EP0940061B1

EP0940061B1 - Distributed network control of a dimmable fluorescent lighting system

Info

Publication number: EP0940061B1
Application number: EP97926443A
Authority: EP
Inventors: Electric Inc. Kumho
Original assignee: Kumho Electric Inc
Current assignee: Kumho Electric Inc
Priority date: 1996-05-13
Filing date: 1997-05-13
Publication date: 2006-07-26
Anticipated expiration: 2017-05-13
Also published as: ATE334571T1; EP0940061A4; DE69736392D1; EP0940061A1; KR20000011044A; AU3121197A; WO1997043876A1

Abstract

An ergonomic and energy-saving dimmable fluorescent lighting system using a distributed network (138) of lighting controls and control power sources. A plurality of controls (84) share a proportional-response control line (58) with a default value, wherein the control requesting the largest deviation from the default value prevails. Certain controls (124) may additionally share a polling line (126), in which the controls must agree on a value, before it can be requested on the proportional-response control line. Interactions between controls on the network provide for a method to control the total illumination on a scene, including both artificial (94) and natural (96) light, down to the light level where only natural light (96) is present. Unidirectional control nodes (86) allow a first network to exert control on a second network, without affecting the control levels in the first network. Thereby, hierarchical network control can be established. In addition, the method provides for network power to energize the control network (63), derived either from the lighting ballasts (52) or from external network-compliant power sources (192) strategically located in the network. A self-energizing contactor (80) switches power to the lighting system (56).

Description

The present invention relates to a method and apparatus for controlling the illumination levels in a dimmable fluorescent lighting system using a multiplicity of controls which interact with each other.
It has long been recognized that higher lighting efficiency can be obtained by lamps employing a gas discharge rather than an incandescent filament. As such, fluorescent lamps have dominated commercial lighting applications for more than fifty years. However, while dimming control is found in a large percentage of incandescent lamp installations, such control is found in only a small percentage of locations with fluorescent lamps. Because lighting is a consumer of a large fraction of the total energy use in many commercial situations, and adding dimming control to lighting systems confers many energy-savings benefits, this lack of dimmable fluorescent lighting systems represents a significant lost opportunity to extract both energy savings and the concomitant environmental savings.
One reason for this low percentage of dimmable fluorescent lighting systems is the lack of good controls at a reasonable cost. The dimmable lighting systems widely available today include the Mark VII and Lutron control protocols. Mark VII dimmable ballasts, which have been widely adopted in the lighting industry, respond to a 0-10 V signal. While this is the most widely available system in use today, the system generally supports only a single manual, photosensor or occupancy control, whereas the most beneficial control involves the use of many controls working in a coordinated fashion.
Another system, available from Lutron, involves a central control box to which many controls and ballasts must be connected. This control box allows for a high degree of flexibility in the control of different fixtures, but in general, installation is expensive since each fixture under control must be connected to a costly central control box, even if the fixture is remotely located.
The dimmable fluorescent lighting systems available today are generally expensive to install, and with the exception of the Lutron system, permit only a limited range of control possibilities. However, it is the availability of flexible control configurations that provides both the energy-savings as well as the additional benefits of ergonomic response.
US 5,357,170 discloses an energy-saving lighting control system that may comprise a plurality of lighting controllers such as a wall box control, an occupant sensor, a photosensor, a time clock and a fire security switch; a programmable lamp controller adapted to receive control signals that are output by the lighting controllers and adapted to output ballast dimming signals; and a plurality of fluorescent lamp fixtures each comprising a fluorescent lamp and a dimming ballast adapted to receive a set of said ballast dimming signals. The lighting control system is selectively operable in either a normal mode or in an off-normal mode. In the normal mode, certain lighting parameters, for example maximum and minimum lighting levels, fade rates, etc. are preset and the lighting level is determined by which of the plurality of ballast dimming signals requires the least electrical power, or by which the lighting level is the maximum requested. In the off-normal mode, for example a calibration or light-adjustment mode, certain parameters are adjustable by manually adjusting a movable member. The micro-processor-based lamp controller is adapted to automatically switch from the normal mode to the off-normal mode in response to an adjustment at the movable member. After a predetermined time period following the most recent manual adjustment at the movable member, the lamp controller stores the new level of the adjusted parameter and returns to the normal operating mode. Preferably, the controller is adapted to provide multiple outputs in multiple protocols, for example high voltage control signals and/or low voltage control signals, by which different types of fluorescent lamp ballasts can be controlled.
In addition, conventional lighting controls are difficult to calibrate during installation and maintenance. For example, photosensor sensitivity must generally be set in the absence of external lighting. This limits calibration in general to evenings, when labor is more expensive. Furthermore, the process is typically iterative, requires manual intervention, and must be performed when the photosensor is in place. This, therefore, requires many ascents and descents on a ladder while gaining manual access numerous times to the photosensor. Adjusting a photosensor properly can take as long as thirty minutes, and the cost of the labor involved can substantially reduce the financial incentive for installing the photosensor. Thus, the clumsy calibration process is a substantial deterrent to the widespread acceptance of many lighting controls.
Furthermore, sophisticated lighting controls, such as photosensors, occupancy sensors, and infrared manual remote controls, require external power for their operation. Frequently, providing this power involves the installation of conduit to bring line-related wiring to the control and step-down transformers to supply power for individual controls. The cost of these installations is frequently prohibitive in retrofit situations, and is very significant compared to the potential energy-savings that could be expected from the use of the controls. The need for external power also has an inhibiting effect on the use of multiple controls to affect the lighting in an area, since each control requires its own low-voltage source. Thus, only those controls which provide the greatest single energy-savings are typically installed in each situation, even though incremental energy-savings and ergonomic benefits could result from the installation of additional controls.
In light of the deficiencies of existing controllable fluorescent lighting systems, it is an object of the present invention to provide greater flexibility in the control of fluorescent lighting systems.
It is also an object of this invention to provide a fluorescent lighting system that can be profitably installed in retrofit situations, as well as new construction.
It is additionally an object of this invention to provide a a fluorescent lighting system that is easy to calibrate.
It is another object of this invention to provide a fluorescent lighting system with improved ergonomic function.
It is a further object of this invention to provide a fluorescent lighting system with energy-savings over systems lacking such control.
It is an additional object of this invention to provide a fluorescent lighting system with a lower cost of installation.
It is still another object of this invention to provide occupants with direct control of the fluorescent lighting systems that illuminate their workspaces.
Additional objects, advantages and novel features of this invention shall be set forth in part in the description that follows, and in part will become apparent to those skilled in the art upon examination of the following specification or may be learned by the practice of the invention. The objects and advantages of the invention may be realized and attained by means of the instrumentalities, combinations, and methods particularly pointed out in the appended claims.
To achieve the foregoing and other objectives and in accordance with the purposes of the present invention, as claimed, there is provided a lighting control system for controlling the amounts of illumination emanating from each of a plurality of artificial lights for illuminating corresponding areas. The lighting control system comprises a plurality of lighting controllers; a plurality of lighting modulators, each of which provides a variable amount of power to the corresponding light and is adapted to vary the power to the corresponding light.
According to the invention, the lighting control system further comprises an electrically-conductive control wire connecting the controllers and the lighting modulators, in which a control signal is shared between the controllers and the modulator, this control signal having a default value. Each lighting controller comprises an uni-directional signal device, wherein each of the uni-directional signal devices in the controller is adapted to change the level of the control signal uni-directionally from said default value, if the controller takes control. Further, the control signal is fixed at the level of the particular lighting controller attempting to set the level of the control signal furthest from the default value. And the lighting modulator varies the power to the corresponding light in response to the value of the control signal.
According to another aspect, as claimed, there is provided a method for controlling lighting in an area using a plurality of lighting modulators and a corresponding plurality of artificial lights, for example gas discharge lamps, wherein the amount of illumination emanating from each light is controlled by the corresponding modulators. The method comprises providing at least one lighting controller chosen from the set comprising occupancy controls, photocontrols, computer controls and remote control receivers.
According to the invention, the methods further comprises the steps of providing common electrical connection between said plurality of lighting modulators and said at least one lighting controller using an electrically-conductive control wire, in which an analog electrical control signal with a default value is established; regulating the electrical control signal with the lighting controllers, wherein all controllers regulate the level of the control signal uni-directionally from the default value; fixing the control signal to a value at the level of the lighting controller that is attempting to set the level at a value that is furthest from the default value; and affecting the illumination in the area by means of the lighting modulators, wherein the lighting modulators are responsive to the level of the electric control signal in the control wire.
In preferred embodiments, the plurality of lighting controls may be chosen from the set including occupancy controls, photocontrols, computer controls, and remote control receivers, in which a plurality of these controls may be commonly housed. These controls may be connected to at least one lighting modulator, which may provide variable amounts of power to at least one gas discharge lamp in order to illuminate an area, by an electrically-conductive control wire connecting the controls and the lighting modulator. In this wire, a control signal may be shared between the controls and the modulators, and this control signal may have a default value wherein each of the controls change the level unidirectionally from the default value. The control signal may be fixed at the level of the lighting control attempting to set the level furthest from the default value, and the lighting modulator may vary the power to its connected gas discharge lamps in response to the value of this control signal. The system may additionally comprise at least one regulated power supply, which may be connected to at least one control through a power wire through which regulated DC voltage power may flow for energizing the controls that are connected to the power wire. An infrared control receiver may be connected to at least one of the controls through a separate infrared control receiver signal wire, which may pass signals from an infrared transmitter to at least one of the controls, and in which the signals may include both mode-changing signals and remote lighting level requests. Furthermore, an electrically-conductive polling wire may connect at least two occupancy sensors. The polling wire signal value may switch from an inactivated value representing the absence of people in the area under surveillance, to an activated value, representing the presence of people in the area, when at least one occupancy sensor connected to the polling wire has detected people within the area. This polling signal activated value may force all occupancy sensors connected to the polling wire to attempt to modulate the signal in the control wire to a lighting level suitable for occupied space. The apparatus may contain at least one power wire, connected to AC line voltage, supplying power to at least one of the lighting modulators. A self-energizing line contactors may be used to switch AC line power to at least one of the modulators, where the energy used in controlling the contactor may be stored transiently from a line-related AC source within an earth-ground referenced energy-storage device and may be released through an opto-isolator when the contactor is in a switched state, so that the opto-isolator may trigger an electric switch so as to energize a relay coil using a source of line-related AC voltage used in powering the lighting modulator.
Non-limiting modes for carring out the present invention will be the following detailed description with reference to the drawings, in which:

Fig. 1 is a block diagram of a local control network, in which the present invention can be applied;
Fig. 2 is a block diagram of a local control network connected with a shared proportional-response control line according to the invention;
Fig. 3 is a block diagram which depicts a shared proportional-response control system as shown in Fig. 2, additionally with polling of polling-responsive environmental controllers;
Fig. 4 is a block diagram depicting a polling wire control system as shown in Fig. 3, showing the internal components of the polling-responsive controllers;
Fig. 5 is a block diagram depicting components and connections in a uni-controller not representing the present invention and shown as an illustrative example;
Fig. 6 is a block diagram of a dynamic illumination stabilizer not representing the present invention and shown as an illustrative example;
Fig. 7 is a schematic of a photosensor with two loop-response speeds operating without dynamic illumination stabilization.
Fig. 8 is a schematic of a dynamic illumination stabilizer as shown in Fig. 6, with dual loop speeds and automatic calibration, not representing the present invention and shown as an illustrative example;
Fig. 9A is a schematic of an dynamic illumination stabilizer with dual loop speeds and automatic calibration, carried out by means different from that of Fig. 8, not representing the present invention and shown as an illustrative example;
Fig. 9B is a schematic of a dynamic Total Effective Illumination stabilizer with dual loop speeds and automatic calibration, not representing the present invention and shown as an illustrative example;
Fig. 10A is a schematic of a motion sensor allowing remotely actuated calibration, not representing the present invention and shown as an illustrative example;
Fig. 10B is a schematic depicting the output stage of the occupancy sensor of Fig. 10A;
Fig. 11 is a block diagram depicting a lighting network with a single level of control, not representing the present invention and shown as an illustrative example;
Fig. 12 is a block diagram depicting a lighting network with three levels of control, not representing the present invention and shown as an illustrative example;
Fig. 13a through 13g are schematics depicting alternate embodiments of unidirectional signal devices not representing the present invention and shown as an illustrative example;
Fig. 14 is a schematic of a uni-controller motherboard not representing the present invention and shown as an illustrative example;
Fig. 15 is a schematic exploded top-view of the physical layout of the uni-controller shown in Fig. 5 not representing the present invention and shown as an illustrative example;
Fig. 16 is a cross-sectional schematic of the uni-controller of Fig. 15, taken along line 16-16 shown in Fig. 15 not representing the present invention and shown as an illustrative example;
Fig. 17 is a block diagram depicting a lighting network similar to that of Fig. 12, in which compliant power supplies are used to power the network, not representing the present invention and shown as an illustrative example;
Fig. 18A is a block diagram of a dimmable ballast with a line-isolated control interface, not representing the present invention and shown as an illustrative example;
Fig. 18B is a block diagram of a lighting fixture with dimmable ballasts, in which the line-isolated control interface is not integral with the dimmable ballast, not representing the present invention and shown as an illustrative example;
Fig. 19 is a schematic of an electronic ballast which is designed to provide power for gas discharge lamps and simultaneously provide line isolated, low voltage power for external use, not representing the present invention and shown as an illustrative example;
Fig. 20 is a schematic of an alternate embodiment of an electronic ballast, involving the use of added magnetic components that are not used to power lamps, not representing the present invention and shown as an illustrative example;
Fig. 21 is a schematic of another embodiment of an electronic ballast, involving the use of added magnetic components that are not used to power lamps, as well as an added oscillating circuit, not representing the present invention and shown as an illustrative example;
Fig. 22 is a schematic of a further embodiment of an electronic ballast, involving the use of added capacitive components not representing the present invention and shown as an illustrative example;
Fig. 23 is a circuit diagram of a line-isolation control and power interface not representing the present invention and shown as an illustrative example;
Fig. 24 is a schematic of a compliant, current-limited power supply for use as a distributed power source in a lighting control network, not representing the present invention and shown as an illustrative example;
Fig. 25 is a schematic of a self-powered contactor energizing circuit, not representing the present invention and shown as an illustrative example;
Fig. 26A is a flow chart of a method for calibrating the photosensor of Fig. 8; and
Fig. 26B is a flow chart of a method for calibrating the photosensor of Fig. 9A and the photosensor of Fig. 9B.

The present invention involves a network of lighting controls interacting together to exert control over a dimmable fluorescent lighting system. In order for the network to function, a method is described for governing the interaction of controls with each other and with the controllable ballasts. Furthermore, for maximum control flexibility, illustrative methods for hierarchical control configurations not representing the invention are described. In addition, illustrative methods not representing the invention are described for the controls to obtain described for the controls to obtain power in order to function. Because of the broad availability of control interaction made possible by the present invention, a number of novel control methods also not representing the invention arising out of this interaction are then possible. These novel and useful methods are sequentially described as illustrative examples in this specification.

Local Control Network Overview Functional Description

Control Modes Within the Network

There are a variety of controls which may be present within a lighting control network. Four types of controls are typically found in lighting systems, and the methods by which these controls can operate within the present invention are given further below.

Manual Control

Manual control allows the user direct control over illumination. This direct control may be exerted either through direct electrical connection, for example, with a wall dimmer, or through infrared remote control. In either instance, the user attempts to set the illumination level through an allowed range from full manual illumination to full dim (i.e. the minimal artificial illumination allowed by the ballast). The manual control in control networks of the present invention may incorporate any of a number of different means of user input, including manual adjustment infrared remote control, voice recognition, telephone input, and local computer control.

Photo Control

Photosensors detect the amount of light impinging on the work area from the vantage of the sensor, which is typically located at the ceiling, collecting light from below. The light from the work area may be comprised of both artificial as well as natural light, if the work area receives direct or indirect lighting from windows or skylights. The object of photosensing is generally to reduce the amount of artificial light, and the corresponding energy use generated by the lighting system, to the extent that natural light provides adequate illumination. This method of reducing energy consumption is frequently called "daylight harvesting."
The process of daylight harvesting requires sensitive calibration of the photosensor. The object of the calibration is to adjust the operating point of the photosensor. Typically, the photosensor is adjusted such that the amount of artificial lighting is reduced as soon as levels of light exceed that of the maximum artificial lighting on its own. In order to set this calibration point, calibration is typically carried out at night, when the photosensor can observe the scene with only artificial lighting present. This method has a number of significant disadvantages, including very lengthy and expensive calibration. However, the most significant limitation is that the system is inherently set to provide only a single predetermined amount of light, typically that of full artificial illumination brightness.
The method of the present invention makes use of the cooperation of the photosensor with either direct-manual or infrared remote manual control. Instead of the prior art, in which a single predetermined illumination level is maintained by the system, the method of the present invention provides control to the user of this level. Thus, while working at a computer terminal, the user may request lower light levels of combined artificial and natural light than during desk work. This form of lighting control depends on the network capabilities of the present invention, which allows manual input control to the photosensor. This form of total illumination control is called "dynamic illumination stabilization."

Occupancy Control

Occupancy sensors indicate the presence of an animate object in an area, sensing presence either through an infrared and/or a sonic detector. Because presence is usually detected through movement of the person sensed, occupancy detectors are often called motion detectors. The purpose of such sensors in a lighting system is to reduce the lighting level and thus the energy use when the illuminated area is unoccupied. Until recently, occupancy detectors only switched the controlled lights on when motion was detected. Switching fluorescent lamps on and off frequently, however, significantly reduces lamp life, and therefore concomitantly reduces the financial savings of using an occupancy sensor.
By placing occupancy sensors in networks containing controllable ballasts and incorporating a novel control logic, the method of the present invention provides the more appropriate ergonomic responses. Most particularly, dual timing methods are used to dim lamps slowly while restoring lighting much more rapidly. Dimming is performed slowly so that if a person is present but not detected by the motion detection circuitry, the person may make his presence known to the system by deliberately moving, and thus prevent the lights from placing him in relative darkness. However, when the illumination is at a dim level due to the action of the occupancy sensor, and an occupant is then detected, the illumination is restored to original level rapidly. This prevents, for example, a large delay between entering an unoccupied room and the time at which higher levels of illumination are restored, ensuring that the person does not enter a darkened room.
There are a number of additional limitations with prior art occupancy sensors. For example, in certain lighting scenarios, it is difficult for a single occupancy sensor to detect motion within the entire area of the lights it controls. Thus, in long or "L" shaped hallways, bathrooms, or large work areas, it would be useful to employ multiple occupancy sensors cooperating to sense occupancy anywhere within the area. The present invention describes the logic needed for networks containing multiple occupancy sensors. In functional overview, the logic requires that all occupancy sensors agree that the area under surveillance is unoccupied before the lights could be dimmed. Thus, if a person was detected by only a single occupancy sensor, the other sensors could not dim the lights unilaterally without agreement from the other linked sensor.
Another form of cooperation that draws from the present invention involves interaction with manual controls, whether direct line or infrared remote, to allow remote calibration of occupancy sensors. Conventional occupancy sensors require the user to set physical jumpers, dip switches, or potentiometers within the occupancy sensor in order to set the level of sensitivity. Because this requires physical interaction with the occupancy sensor, and because the occupancy sensor must work with the user at some distance from the sensor, the effort in setting the control is laborious and insensitive. In many occupancy sensors, only a few choices of sensitivity are offered, or require many separate ascents and descents on a ladder to select the proper sensitivity.
The method of the present invention, by facilitating remote commands input to the occupancy sensor, permits novel methods of occupancy sensor control. Using commands entered via an infrared remote control, the user can present movements that are instructed to be at the limits of the desired occupancy sensor range, which allows fine distinctions to be made. Such fine sensitivity could, for example, resolve the needed distinction between a person walking outside of an office door (which should not activate the occupancy sensor) and the moment the occupant enters the door (when the occupancy sensor should activate).
The same logic described here for calibration would be appropriate whether occupancy is detected using sonic and/or infrared detectors. The methods of the present invention for cooperation between occupancy sensors and for the remote calibration of occupancy sensors would operate even with the use of radio frequency or other identifying tags, in which the occupant wears a short-range identification tag which is detected by a receiver in the room.

Energy Management System Control

Local lighting control networks can be as small as a part of an office, or they may be as large as an entire building. For purposes of describing the present invention, a local network is considered to be any network where all of the controls in the network cooperate to control the same ballasts. Hierarchical networks, to be described later, have multiple and complex groupings of controls and the ballasts that they affect. One type of control over a local network is that of a computer or programmable microcontroller arranged to manage energy, generally for purposes of providing financial returns through energy savings. These controls are frequently called energy management systems (EMS), and may operate either over suites of rooms, floors of a building, or building-wide.
One drawback of many of the commercially available energy management systems is that the EMS removes local control from the user. Thus, energy savings may come at the cost of poor ergonomics, worker satisfaction and productivity. With the method of the current invention, energy management systems may be devolved to smaller networks, where individual workers or groups of workers may vary illumination within the bounds of energy maximums or "caps" mandated by the EMS, or even override the EMS when necessary. User input could be effected either through direct computer interaction, for instance at a keyboard or using a mouse, through a scripted telephone interaction in conjunction with a computer connected to the phone system, or other such means. The computer could then interact with the network directly, or through a power line transmission protocol, which utilizes frequency-encoded information.

Shared Proportional-Response

The previous discussion has presented some of the controls which may take part in a local control network. The discussion immediately following presents the modes by which these controls may interact within a local network. While there are a number of specialized modes of control to be discussed later, the most widely used protocol is that of a shared proportional-response system.
Fig. 1 is a block diagram of a local control network, in which all fixtures within the local control network are affected in the same manner by their associated controls. In this figure, a shared control signal is carried over a control wire 58 with electrical connections to both ballasts and controls. Other connections in this network, including electrical ground and control power connections, are not shown. A plurality of fluorescent lamps 50 are powered by a plurality of dimmable ballasts 52. Each of these ballasts 52 contains a line-isolated control interface 54, which transmits the shared control signal from the control wire 58 to the dimmable ballast 52. The fluorescent lamps 50, dimmable ballast 52, and line-isolated control interface 54 are all located in a plurality of fluorescent lighting fixtures 56, represented by the enclosing dotted boxes.

The method of the present invention is to provide a scheme whereby, among a plurality of sensors which provide proportional-response control, the sensor requesting the lowest lighting level prevails. This "OR" logic naturally sets the lighting level to that which is most ergonomically appealing as well as most energy-conserving. Consider, for instance, an office in which a manual dimming control with pilot light and on/off switch 60, a photosensor 64. an occupancy sensor 62, an infrared control 61. an infrared remote receiver 59, and a local computer control 63 are simultaneously actively responding to both the user and the environment. The infrared remote receiver detects signals from a manual remote transmitter operated by the room occupant, or any other local infrared transmitting device which may be automatic, and passes commands to the photosensor 64, the occupancy sensor 62, and the infrared control 61 via a first photosensor supervision wire 65, an occupancy sensor supervision wire 67 and an infrared control supervision wire 69, respectively, which are used to either calibrate or adjust the normal operation of each control. The following table tracks the scenarios described below, in which the combined manual and infrared manual remote signals are collectively referred to as manual.

control inputs				light output	scenario
Manual	light sensor	motion sensor	Building EMS	light output	scenario
90%	100%	100%	100%	90%	room set to user preference
90%	50%	100%	100%	50%	sun brightens: illuminates room
90%	100%	10%	100%	10%	room unoccupied
90%	100%	100%	60%	60%	janitor cleaning at night
40%	100%	100%	100%	40%	working at computer: low light desired

As the user enters his office in the morning, using either the manual dimmer 60 or the infrared control 61, he will set the lighting level to a comfortable level, which may be less than the full lighting level -- for example, 90% of the maximum. Changes in the lighting level are produced by changes in the fluorescent lamps 50, which receive the appropriate power levels from the dimmable ballasts 52, which in turn receive the local network lighting level signals from the control interface 54. As long as the occupancy sensor 62 detects the user presence, and the ambient, external (natural) lighting is insufficient to meet the occupant's desired lighting level, the light output will be set to the 90% level. If the external lighting increases substantially (e.g. the sun shines in the window), there may be enough ambient lighting to satisfy some fraction of the user's request. If the light sensor 64 now determines that it needs only 50% artificial light output to maintain the user's previous request of 90% of maximum artificial lighting, the lighting system will be reduced to the 50% level. If the user leaves the room, the occupancy sensor 62 will detect non-occupancy and command that the light output be reduced to some predetermined level, in this case 10%. Generally, decreases in lighting will be gradual, so that in case the occupancy sensors 62 fail to identify a user who is in fact present, this gives the user time to make his presence known to the sensor 62. When the user returns, the system rapidly readjusts itself to deliver the lighting set by the dimmer 60, the infrared control 61, and the light sensor 64.
A second photosensor supervision wire 71 connects the infrared control 61 to the light sensor 64 through an infrared control mode switch 134. The second photosensor supervision wire 71 carries a signal equivalent to the total light level requested by the user for dynamic illumination stabilization that is detected by the infrared remote receiver 59 and transferred to the infrared control 61, where it is maintained. That is, the user sets the total amount of light desired in the area of control, including artificial and natural light, in the manner described above.
The photosensor 64 is also connected directly to the infrared manual remote receiver 59 via the first photosensor supervision wire 65, which carries requests from the user, the lighting installer, or the maintenance worker to automatically calibrate the photosensor 64.
The occupancy sensor supervision wire 67 connects the infrared remote receiver 59 to the occupancy sensor 62. The first photosensor supervision wire 65 carries signals from the user, via the infrared remote control receiver 59 to the occupancy sensor 62, placing the occupancy sensor 62 into calibration mode.
In the absence of dynamic illumination stabilization, commands received by the infrared remote receiver 59 are transferred to the infrared control 61 by infrared control supervision wire 69, where they are maintained and transferred directly to the control wire 58 by switch 134 in the second position, without interaction with the photosensor 64.
In Fig 1, the shared proportional-response control signal is transmitted along control wire 58, which in the preferred embodiment carries a low-voltage DC voltage signal between the controls and the ballasts 52. In a simple installation, the control wire 58 is a pair of wires (optionally in conjunction with other wires carrying power and power return, as will be described), one carrying the signal and the other a signal return. The manner by which separate controls coordinate the control signal established in control line 58 is described in the next section.

Physical Implementation Of Local Control Network Features

Shared Proportional-Response Control System

Fig. 2 is a block diagram of a local control network connected with a shared proportional-response control line according to the present invention. A plurality of environmental controllers 84 is connected to the electrically conductive control signal wire 58 via a pull-down diode 86 which allows only unidirectional influence over the voltage potential of the wire. Each of a plurality of light modulators 88 contains a pull-up resistor 90 connected to a source of voltage 92, denoted by V+. If no environmental controller 84 is active in controlling the potential of the wire 58, the pull-up resistors 90 keep the voltage potential at the default value, in this case, V+ of the voltage source 92. When this voltage potential is at the default value, the output of a plurality of artificial lights 94 are adjusted by the light modulator for full illumination. If any of the environmental controllers 84 takes control by lowering the potential of the cathode of its pull-down diode 86, the potential of the signal wire 58 is also lowered to the potential of the anode of that controller's pull-down diode 86. Then, if any other environmental controller 84 reduces its pull-down diode 66 cathode potential beyond that of the first controller, the potential of the electrically conductive wire 58 is lowered beyond the influence of the first controller. Thus, control of the potential of the electrically conductive wire 58 is passed to the environmental controller 84 with the lowest internal potential.
It should be noted that devices other than the pull-down diode 86 may allow unidirectional control over the signal in the electrically conductive wire 58. Devices which permit such unidirectional control may be called "unidirectional signal devices" and, subsequently, a number of the electronic circuits which behave as unidirectional signal devices, such as unidirectional current gates and current buffers, will be described.
Each light modulator 88 is arranged so that the amount of light produced by the artificial light 94 is related to the control signal in the signal wire 58 in the following way. When the control signal is at the default value, the amount of illumination emanating from the light 94 is maximal. As the control signal varies from the default, the amount of illumination from the light 94 is reduced, the amount of reduction determined by a gain factor set within the light modulator 88 in a manner to be described later.
It should be understood that there are a number of alternative arrangements, devices and components, polarities of potential, or other control transmission protocols that embody the spirit of this invention. The sensors within the environmental controller 84 that utilize the shared proportional response control line can include a variety of types, including manual controls, motion/occupancy detectors, remotely-adjusted controllers (e.g. from infrared-mediated or other communication devices), digitally-addressed electronic controllers, controllers directed by digital computers, or building- or area-wide energy management systems. It should also understood that the control signal on the control wire 58 need not be voltage, and other schemes which affect current levels are also envisioned. Alternatively, the control signal could be encoded in the frequency modulation of an alternating current, which could be adjusted in the control line. Indeed, several of the herein suggested means could be employed together to simultaneously implement multiple modes of control. The essence of the teachings of the present invention related to the shared proportional-response control line are that whichever controller 84 attempts to set a control signal in a shared control line the furthest from some predetermined default value asserts control over the shared signal, and therefore, the amount of light from the light modulator 88.
It is of considerable value that this simple logic naturally adjusts the level of lighting in a system with multiple controlling devices such that an ergonomically-correct light output results. Furthermore, because of the independent logic control over the network by each sensor, the "intelligence" of the network is distributed. This means that the connectivity of the network can be quite varied, reducing the amount of effort required in designing the network, as well as the amount of labor and wiring required.
It should be noted that the control wire 58 connecting the controls does not also carry power for the controls in the preferred embodiment. Instead, a control power wire 66 carries power for the controls. As shown in the diagram, all of the controls may share the same control power wire 66. It should be noted that depending on the nature of the signal, control signal wire 58 and control power wire 66 may be the same. For example, if a single wire carrying both signal and power were employed, the signal could be carried by frequency modulation, with the power supplied by a DC component.

Polling Signal Control Coordination

An exception to the shared-proportional response control logic described above is that which occurs when multiple occupancy sensors are scanning an area. For example, a long hall might have occupancy sensors at both ends. While each has a field of view that overlaps the other, neither can cover the entire area. In such a case, the proper control requires a polling mechanism -- that is, the hall can be considered empty only when both occupancy sensors indicate so. If only one occupancy sensor detected an empty hall, one would not want it to lower the lighting according to the shared-proportional response control logic described above, since the person in the hall might only be detectable by the other occupancy sonsur. In this case, a means must be provided to allow the occupancy sensors to "poll", one another to determine whether all of them agree that the area under surveillance is empty.
The hall could still be controlled by other sensors - e g. dimmers, light sensors, or building-wide EMS systems using the shared-proportional response control logic. It is only the occupancy sensors that must be polled. The example given here is but one of several multiple-sensor scenarios that benefit from a polling signal.
Fig. 3 is a block diagram which depicts a shared proportional-response control system according to the invention with polling of polling-responsive environmental controllers. In addition to the environmental controller 84, which could be a photosensor, a plurality of poll-responsive environmental controllers 124 are connected to the network through the shared proportional-response control line 58. As in the shared proportional response control system described above, each environmental controller 84 and 124 is connected to the electrically conductive wire 58 via a pull-down diode 86 which allows only unidirectional influence over the voltage potential of the wire. A subset of the environmental controllers are poll-responsive to a polling signal shared by them. A polling wire 126 is present through which the poll-responsive controllers 124 communicate in such a way as to act as a single element in the system, in response to stimuli that affects only one or some of the poll-responsive controllers 124. The circuitry of each poll-responsive controller 124 is so arranged that it prohibits control of the control signal wire 58 unless all poll-responsive controllers 124 agree that control is required. As an example, if one or more of the poll-responsive environmental controllers 124 is responding such that it would not pull-down the potential of the control signal wire 58, the polling wire 126 inhibits all linked poll-responsive controllers 124 from pulling down. Only when all linked, poll-responsive controllers 124 agree to pull down the potential of the control signal wire 58 does the polling wire system allow the potential to be lowered. When this potential is lowered all the light modulators 88 attached to the control signal wire 58 modulate the light output of the attached lights 94 accordingly.
The polling medium, in this case a polling wire, is shown separate from the shared proportional response control wire 58, but it could be a separate signal sharing the same control signal wire 58. Such a sharing of the network control wire 58 would reduce the cost and complexity of installing the network, although it could increase the cost and complexity of those sensors that are part of the polling network. Alternatively, the polling medium could include other means of communication, such as infrared light or radio transmission.
As in Fig. 3, the control wire 58 connecting the controls does not also carry power for the controls in the preferred embodiment. Instead, a control power wire 66 carries power for a plurality of controls. As shown in the diagram, one of the poll-responsive environmental controllers 124 shares the control power line 66, while the other poll-responsive environmental controller 124 has an external power source 125 that powers only the controller 124 to which it is connected. The choice of whether or not to share powering means depends in part on the nature of the control the distance and configuration of control connections, the availability of access to line power, and other considerations.
Fig. 4 is a block diagram depicting a polling wire control system according to the invention. Resistors R1C and R2C are illustrative of several pull-up devices which could keep the potential of the control signal wire 58 at the voltage source 92 of V+, when not influenced by any controller. Two polling-responsive controllers 124 are shown. For illustration purposes, each controller 124 shown is a motion detector. Within each controller 124 is a pull-down diode D1C which allows only one controller at a time to influence the potential of the control signal wire. Each motion detector 124 contains a motion detection circuit 128, whose designs are well known within the prior art. These circuits 128 may include infrared or sonic detectors or both infrared and sonic detectors or other effective sensing techniques. Within each circuit is an integrated circuit U1C, which represents the output stages of the motion detection circuits 128 and which is arranged to pull down the base of transistor Q1C, turning it on when motion is sensed. The potential at the base of a transistor Q3C is that of the collector of the transistors Q1C. The transistors Q3C are PNP emitter followers which influences the potential at the cathode of the diodes D1C. The polling wire 126 provides a link between the collectors of the transistors Q1C within each poll-respeinsive controller 124, and the like collectors of any other linked, poll-responsive controllers 124 on the polling wire 126. If motion is detected by any integrated circuit U1C, the collector of any transistor Q1C respectively will, by the link provided by the polling wire 126, force high the potential of the other collector and any further collectors in other poll-responsive controllers 124 that are so linked. This inhibits any linked poll-responsive controllers 124 from lowering the potential of the control signal wire 58. Therefore, only when all linked poll-responsive controllers 124 agree that there is no motion can the potential of the control signal wire 58 be influenced.
Such a polling mechanism would be of particular use in a situation where a plurality of motion/occupancy detectors were operating. Only in the case where all of the motion detectors indicated that the area was unoccupied would the sensors set the shared proportional response control line low, appropriately reducing the lighting.

Control Coordination and Illumination Stabilization

The present invention teaches the use of multiple control within a local lighting system network to improve energy-savings and provide a comfortable and productive work environment. In order to obtain the largest advantages of the present invention, two to three controls will be installed within a room or work space. Often, the optimal location for photo, occupancy, and manual dimming controls is just over the work area, since from this vantage point, the occupancy control can easily monitor personnel, the photo control can regulate light impinging on the work surface, and the receiver for a remote control dimmer is close to the people affected.
The co-placement of multiple controle at a single location encourages the close integration of the controls into a single housing, with shared electronic and physical components. One embodiment of the present invention, based as it is on the cooperation of multiple controls, includes a common controller that incorporates multiple control devices. As will be seen, this not only simplifies installation and reduces cost, but also provides for novel interactions between different control devices. The unified controller incorporating multiple controls within a single housing will be hereinafter referred to as a uni-controller. It is understood, however, that the physical placement of cooperating controls into a single housing is simply a convenience, and the methods of the present invention work as well when the controls are located in separate housings.
Fig. 5 is a block diagram depicting components and connections in a uni-controller 138 incorporating multiple controls shown as an illustrative example and not representing the invention. An internal control bus 142 is influenced by the outputs of the photosensor 64, the occupancy/motion sensor 62, and the infrared control 61. As we will see later, the internal control bus 142 functions as a shared, proportional-response control line. Furthermore, the uni-controller 138 contains the infrared remote receiver 59, which receives controls from the user who enters commands through an infrared transmitter, or other infrared transmitter means (not shown). Each of these controls is connected to the internal control bus 142. which functions as a shared proportional response control line to allow the control requesting the lowest control response to dominate. In order to maintain the shared proportional response line, each of the control attached to the internal control bus 142 will contain a unidirectional signal device to permit unidirectional control over the signal in internal control bus 142. Controls without internal unidirectional signal devices will be attached to the internal control bus 142 through an external unidirectional signal device.
The control value present at the internal control bus 142 is communicated to light modulators connected to the uni-contraller 138 through a connector J7. The uni-controller 138 is constructed so that from one to three of the controls are emplaced in the uni-controller 138 housing. When the motion sensor 62, the photosensor 64, and the infrared control 61 are all emplaced in the uni-controller 138, the uni-controller 138 is configured as is shown in Fig. 1. In general, the infrared remote receiver 59 will be in place in order to receive commands from the external transmitter sending signals to controls resident within the uni-controller 138.
The infrared control mode switch 134 sets the functioning of the infrared control 61 and the photosensor 64. When the infrared control mode switch 134 is in the "DN" position, so that the infrared control 61 is directly connected to the control bus 142, the infrared control 61 functions very similarly to a manual dimmer, maintaining dimming signals from a remote infrared transmitter received by the infrared remote receiver 59. When operating simply as a manual dimmer, the infrared control 61 sets the control signal of the internal control bus 142 directly.
However, when the infrared control mode switch 134 is in the "UP" position, connecting the infrared control 61 directly to the photosensor 64, the infrared control 61 acts to fix the point at which the photosensor 64 regulates the light output. In this way, the user does not set the amount of light output by the artificial light sources alone, but rather sets the total amount of illumination, comprised of both natural and artificial in the region sensed by photosensor 64. As mentioned before, this method of lighting control may be called "dynamic illumination stabilization".
It should be noted that alternative arrangements and packaging of controls will have similar function. For example, the infrared remote receiver 59 and the infrared control 61 may be physically and electronically integrated. Furthermore, the infrared control mode switch 134 may be absent, with a preferred function (either direct dimming or dynamic illumination stabilization) for infrared control 61 permanently fixed by the wiring layout.
An inter-controller control wire 165 connects the connectors J1, J2, and J3 with other lighting controllers, and J7 with other controllers or with ballasts. These inter-controller control cables 165 provide a pathway for control signal communication between uni-controllers, other lighting controllers, and ballasts in a manner to be described later.
Fig. 6 is a block diagram of a dynamic illumination stabilizer shown as an illustrative example and not representing the invention. A light transducer 96 responding to input light from the artificial light 94. as well as natural light 95, generates a signal at its output. The light transducer may be any element which reacts to variations in lights energy with a changing electrical characteristic. Such transducers may include photodiodes, photoresistors, or any photosensitve device which may be further associated with circuitry which transforms or amplifies the output signal to a signal analogous to the light stimulus. A dual-speed error amplifier 98 compares the signal output by the transducer 96 to a reference voltage 100 and generates an output signal with a polarity and slew-rate magnitude proportional to the difference between the light transducer 96 signal and the internal reference voltage 100 (the "error"). The term dual-speed refers to the fact that the proportionality constant relating the slew rate of the dual-speed amplifier 98 to the "error", as defined above, may be set to two distinct values, thus affecting the slew-rate for a given error. It is understood that the proportionality constant may be alternatively set to a multiplicity of values, or may be continuously adjustable.
The voltage reference 100 signal may be overridden by a signal communicated over an external illumination stabilization signal wire 136, such that the external signal can take control of the light level regulation point because the output impedance of the voltage reference 100 is sufficiently high to allow an external control signal from signal wire 136 of lower impedance to override it.
It is understood that the external illumination stabilization control signal emanates from the infrared control 61, and is transmitted to the photosensor 64 through the infrared control mode switch 134 in the "UP" position, thence to the error amplifier 98 through external illumination stabilization signal wire 136. Thus, the user sets the regulation point for the error amplifier 98 through infrared remote control. When the infrared control mode switch 134 is in the "DN" position, no external signal is received through signal wire 136, and the resident voltage reference 100 establishes a single regulation point for the photosensor 64.
It is further understood that the illumination stabilization signal need not be sent from the infrared control 61, and that signals may be alternatively sent from a manual dimmer arrangement, through computer control, or any other such means for establishing a voltage control signal of lower impedance than the voltage reference 100 in the signal wire 136. Also, while the physical proximity of the photosensor 64 and the infrared control 61 within uni-controller 138 facilitates the communication of the control signal through the signal wire 136, other arrangements are possible in which the photosensor 64 and the infrared control 61 are remote from one another.
The output of the dual-speed error amplifier 98 is examined by a window comparator 102 that determines if the error amplifier output is within a given linear range. The output of the window comparator 102 drives a threshold indicator LED D2K to indicate whether the signal is within the detection range of the window comparator, for purposes of manually adjusting the photosensor light response. When setting the control loop saturation thresholds for the photosensor, it is desirable to have an indication of the state of the control loop. This may be had by observing the light output response of the system in response to movement of the control set point. In the present invention, an error threshold indicator is provided which indicates when the system has exceeded either the minimum or maximum loop saturation threshold, or both, of the system. In general, this indicator will be a light source visible from the outside of the control enclosure, so that the technician adjusting the error threshold can perceive the threshold while the control is in an operational state. There are a number of different ways in which the light could indicate the threshold positions, but in the preferred mode, the indicating light source is illuminated between the minimum and maximum error signals, and switched off when the loop is in saturation.
Switch S1E represents both a manual as well as an electronic switch, which when closed increases the slew-rate of the error amplifier 98 for a given input error signal. When used as a manual switch, the increased slew-rate of the amplifier serves to assist circuit adjustment and functional testing relative to that found during normal operation. In normal operation, the slew-rate is made slow, so that rapid changes in ambient light do not cause large, sudden changes in the amount of artificial light, which would be uncomfortable for the user. However, during adjustment of the saturation thresholds for the photosensor, such slow slew rates lengthen the amount of time required for calibration. Thus, a fast slew-rate is advantageous during manual adjustment. In one form, the system may function only as a simple photo control, without dynamic illumination input or electronic actuation of switch S1E by the window comparator. In this simple form, S1E may still benefit operation through electronic actuation during initial energization, when a rapid slew-rate facilitates rapid system stabilization.
With dynamic illumination stabilization, the switch is also electronically adjustable. The window comparator 102 connected to the output of the dual-speed error amplifier 98 is designed to detect when the output of the error amplifier 98 is outside of a pre-determined range, or is slewing faster than a pre-determined speed which indicates a high rate of change of either the light level impinging on the light transducer 96 or high rate of change of the external control voltage transmitted through signal wire 136. If either of these circumstances occur, the window comparator 102 closes the switch S1E. increasing the loop speed. The window comparator 102 is designed to keep the switch S1E closed for a period of time after the error amplifier 98 comes out of saturation or stops slewing rapidly. This allows the amplifier 98 to return quickly to regulation after the condition that caused saturation terminates, or after a significant change in external input control voltage.
The electronic control of the dual-speed error amplifier 98 slew-rate through the switch S1E allows changes made by the user in illumination stabilization to be immediately registered throughout the system, affecting the output of the light 94. Such user initiated changes will rapidly change the external input control voltage through wire 136, affecting the closing of the switch S1E. Furthermore, this electronic control of the switch S1E allows the system to rapidly reestablish equilibrium when the equilibrium is disturbed. The system often will be far out of equilibrium when the lights 94 are just turned on, since the light modulators 88 might be adjusted to provide full illumination from lights 94 on startup. A further purpose of this adjustment is so that the user obtains immediate feedback that the lights are responding to his command, and that the lighting system is not defective. However, it is further ergonomically pleasing for the system to rapidly come into regulation, and to allow the user to set the desired lighting level. In this case, rapid re-establishment of regulation is provided by the electronic setting of the switch S1E as the illumination detected by the system rapidly changes during the system startup. The resulting rapid equilibrium could be provided by means alternative to switch S1E, including a variety of combinations of both digital and analog circuitry.
The conditions under which the electronic closing of the switch S1E is performed must be set so that during normal operation of the lighting system, the slew-rate of the dual-speed error amplifier 98 is not constantly changing. Limiting the electronic closing of the switch S1E to cases where the lighting changes more than 5% in less than 500 milliseconds provides for regulation pleasant to the observer. Switch S1E should then normally open within one second of closing under the conditions described above. Of course, in different office and work environments, settings other than this may suffice.
It should be noted that this system is a closed feedback loop in which the dual-speed error amplifier 98 generates a control signal which influences the light modulator 88, which adjusts light output from artificial light 94, which affects the incident light impinging on light transducer 96. This signal is then input to the error amplifier 98 to complete the loop. This feedback loop acts to regulate the total incident illumination on the scene sensed by the light transducer 96. The photosensor is adjusted to provide the desired light level by adjusting the value of the reference voltage 100, and therefore the amount of light that must impinge on the light transducer 96 to provide equilibrium at the input error amplifier 98.
The present invention teaches that the window of the window comparator 102 can be set to encompass all signals that are valid only if the loop is in regulation and thereby detect if the loop is in regulation or in saturation. As shown, the threshold indicator LED D2K is illuminated only during saturation. The circuit can be easily arranged, however, so that the LED D2K is illuminated for other indication is made active) only during normal regulation. In this case, the indicator LED 104 serves a secondary purpose in that it also indicates that operating power is present within the circuit, thus serving a dual function of both indicating operating power and normal loop regulation.
It should be understood that the threshold indication can function via a variety of different alerting means other than through LED D2K, including audible or electrical indicators. Furthermore, the error signal indicators need not only be the maximum and minimum attainable values of the error signal, but may also include other values as well. It might be beneficial in certain circumstances to include multiple indicator lights, such as one to indicate the maximum attainable value, and another to indicate the minimum attainable value, which would reduce the potential ambiguity to the technician servicing the photosensor. It should be understood that it is advantageous to close switch S1E when the unit is first energized, when large changes in signal from any source are encountered, or during recovery from loop saturation.
Fig. 7 is a schematic of a photosensor with two loop-response speeds operating without dynamic illumination stabilization shown as an illustrative example and not representing the invention. This method corresponds to that of Fig. 6, wherein the voltage reference line 136 terminates at the voltage reference 100, and does not accept outside inputs. A photodiode D1R in conjunction with a current-to-voltage converter 106 comprise the light transducer 96, in this case a light-to-voltage converter, with an output at a node E5K. In normal operation, a plurality of switches S1K and S2K are open. Light related voltage at a node E5K is compared with reference voltage at a node E1K generated by the reference voltage element 100, and seen through a resistor R3K at the non-inverting input of an error amplifier 97. The error amplifier 97 output responds by slewing with a speed and direction relative to the difference in magnitudes of voltages at the nodes E5K and E1K. A capacitor C1K and a resistor R1K define the slew-rate of the amplifier 97 for a given error input. The output of the amplifier 97 is current buffered by a unidirectional signal device 130 and provided to output at an output node E4K. Closed loop operation is achieved by connecting the output node E4K to the control input of an external variable light source with an intensity responsive to the voltage at the output node E4K, and with the source of light at least partially illuminating the diode D1R. The circuit is so arranged that an increase in light impinging on the diode D1R reduces the light output from the external, variable light source such that equilibrium is maintained. The resistor R4K adjusts the reference voltage, and hence, through loop regulation, the amount of illumination impinging on the diode D1R.
The capacitor C1K and the resistor R1K provide a slew-rate control for the amplifier 97 which is appropriate for normal operation. When calibration or adjustment is required, the loop speed, dictated by the feedback components of the amplifier 97, can be modified by the closure of the switch S2K. With the switch S2K closed, the resistor R2K is placed in parallel with the resistor R1K. The reduced equivalent resistance speeds up the slew-rate of the amplifier 97. This allows the results of manipulating the reference voltage by adjusting the resistor R4K to be had with less delay, because of faster loop response.
The switches S1K and S2K together have the functionality of both electronic and manual switches which combined have the function designated S1E in Fig. 6. Furthermore, the circuitry comprising the error amplifier 97, the capacitor C1K, the switch S1E (from the switches S1K and S2K), and the resistors R1K and R2K combine to give the functionality of a dual-speed error amplifier 98 of Fig. 6.
A plurality of comparators 114 and 115, along with a pair offsets 240 and 242 comprise a window comparator, with inputs comprising the voltage at the node E5K and with the voltage of the node E1K plus and minus the offsets 240 and 242. During normal operation, with the switch S2K open, when the voltage at the node E5K lies within the window created by the voltage at the node E1K plus and minus the offsets of 240 and 242, the outputs of both comparators are in a first state. When the voltage at the node E5K is outside the voltage window, one of the comparators will change state. Since the outputs of the comparators 114 and 115 are wired together with an OR logic, when either activates, a one-shot 122 is triggered via an "OR" gate 140. The one-shot 122 is so designed as to change state during the entire time that the trigger is active, plus a pre-determined "one-shot" time after the trigger terminates. When triggered by the "OR" gate 140, it closes the electronic switch S1K. This places the resistor R2K in parallel with the resistor R1K. The parallel combination has a reduced equivalent resistance which speeds up the slew-rate of the dual-speed error amplifier 98 and speeds up loop response. By this means, large sudden changes in the value of the voltage at the node E5K or the node E1K trigger a faster response than would otherwise occur. Further, loop saturation will be detected as a constant difference between the voltage at the nodes E5K and E1K. Thus, a return from loop saturation is hastened by the faster loop speed afforded by the detection of the window comparator. A current buffer 120 drives the LED D2K, to indicate the state of the control loop speed.
The unidirectional signal device 130 contains circuitry that detects whether a load exists at the output, E4K. If no load exists at the output, the sensing circuitry issues a trigger, E6K to the one-shot 122 through the "OR" gate 140. Thus, loop speed is increased and remains increased while no load exists at the output E4K, and for the "one-shot" time period after a load is placed at the output E4K. This allows faster return to regulation when the load, which is typically the control signal to the variable light source, is applied.
Fig. 6 is a schematic of a dynamic illumination stabilizer with dual loop speeds, including automatic calibration. A plurality of switches S2R and S1R are shown here in the mode for normal operation. A photodiode D1R, in conjunction with the current-to-voltage converter 106 and a feedback element, a non-volatile digital potentiometer 222, form the light transducer 96, in this case a light-to-voltage converter. It should be noted that other light-transducing devices may be employed, which may include, for example, photoresistors. The essence of the teachings of the present invention require only a device which outputs a signal analogous to the light impinging on the transducer, and that the ratio of the output signal to a given light stimulus be variable under the control of the calibration means. The converter output voltage at a node E7R relates to the current generated by the light level impinging on the photodiode D1R, and the value of the feedback resistor in digital potentiometer 222. The error amplifier 97 amplifies the difference between the voltage at the node E7R seen through a resistor R1R, and the voltage at a node E1R, the reference voltage 236 seen through a resistor R3R at the non-inverting input. The resistor R1R and a capacitor C1R set the slew-rate and response speed of the amplifier 97. The output of the amplifier 97 is current buffered by the unidirectional signal device 130 and provided to an output node E4R. Closed loop operation is achieved by connecting the output node E4R to the control input of an external variable light source with an intensity responsive to the output node E4R, and with this light source at least partially illuminating the photodiode D1R. The circuit is so arranged that an increase in light impinging on the photodiode D1R reduces the light output from the external, variable light source such that equilibrium is maintained.
The resistor R3R provides an impedance between the reference voltage 236 and the non-inverting input of the amplifier 97, such that if an external voltage is connected to an external control voltage input E6R, it can override the voltage seen through the resistor R3R if the voltage into the control voltage input E6R has a lower source impedance and the switch S1R is in the normal operating position.
Together, the error amplifier 97, the capacitor C1R, the electronic switch S2R and the resistors R1R and R2R function as a dual-speed error amplifier 96.
A summary flow-chart of the steps involved in calibration is shown in Fig. 26A not representing the invention, which may be understood in conjunction with Fig, 8 also not representing the invention and shown as an illustrative example. Calibration of the circuit is initiated by activating a calibrate-mode signal input E5R. With signal input E5R activated, the switch S1R changes state. removing external control voltage input E6R from the non-inverting input of the amplifier 97. and placing the non-inverting input of the amplifier 97 at the reference voltage 236 potential, plus an offset 234. Further, signal input E5R closes the switch S2R through the "OR" gate 140 and the one-shut 122. A resistor R2R, now in parallel with the resistor R1R, increases the slew-rate of the amplifier 97. This, in conjunction with the large voltage impressed at the node E1R, drives the output of the amplifier 97 to the maximum control limit, ensuring a maximum output signal at the output E4R, and therefore, maximum illumination from the variable light source under control. The signal input E5R also activates the enable line into the non-volatile digital potentiometer 222. The reference voltage 236 is related to the maximum functional control voltage and therefore the maximum light level available from the variable light source under control. With no other light source present, and with the external, variable light source at maximum, the light impinging on the photodiode D1R is at the maximum. The voltage at the node E7R is then adjusted to a voltage equal to the maximum control voltage. By this means, calibration is achieved because voltage at the node E7R will then be at the maximum control voltage value when the external, variable light source is also at the maximum.
In closer detail, and with reference to Fig. 8, calibration is achieved by a window comparator examining the voltage of the node E7R in relation to the reference voltage 236. The offset 232 creates a voltage window for the comparators 114 and 115. If the voltage at tho node E7R falls outside the window, one of the comparators will trip. The circuit is so arranged that the comparator which trips, adjusts the non-volatile potentiometer 222 in a direction which returns the voltage at the node E7R towards the center of the voltage window. This automatically adjusts the gain of the amplifier 106. The feedback resistance value, thus automatically chosen, is stored in the non-volatile digital potentiometer 222 when the calibration mode signal input E5R deactivates. Because the potentiometer 222 is non-volatile, the resistor value remains stored when the system is de-energized, and is present when power returns, avoiding the necessity to recalibrate each time power returns. As the signal input E5R de-activates, the one-shot 122 remains active for a preset time. This keeps the amplifier 98 in the faster-response speed mode and therefore decreases the time required for the circuit to return to regulation. The unidirectional signal device 130 contains circuitry that detects whether a load exists at the output node E4R. If no load exists at the output node E4R, the sensing circuitry issues a trigger to one-shot 122 through the "OR" gate 140. Thus, loop speed is increased and remains increased while no load exists at the output node E4R, and for the one-shot 122 time period after a load is placed at the output node E4R. This allows faster return to regulation when the load, which is typically the control signal to the variable light source, is applied.
Fig. 9A is a schematic of an dynamic illumination stabilizer with dual loop speeds and automatic calibration, not representing the invention and carried out by means different from that of Fig. 8. In normal operation, the electronic switches S2R and a switch S3R are open, while an electronic switch S4R connecting the node E1R and the external control voltage input E6R is closed. The photodiode D1R in conjunction with the current-to-voltage converter 106 and a feedback element, the non-volatile digital potentiometer 223, form a light transducer 96, in this case a light-to-voltage converter with an output voltage at node E7R relating to the light level impinging on the photodiode D1R, and the value of the feedback resistor in the potentiometer 223. The error amplifier 97 amplifies the difference between the voltage at node E7R seen through the resistor R1R, and the voltage at the node E1R at the non-inverting input of the error amplifier 97. The reference voltage at the input E1R is a replication of the reference voltage 228 which has been processed by a micro-controller 238, by first analog-to-digital (A/D) from node E7R and then by digital-to-analog (D/A) processing internally, which is then passed to a node E8R, where it is seen by the non-inverting input of the amplifier 97 through a resistor R5R. The resistor R5R provides an impedance between the reference voltage E8R and the voltage E1R at the non-inverting terminal of the error amplifier 97, such that a signal received by the external control voltage input E6R will override the voltage at the node E8R if the impedance of that signal is lower than the impedance offered by the resistor R5R. The microcontroller 238 continuously monitors the external control voltage input E6R on the line 136 using the A/D converter on the microcontroller 238. The circuit operates so as to maintain a constant total illumination (natural plus artificial illumination) detected by light transducer 96.
The resistor R1R and the capacitor C1R set the slew-rate and response speed of the amplifier 97. The output of the amplifier 97 is current buffered by the unidirectional signal device 130 and sent to the output node E4R. The unidirectional signal device 130 contains circuitry that detects whether a load exists at the output node E4R. If no load exists at the output, the sensing circuitry issues a trigger to the one-shot 122 through the "OR" gate 140. Thus, loop speed is increased and remains increased while no load exists at the output node E4R, and for the one-shot time period after a load is placed at the output node E4R. This allows faster return to regulation when the load, which is typically the control signal to the variable light source is applied. Alternatively, if the microcontroller 238 senses that a change the external control voltage input E6R at the input 136 greater than a predetermined threshold, the microcontroller 138 issues a command to the input of the OR gate 140 through the output line 274, triggering the one-shot 122, which increases the slew rate of the error amplifier 97 for a predetermined time. The result of this is that when the system attempts to change the regulated illumination either automatically or manually, rapid regulation restabilization occurs.
Closed loop operation is achieved by connecting the output E4R to the control input of an external variable light source with an intensity responsive to the voltage at the output node E4R, and with the illumination source at least partially illuminating the photodiode D1R. The circuit is so arranged that an increase in total light impinging on the photodiode D1R reduces the light output from the external, variable light source such that equilibrium is maintained. The resistor R5R provides an impedance between the replicated reference voltage at the node E8R and the non-inverting input of the amplifier 97, such that if an external voltage is connected to external control voltage input E6R, it can override the voltage seen through the resistor R5R if the voltage into the external control voltage input E6R has a lower source impedance and the switch S1R is in the closed position.
Calibration is carried out with or without the presence of a source of non-variable ambient light. It is the purpose of calibration to adjust the gain of the current-to-voltage converter 106 so that in the absence of light other than the external, variable light source, and with said external variable source at maximum brightness, the voltage at the converter 106 output node E7R is equal to the maximum value of control voltage E4R corresponding to the maximum brightness of the external variable light source. To do this, the calibration procedure effectively computes and subtracts the effect of ambient light that is present, so that accurate calibration may be attained.
A summary flow-chart of the steps involved in calibration is shown in Fig. 26B, which may be understood in conjunction with Fig. 9A. Calibration of the circuit is initiated by activating the calibrate mode signal input E5R. With the signal input E5R activated, the switch S4R is opened, disconnecting the external control voltage input E6R from the node E1R. Furthermore, the signal input E5R closes the switch S3R, converting the amplifier 97 to a voltage follower for the voltage seen at the non-inverting input E1R. The signal input E5R, through the "OR" gate 140, triggers the one-shot 122, closing the switch S2R, placing the resistor R2R in parallel with the resistor R1R. The signal input E5R further enables the calibration sequence resident in the micro-controller 238.
In response to the enable signal E5R, the micro-controller 238 initiates an internal program which performs an automatic calibration process. Voltage at the node E8R is set to a voltage equivalent to the maximum control voltage that the external variable light source is known to respond to when yielding the brightest light. For the purposes of illustration, let us assume that the maximum control voltage is equal to 10 volts. Since the resistance of the non-volatile digital potentiometer 223 and hence the gain of the amplifier 106 is proportional to the binary number of a data bus 244 when digital potentiometer 223 is enabled, the micro-controller 238 sets the data output bus 244 to the maximum binary number available, so that the digital potentiometer resistance provides maximum gain for the amplifier 106, and enables digital potentiometer 223 by activating an enable line E9R.
The voltage at the node E8R is held fixed at 10 volts. This voltage is passed by the amplifier 98, acting as a voltage follower, to the unidirectional signal device 130 to the output node E4R and on to the external variable light source, which then responds by providing maximum light. Because the amplifier 98 has been converted to a voltage follower for the non-inverting input, it ignores the voltage at the node E7R. Then, the micro-controller 238 sequentially decrements the data bus 244, incrementally reducing the gain of the amplifier 106 via the digital potentiometer 223, until the micro-controller 238 computes the voltage at the node E7R to be 6 volts. The binary number present at the data bus 244 at this point is sent to memory within the micro--controller 238. The controller 238 then reduces the voltage at the node E8R 5 volts. This voltage is again passed to the output E4R where the external, variable light source brightness is reduced to 50% of full. The new lower voltage at the node E7R is then digitized and stored. For purposes of explanation, we will posit the value of 4 volts at the node E7R. This value is subtracted from the digitized equivalent of the previous reading of 6 volts. The result (6 V - 4V = 2 V1) is divided by the original 5 volt difference in control signal, revealing an amplifier gain of 2/5. If the gain of the amplifier were calibrated, the gain term should have been 1/1. That is, a 5 volt change in the amplifier is desired for a 5 volt change in the control signal. The reciprocal of the gain term of 2/5, or 5/2, is calculated because it is the multiplier required for the amplifier to achieve a gain of 1. The binary number stored when the amplifier first reached 6 volts is multiplied by this reciprocal, and the new binary number is set at the data bus 244. The enable line E9R is released, storing the non-volatile gain information in the digital potentiometer 223. The system is now calibrated.
When the calibration mode input E5R is released, the amplifier 98 returns to error amplifier mode and normal operation resumes with the exception that the one-shot 122 is still active, keeping the switch S2R closed. This keeps the response speed of the amplifier 98 fast, and returns the loop to regulation quickly. The one-shot 122 then times out and the loop returns to slow speed operation.
Fig. 9B is a schematic of a dynamic illumination stabilizer with dual loop speeds, automatic calibration, and dynamic Total Effective Illumination stabilization all not representing the invention. The purpose of this circuit is to maintain a constant Total Effective Illumination (TEI), which is the sum of artificial illumination and a factor times the external light. This is distinguished from the dynamic illumination stabilizer depicted in Fig. 6, Fig. 8, and Fig. 9A, which stabilizes the sum of artificial and external light, without a factor weighting the external light. The dynamic illumination stabilization can be considered the special case where the factor used to weight external illumination in the computation of TEI is always equal to one. The usefulness of TEI derives from the different spectral and Lambertian characteristics of artificial and external light, which for the purposes of this discussion we will consider to be natural light (although other sources of light, such as high intensity discharge and area lighting may apply). Therefore, compensating for variations in natural light by an equal amount of artificial light may provide effective changes in perceived illumination by the user. Thus, it may be preferable to compensate for variations an natural light with larger or smaller amounts of artificial light, depending on the source and location of the natural and artificial illumination, as well as the illuminated environment.
In order to regulate the TEI output of the dynamic illumination stabilizer, the amounts of artificial and natural illumination components must be known at all times. The following discussion assumes proper calibration of the device, by which the value at the node E7R is caused to be equal to the maximum output value to which the external light modulator will respond under control output E4R. This calibration procedure will be discussed below. The artificial light is computed as a factor of the output control voltage E4R, which is derived from a D/A output 270 of a microcontroller 239. Thus, this information is directly accessible to the microcontroller 239. The total amount of illumination is available from the output of the light transducer 96, which is directly sampling the area under control. This information is input to the microcontroller 239 the node E7R. From these two values, the amount of natural light can be computed by subtraction. Thus, if T is the total illumination, N is the natural illumination, and A is the artificial illumination, by definition, T = A + N, and thus, N = T - A.
The Total Effective Illumination (TEI) is then computed by the microcontroller 239 by the formula TEI = A + XN, where X is a weighting factor which may be less than or greater than 1, though in general it will be greater than 1. In the following discussion, X is considered a constant, although it is within the teachings of the present invention that X may be a variable function of external control voltage input E6R, time of day, or total illumination T. The microcontroller participates in a closed-loop feedback system wherein the output control voltage E4R is varied so that the difference between the external control voltage input E6R and the computed TEI is minimized.
The rate of change of the output 270 from the microcontroller 239 through the unidirectional signal device 130, and thence to the light modulators connected to the output E4R, is regulated by an internal time constant which can be changed on the basis of the value of an input 272. In a first state, this time constant provides for smooth and gradual artificial light change in normal operation. In a second state, this time constant provides for rapid restabilization of lighting so that the TEI returns after disruption to the desired light registered at the external control voltage input E6R. Such instances of disruption are input to the microcontroller 239 through the OR gate 140 and the line 272, which occurs at least under the three following circumstances. In the first case, as calibration ends and for a predetermined period stored within the microcontroller 239, it is beneficial for the system to regain regulation as quickly as possible. In a second case, the unidirectional signal device 130 contains circuitry that detects whether a load exists at the output node E4R. If no load exists at the output, the sensing circuitry issues a signal through the "OR" gate 140 to the microcontroller 239 via the line 272. The loop speed is increased and remains increased while no load exists at the output node E4R, and for a predetermined period stored within the microcontroller 239 after a load is placed at the output node E4R. In the third case, the microcontroller 239 constantly monitors the change in external control voltage input E6R. When this input change exceeds a predetermined threshold, indicating that the lighting network, by either automatic or manual means, is attempting to change the desired light input, rapid restabilization of the system is performed.
Calibration is carried out with or without the presence of a source of non-variable ambient light, in a manner similar to the dynamic illumination stabilizer of Fig. 9A not representing the invention. The following discussion may be read with reference both to Fig. 9B and Fig. 26B, both not representing the invention.
Calibration of the circuit is achieved by activating the calibrate mode signal input E5R. The signal input E5R, through the "OR" gate 140 via line 272, changes the internal time constant within the microcontroller 239 to allow for rapid restabilization of lighting, as described above. The signal input E5R further enables the calibration sequence resident in the micro-controller 239.
In response to the enable signal E5R, the micro-controller 239 initiates an internal program which performs an automatic calibration process. Output voltage on line 270 is set to a voltage equivalent to the maximum control voltage that the external variable light source is known to respond to when yielding the brightest light. For the purposes of illustration, let us assume that the maximum control voltage is equal to 10 volts. Since the resistance of the non-volatile digital potentiometer 223 and hence the gain of the amplifier 106 is proportional to the binary number of the data bus 244 when digital potentiometer 223 is enabled, the micro-controller 239 sets the data output bus 244 to the maximum binary number available, so that the digital potentiometer resistance provides maximum gain for the amplifier 106, and enables digital potentiometer 223 by activating an enable line E9R.
The voltage on the control output line 270 is held fixed at 10 volts, and hence the output node E4R responds by providing maximum light. Then, the micro-controller 238 sequentially decrements the data bus 244. incrementally reducing the gain of the amplifier 106 via the digital potentiometer 223, until the micro-controller 239 computes the voltage at the node E7R to be 6 volts. The binary number present at the data bus 244 at this point is sent to memory within the micro--controller 239. The controller 239 then reduces the voltage at the output line 270 to 5 volts. This voltage is again passed to the output E4R where the external, variable light source brightness is reduced to 50% of full. The new lower voltage at the node E7R is then digitized. For purposes of explanation, we will posit the value of 4 volts at the node E7R. This value is subtracted from the digitized equivalent of the previous reading of 6 volts. The result (6 V - 4 V = 2 V) is divided by the original 5 volt difference in control signal, revealing an amplifier gain of 2/5. If the gain of the amplifier were calibrated, the gain term should have been 1/1. That is, a 5 volt change in the amplifier is desired for a 5 volt change in the control signal. The reciprocal of the gain term of 2/5, or 5/2, is calculated because it is the multiplier required for the amplifier to achieve a gain of 1. The binary number stored when the amplifier first reached 6 volts is multiplied by this reciprocal, and the new binary number is set at the data bus 244. The enable line E9R is released, storing the non-volatile gain information in the digital potentiometer 223. The system is now calibrated, and as a result of calibration, in the absence of natural or uncontrolled light, the values at the input E6R and E7R will be nearly equal. When the calibration mode input E5R is released, the internal time constant of the microcontroller is restored after a predetermined time stored within the microcontroller 239, and the feedback loop returns to slow speed operation.
In the absence of external control voltage input E6R, the resistor R3R holds the microcontroller 239 input line 136 at the reference voltage 228. The reference voltage 228 is made equal to the maximum control voltage to which the external light modulator is known to respond.
The previous discussion has described the cooperation of an infrared remote control with a photosensor to provide a novel mode of interaction called illumination stabilization. Further novel interactions involve the use of external signals, most conveniently from the infrared remote, to initiate calibration of the photosensor. The following discussion extends the modes of sensor cooperation to include that between an occupancy sensor and an infrared remote.
As mentioned before, calibration of the occupancy sensor is made difficult because the person performing the calibration cannot be located physically proximate to the occupancy sensor, since their presence would affect the calibration. The infrared remote control provides a means of communicating with the occupancy sensor to effect calibration. Fig. 10A is a schematic of a motion sensor allowing remotely actuated calibration, including both manual and self-calibrating modes. The signal from a motion sensing element 210 is amplified by a first amplifier 212. The gain of the first amplifier 212 is governed by a non-volatile, digital potentiometer 225, arranged as a feedback element for the first amplifier 212. The output of the first amplifier 212 may optionally be further amplified by a second amplifier 214 and applied to one input of a comparator 216. A reference voltage 228 is applied to the other input of the comparator 216. In the normal, detecting mode, the output of the comparator 216 is passed to the trigger input of a timmg circuit 218. The timing circuit 218 provides an output for a predetermined length of time after being triggered by the output of the comparator 216. The output of the timer 218 is passed to the input of an output stage 220. and passed to an output node E8C As described above, the output node E8C is further under the control of the polling signal E11C, which inhibits the output stage 220 from pulling down on the output node E8C unless agreement is found among all occupancy sensors connected to the polling signal input E11C. The output stage 220 performs a number of different functions, including current buffering, setting of the dim level that is produced when the occupancy sensor determines that the room is unoccupied, as well as the timing for raising and lowering illumination. The internal components comprising the output stage are discussed in more detail below.
Upon receiving a calibration command from any of a plurality of calibration command inputs E10C, E12C, E13C, and E16C, the calibration cycle is initiated. During this time an "OR" gate 230 output at a node E15C goes active. This enables adjustment of a non-volatile digital potentiometer 225, and inhibits the timer 218 from operating and issuing an output. With an electronic switch S1C in the de-energized position as shown, the output of the comparator 216 is tied to the up/down input of the digital potentiometer 225. Potentiometer 225 is thus enabled and the circuit is so arranged that if the comparator 216 has not tripped, the output E14C incrementally adjusts digital potentiometer 222 in the direction required to raise the gain of the amplifier 212. Conversely, if comparator 216 has tripped, the circuit is arranged to lower the gain of amplifier 212.
If the calibration command input E10C (auto-calibrate until release) is active, and a given stimulus is applied to the motion sensor 210, the circuit will continue to adjust the gain of the amplifier 212 in the correct direction until the output of the comparator changes state. The change of state of comparator 216 indicates that calibration has been achieved. Then, the circuit will reverse the direction of gain adjustment which will take the comparator 216 back across the threshold. Thus the circuit will continue to dither about the calibration point until the calibration command input E10C is released. Because each increment and decrement of the digital potentiometer 222 is small, and hence the gain change of the amplifier 212 is also small, the system can be considered calibrated whatever the state of the comparator 216 at the moment of calibration command release.
If the calibration command input E12C (timed calibrate) is activated, a timer 224 is triggered. The output of the timer 224 enables the digital potentiometer 225 through the "OR" gate 230. At this point, the calibration command input E12C can be released and calibration proceeds via the timer 224 output, with the comparator 216 eventually dithering about the calibration point as described previously, until the timer 224 times out, terminating the calibration cycle.
If the calibration command input E13C (auto-calibrate until threshold is reached) as activated, a flip-flop 226 is "set". The flip-flop 226 output enables the digital potentiometer 225 through the "OR" gate 230. Once this occurs, the command line input E13C may be released, and the calibration mode continues via the flip-flop 226 output until the comparator 216 output changes state. The circuit is so arranged that a change in state of the comparator output E14C in either direction "resets" the flip-flop 226, terminating the calibration cycle.
Manual adjustment is achieved by activating the manual calibration command input E16C. Activation of input E16C enables digital potentiometer 225 and disables the timer 218 through the "OR" gate 230. Simultaneously, E16C switches the switch S1C, this disconnects the comparator output E14C from the "up/down" input of potentiometer 225, and connects the "up/down" input of potentiometer to the manual adjustment input E9C. While E16C is active, potentiometer is adjusted in the direction commanded by the state of the manual input, E9C.
The various calibration commands described above may be permitted within the same embodiment, us indicated in Fig. 10A not representing the invention, or only some of the commands may be available in any one embodiment. Setting of the commands through inputs E9C, E16C, E10C, E13C, and E12C will in general be mediated through use of an input device separate from that of the occupancy sensor. Since the device is then physically distant from the occupancy sensor, the user may initiate the calibration from a distance, participating in the calibration (e.g. waving hands) or moving outside of a doorway, without having to come close to the occupancy sensor.

Occupancy Sensor Dual-Timing Circuitry

As mentioned before, prior art occupancy sensors respond by turning lights on and off, rendering them largely unsuitable for use in networks in which the primary mode of lighting control is modulation of the brightness of individual lamps, rather than turning lamps on and off. The method of the current invention can use the occupancy sensor to modulate illumination levels from individual lamps. This provides two primary benefits. Firstly, because fluorescent lamp life is greatly reduced by frequent on-off cycling, lamp life is greatly extended by using lamp brightness modulation to provide the energy-benefits of occupancy sensing. This provides significant financial benefits through elimination of the cost of frequent lamp replacement and attendant labor costs. Secondly, people do not then need to enter and navigate dark rooms, such as bathrooms, prior to triggering the occupancy sensor. In such a case, the occupant may feel discomfort and fear, or injure himself. This problem is especially exacerbated by the use of on-off cycling, where lamp ignition takes an appreciable period of time subsequent to triggering the occupancy sensor. Rather, in the method of the present invention, the rooms may be lit to a predetermined low illumination level when the occupancy sensor determines that the room is unoccupied, so that a person entering the room has sufficient illumination to view their surroundings before the occupancy sensor triggers, yet the energy-savings benefits are largely preserved.
The method of the present invention provides for differing rates of illumination change when the illumination is being raised or lowered by the occupancy sensor. The purpose for this difference in timing is to provide ergonomic and safety benefits to occupants if the occupancy sensor incorrectly determines that a room is unoccupied, and begins to reduce the illumination levels. Reducing illumination slowly allows an occupant to make their presence known to the occupancy sensor by deliberate motion before the lighting levels are significantly reduced. On the other hand, when entering a room whose lighting has already been reduced by the occupancy sensor, rapid restoration of higher lighting levels is essential for occupant safety and comfort.
Fig. 10B is a schematic depicting the output stage 220 not representing the invention of the occupancy sensor of Fig. 10A not representing the invention. Three separate functions are shown, including the ability to set the dim level, to exercise polling between occupancy sensors in the manner described above, and to provide distinct speeds for lowering and raising the output signal. The circuitry depicted in this figure bears some similarity to that shown in Fig. 4 according to the invention. Because of the additional functionality described above, however, differences between the different embodiments will become apparent.
In the absence of an active polling signal at the polling signal input E11C, and with the input E17C from the detection circuitry to the output stage 220 sufficiently low in voltage (deactivated) to back-bias a diode D11C, a capacitor C11C charges to the potential set by a potentiometer R13C through a resistor R12C. The resultant potential stored on the capacitor C11C is buffered by the unidirectional signal device 130, and passed to the output node E8C. Therefore, by adjusting the value of the potentiometer R13C, the output node E8C can be set to the desired voltage level. By this means, the dim level directed by the occupancy sensor when the sensor determines that the area is unoccupied, may be set. It is within the teaching of the method of the present invention that the potentiometer R13C may be replaced by a fixed voltage divider, or by any electronic means for providing a desired fixed or variable voltage.
When the occupancy sensor determines that the area examined is occupied, the input E17C is activated and rises to a potential at least equal to the highest control voltage to which the lighting system responds. The diode D11C is forward-biased, presenting nearly this potential at the top of the resistor R11C through a forward-biased diode D12C, and to the polling signal at the polling signal input E11C. The impedance of the resistor R11C is made substantially lower than that of the resistor R12C, such that the capacitor C11C charges to a potential nearly equal to that of the cathode of the diode D12C, regardless of the setting of the resistor R13C. The resultant potential on the capacitor C11C is buffered by unidirectional signal device 130 and passed to the output E8C. The value of the resistor R11C may lie between 0 ohms and any value which in conjunction with the capacitor C11C provides a time constant which is rapid (generally on the order of less than a second though other values may suffice), such that there is little delay between activation of the input E17C and resultant voltage change on the output E8C. When the input E17C deactivates, reverse-biasing the diodes D11C and D12C, the capacitor C11C discharges to the voltage value set by the resistor R13C, through the resistor R12C, which is, as mentioned previously, a much higher impedance than the resistor R11C. Therefore, the rate of discharge of the capacitor C11C, and the resultant rate of change at the output node E8C is much longer than the rate of charge provided when the input E17C is active. The rate of charge of the capacitor C11C is governed by the value of the resistor R11C, and the rate of discharge of the capacitor C11C is governed by the value of the resistor R12C. The values of the resistors R11C and R12C may be predetermined and fixed, or may be adjusted manually or electronically during adjustment of the output stage 220 of the occupancy sensor.
When a plurality of the output stages 220 are linked together at the polling signal input E11C, the activation of any input E17C from any of the plurality of output stages 220, will raise the voltage at the polling signal input E11C in all output stages. This forces the capacitor C11C of all output stages 220 to charge nearly to the value of the node E11C, because of the aforementioned ratio of impedance of the resistors R11C and R12C in all output stages 220. Therefore, the output E8C of all output stages 220 will be at the voltage on the capacitor C11C even if the inputs E17C from one or more of a plurality of the rest of occupancy sensors 62 is deactivated if the sensors determine that the area under surveillance is unoccupied. Therefore, the inputs E17C from all occupancy sensors with the polling signal input E11C in common must all be deactivated before the capacitor C11C may be discharged to the voltage set by the resistor R13C.
The combination of different types of control devices (e.g. infrared remote and photosensor) or the interaction of similar types of control devices (e.g. polling among occupancy sensors) represent powerful methods of controlling local networks. Next, we will show how these methods may be extended to include control interactions in large, hierarchical networks.

Network Coordination And Control Hierarchy

Network Logic and Schematic Description

The shared proportional-response control system described above works to coordinate controls within a common environment. In the example shown in Fig. 1, the manual dimmer 60, the light sensor 64, the occupancy sensor 62, the infrared remote 61, and the local computer control 63 are all operating within a single local lighting network. Consider, however, a work room for two people that had sets of light fixtures near to windows offering external illumination, as well as fixtures not located near external illumination. Those fixtures near to windows would be well served with photosensor control, while other parts or the room away from the windows and external illumination would not benefit. Furthermore, consider that the entire room responded to a centrally-placed occupancy sensor. Finally, consider that the entire room was further controlled by a building energy management system (EMS) which controlled the lighting for an entire floor at once. This scenario, which would not be atypical of a desired lighting system, encompasses three levels of control. At the highest level is the building EMS system which controls the entire floor, at the next level, a room-wide control using an occupancy sensor, and lastly, part-of-a-room control, in which only some of the fixtures respond to a photosensor. It would be beneficial if the fixtures near the window respond to all three levels of control, while those away from the window need respond to only the top two levels of control. Fixtures in other rooms on the same floor would respond only to the top level control, as well as local controls that might be resident in the other rooms.
Conventional lighting systems, in general, do not manage multiple controls within a single level. While multiple levels are not unknown, such levels require special circuitry and connections. On the other hand, the shared proportional-response control system of the current invention, as will be described below, handles multiple levels of control transparently. Each control is connected to other controls by one of two points - one point connects to controls on the same or a higher level, while the other point connects to controls at a lower level connection. The special "lower level" connector blocks the control signals from below, preventing them from affecting the control signals in a higher level. Connections to the other point allow for joint control of fixtures. If the other connected controls are at the same level, coordinated control is available as described earlier. If another connected control is from a higher level, this represents supervisory control from a higher level to the current level. This allows hierarchical control to be established simply from the network connecttvity, rather than from special hardware.
In Fig. 5 not representing the invention controls at the same or higher control level are connected to the uni-controller 138 not according to the invention through a multiplicity of input/pass-through connectors J1, J2 and J3, arranged in parallel along an input control bus 144. The control signal on the control bus 144 is transmitted to the internal control bus 142 through a uni-controller unidirectional signal device 130. which prevents the voltages in the internal control bus 142 from affecting the input control bus 144. The voltage on internal control bus 142 is potentially affected by four agents - the pass-through input control voltages connected through the unidirectional signal device 130, the motion sensor 62, the photosensor 64, and the infrared control 61. The internal control bus functions as a shared, proportional-response control line. Thus, the control requesting the lowest control signal on the internal control bus 142 prevails through virtue of the pull-down diodes 86 and the unidirectional signal device 130. This control signal voltage is transmitted to lower level controls and light modulators through the connector J7.
Because of the unidirectional signal device 130, control at the same and higher level than the given uni-controller exert control over the given control and controls on levels below connected at the connector J7. At the some time, controls within the given uni-controller have no affect on controls connected at the connectors J1, J2 and J3. This satisfies the requirements of hierarchical control, and as we will describe, complicated hierarchical control in a network can be organized using the same controls simply be establishing the connectivity of the controls.
Fig. 11 not representing the invention is a block diagram depicting a lighting network with a single level of control. The uni-controller 138 contains the photosensor 64, the infrared control 61, the occupancy sensor 62, the infrared remote receiver 59 and the unidirectional signal device 130, as represented by rounded rectangles enclosing a "P" "I", "O", "R" and a downwards pointing triangle, respectively. Because there is only a single level of control, there are no supervisory or coordinated controls connected to the input/pass-through connectors J1, J2 and J3, represented by the squares in the upper quadrant of the uni-controller 138. Four power-generating dimmable ballasts 146 are connected to the output jack J7.
Because there is only a single level of control, no controls under the supervision of the single control are connected to output connector J7 of uni-controller 138. The ballasts 146 respond to any of the three controls in the uni-controller 138. Because there are no supervisory or coordinated controls operating at or above the level of uni-controller 138, the unidirectional signal device 130 has no effect in this control configuration.
A network cable 166 connects the ballasts 146 to the uni-controller 138. This cable contains a number of wires, which include at least a control signal wire and a return. In addition, the cable 166 may contain a wire carrying DC power to power the ballasts, and a polling wire for poll-responsive occupancy sensors. In this diagram, the uni-controller 138 is shown with only those internal connections related in control, while power, return and polling connections are not shown. The power supplied by the ballasts 146 is transmitted to the uni-controller 138 through the network cable 166.
Fig: 12 not representing the invention is a block diagram depicting a lighting network with three levels of control. In this figure, a number of uni-controllers are depicted, each showing 3 input connectors and one output connector. The designations for these connectors (J1, J2, J3 and J7) are topologically consistent throughout the diagram and follow those depicted in Fig. 5 and Fig. 11. The internal controls are designated with letters within rounded rectangles, with the letter "R" referring to the infrared remote receiver 59, the letter "O" referring to the occupancy sensor 62, the letter "I" referring to the infrared control 61, and the letter "P" referring to the photosensor 64. An inverted triangle designates the unidirectional signal device 130. As in Fig. 11, the uni-controllers are depicted only with those internal connections related to control, while power, return and polling connections are not shown.
A network computer control 148 is at the highest level of control. It does not have network inputs, but only a single output connector 164. The lighting network is divided into two high level zones, labeled Zone 1 and Zone 2, each with ballasts that are governed by single uni-controllers, which are a Zone 1 uni-controller 150 and a Zone 2 uni-controller 152, respectively. Output from the computer controller 148 is transmitted through the network cable 166 into the input/pass-through connector J1 of the Zone 1 uni-controller 150. This uni-controller 150 is in turn connected via the input/pass-through connector J3 to the Zone 2 uni-controller 152 input connector J 1 through a continuation of the network cable 166. The voltage set by the computer controller 148 is transmitted through the unidirectional signal devices of uni-controllers 150 and 152 to their respective control bus. This higher-level supervisory control voltage establishes a maximum control voltage on the control buses of both the Zone 1 uni-controller 150 and the Zone 2 uni-controller 152, and through these uni-controllers, for the entire network.
It should be noted that the choice of the connectors J1, J2, and J3 for network connectivity is arbitrary, since all three jacks are in parallel configuration in the uni-controllers. It should also be noted that the control voltage at the input connectors of the uni-controllers 150 and 152 is solely under the control of the computer controller 148. The uni-controllers 150 and 152 themselves can have no influence over the voltage in their respective input control buses.
In this depiction, the Zone 1 uni-controller 150 contains only an infrared control along with its supporting infrared remote receiver. This might represent, for instance, zonal lighting control over multiple areas of a room or lighting control over a multiplicity of rooms. The network cable 166 emerges from the uni-controller 150 and connects both to a pair of ballasts 146, as well as to a Zone 1. Area 2 uni-controller 156. Thus, it can be seen that, the control signal in the network cable 166 can both transmit lighting control signals to other controllers, but the same signal exerts control over ballast output directly, as well. The Zone 1, Area 2 uni-controller 156 is connected through input connectors to a Zone 1, Area 1 uni-controller 154. Because all input connectors in the uni-controllers 154 and 156 at this third level of control are in parallel, the input control signals to the uni-controllers 154 and 156 are equal. It should be noted that the signals that these uni-controllers receive is the control voltage of the output control bus of the uni-controller 150, which is set by the lower of the control voltages emanating from the infrared control of the uni-controller 150 and the computer control 148. Thus, the uni-controllers 154 and 156 at this third, low level are responsive to both higher level controllers 150 and 148.
It is noteworthy that many of the uni-controllers 150, 152, 154, 156, 158, 160, and 162 have different complements of internal controls. For example, the uni-controller 154 has three controls while the uni-controller 156 has only one control. The uni-controllers need not have the same controls, but each uni-controller nonetheless maintains all the remaining characteristics, including the input communications via infrared manual remote input, the internal control bus, the unidirectional signal buffer, the control power network, and the control signal network. This commonality in interface supporting lighting networks, makes for simple design, installation, maintenance, and operation.
Also, the order of connections between the output connector J7 of the supervisory uni-controller 150 and the input/pass-through connectors J1. J2 and J3 of the uni-controllers 154 and 156 is arbitrary. The system would operate identically if the uni-controller 150 output connector J7 were connected to the input/pass-through connector J2 of the uni-controller 154 and the input/pass-through connector J1 of this uni-controller 154 were connected to the input/pass-through connector J3 of the uni-controller 156.
Each lowest tier uni-controller can be considered to influence the artificial light output within an area illuminated by the ballasts 146 under its control. For this description, we will refer to these with area number designations for each uni-controller. The Zone 2 uni-controller 152 supervises three uni-controllers -- a Zone 2, Area 1 uni-controller 158, a Zone 2, Area 2 uni-controller 160, and a Zone 2. Area 3 uni-controller 162. In this example, the uni-controllers are not connected as in a "daisy chain", as are the Zone 1 Area 1 and Area 2 uni-controllers 154 and 156. Rather, the uni-controllers are connected in a "star" fashion, with multiple connections established with the uni-controller 160. The ability to use both daisy chain and star configurations allows a wide range of connectivity between uni-controllers, conserving the amount of cabling that is necessary during installation. It is possible that uni-controllers may be configured with larger numbers of input connectors than the three described here, and a larger number of connectors would allow for wide flexibility in installations. In addition, special "star connectors" may be provided, which allow a star configuration of connections among a plurality of uni-controllers at their input connectors.
With all of the third level uni-controllers 154, 156, 158, 160 and 162, connections to dimmable lighting ballasts are made through the uni-controller output connectors J7. Those uni-controllers that affect any given ballast 146 can be determined by following the network cable 166 from the ballast to output connectors J7, and from the input/pass-through connectors J1, J2 and J3 on the same uni-controller to a network cable 166 connected output connector J7 on another uni-controller. Using this logic, one can see that a ballast connected to the output connector J7 on the uni-controller 158 is controlled by local uni-controller 158, by the Zone 2 uni-controller 152 and by the computer controller 148, and by no other controllers.

Unidirectional Signal Devices

The construction of hierarchical networks using the method of the current invention requires the use of unidirectional signal devices 130. It is these devices 130 that prevent lower level lighting controls from affecting the function of higher level controls, or exerting effects on ballasts not in direct hierarchical supervision from the control.
Fig. 13a through 13g are a series of simplified schematics depicting different embodiments of unidirectional signal devices not according to the invention. All of these devices assume an analog DC control signal, although other unidirectional signal devices are possible for alternative signal transmission schemes. In most cases, the unidirectional signal devices of Fig. 13a through 13g may be substituted one for the other, with regard to the differences in performance as outlined below.
Fig. 13a not representing the invention is a schematic of a diode acting as a unidirectional current gate. The method of unidirectional gating using a diode 86 was used extensively in discussions related to Fig. 3. Fig. 4, and Fig. 5, which used diodes 86 as examples because of their simplicity. The advantage of diode devices is that they are very inexpensive, and take little space in case of crowded electronic layouts. The disadvantage of a silicon diode, however, is that its junction voltage drop of 0.7 V is appreciable. A Schottky diode mny be used instead, inasmuch as it confers all the advantages of a regular diode, but that it exhibits a lower voltage drop.
While the diode represents a simple and practical unidirectional current gate, the following devices provide additional benefits such as current buffering. Because of their buffering capacity, these devices can provide additional "fan-out" for commutating control signals and so offer improved operation in large networks.
Fig. 13b not representing the invention is a schematic showing two uses of a transistor neting as a unidirectional current buffer. A bipolar PNP transistor 252 and an NPN transistor 254 have a signal input to the base, tie the emitter to the output, and the collector to ground or power, depending on the polarity of the device and the polarity of the signal. The PNP transistor 252 and the NPN transistor 254 are therefore used in a classical emitter follower configuration, with either a PNP or NPN employed, depending on the direction of the load. The advantages of the use of these transistors as current buffers are the current gain, in which a small input may control a relatively large output. This is particularly advantageous with large networks. However, transistors have the disadvantage of having an appreciable voltage offset from input to output due to forward junction drop across the base-emitter junction, thereby affecting the control signal that is being buffered.
Fig. 13c not representing the invention is a schematic showing two uses of a MOSFET acting as a unidirectional current buffer. An n-channel MOSFET 256 and a p-channel MOSFET 258 are shown, connected so that the gate is the input, the source is the output, and the drain is connected to either ground or power, depending on the polarity of the device and the polarity of the signal. Therefore the MOSFET 256 and the MOSFET 258 are hooked up in a classical source follower configuration. The advantage of using a MOSFET as a current buffer is the near infinite current gain from input to output. Furthermore, the MOSFETs 256 and 258 exhibit extremely high input impedance. However, there is a appreciable voltage drop between the gate and the source, caused by the intrinsic threshold voltage required for channel conduction. Furthermore, there is in general, a large variation in threshold voltage unit-to-unit for the same device type. Both of these latter two factors make MOSFETs less advantageous to use as current buffers in low-voltage control networks or where the limitations cited above degrade accuracy or control range to an unacceptable level.
Fig. 13d not representing the invention is a schematic showing an op-amp 260 arranged as a voltage follower, with a diode 86 providing unidirectionality. The advantage of this current buffer is its very high current gain and very high input impedance. Furthermore, the voltage offset caused by the diode 86 is compensated for by voltage feedback to the non-inverting input of the op-amp 260 and the high loop gain. However, the output remains one diode voltage drop higher than the op-amp output when in saturation, limiting the range of control in the control network to which this current buffer is attached.
Fig. 13e not representing the invention is a schematic showing an op-amp 260 arranged as a classical voltage follower with the output current boost transistor 252. This configuration shows an even higher current capability than a typical op-amp alone, showing both very high input impedance as well as very high output current capability. The transistor base emitter junction is compensated for by the feedback loop and high loop gain. In this configuration, however, the minimum voltage is offset by one base emitter voltage drop, limiting the range of control.
Fig. 13f not representing the invention is a simplified schematic showing an op-amp 262 with an integral NPN open-collector output structure arranged as a classic voltage follower. The advantage of this configuration is once again its very high current gain and very high input impedance. The open collector output structure of the NPN transistor integral to the amplifier 262 provides an output voltage approaching 0 Volts, limited only by the integral transistor saturation voltage. However, this configuration generally provides less fan out capability than configurations employing external, discrete transistors.
Fig. 13g not representing the invention is a schematic showing an op-amp 260 which has connected at its output an external NPN transistor. The inverting and non-inverting terminals at the input of op-amp 260 are correct for a classical voltage follower, which includes the signal polarity reversal characteristtc of transistor 254 connected in a common emitter configuration. Therefore, the inverting and non-inverting signatures shown in parentheses are correct for a classical voltage follower when the circuit is considered in its entirety. This embodiment offers the same advantages as the arrangement depicted in Fig. 13f, but employs a discrete NPN transistor connected to the output of op-amp 260, which additionally offers higher output current capabilities, and therefore higher fan out, than is typically available from general purpose, commercially available op-amps.

Uni-controller Physical and Electronic Construction

The uni-controller 138 is a convenient device for configuring a lighting control network, allowing for multiple control cooperation using shared proportional-response control lines, polling of poll-responsive environmental controllers, network hierarchical configuration, and as will be described below, network powering of control devices. Furthermore, the uni-controller 138 is inherently easy to install By careful design, as described below, the uni-controller can be made inexpensive to produce. This section describes designs for uni-controllers 138 with many of the beneficial features listed above.
Fig. 14 not representing the invention is a schematic of a typical uni-controller motherboard 168 not according to the invention. The motherboard is a central circuit providing connectivity for resident controls as well as external controls. Further, the motherboard 168 provides local voltage regulation and signal current buffering for external signal inputs affecting the local output.
The connectors J1, J2 and J3 are input/pass-through connectors wired in parallel. A plurality of connectors J4, J5 and J6 are used to connect the motherboard 168 with the photosensor 64, the infrared control 61 and the motion sensor 62, respectively. A connector J8 is provided for input from a manual dimmer directly connected to the uni-controller. Finally, the output connector J7 is provides the output control voltage for controlling light modulators (ballasts) and uni-controllers at lower levels of control. It should be noted that multiple connectors, wired in parallel with the connector J8 would provide additional flexibility in wiring lighting networks. The connectors J1, J2, J3 and J7 will in general be connected to other uni-controllers, lighting controllers, and ballasts through network cables 166, as described in Fig. 11 and Fig. 12, both not representing the invention.
Throughout the figure, pins 1 through 4 of all connectors are in parallel. External low voltage DC providing power to all controls is resident on pin 1. Pin 2 is the power return. Pin 3 is the pass-through dimming signal from controls on the same control hierarchy or above. Pin 4 is a pass-through polling wire for shared-proportional response control logic as described above, particularly for use with multiple occupancy sensors.
A voltage regulator 176 receives unregulated power from any connector Pin 1 and sends voltage regulated by a uni-controller voltage regulator 176 to all resident modules on Pin 5 of their respective connectors. Pin 2 of these connectors are the power return.
Pin 7 is the output control bus, transmitting control signals from any resident modules and the unidirectional current buffer, which in turn receives its input from any connector pin 3.
The infrared control mode switch 134 sets the functioning of the infrared control 61, to determine whether its output directly affects the internal control bus 142, or whether its signal is transmitted to the photosensor 64 in order to participate in illumination stabilization. The switch 134 connects J5 pin 7 to the output control bus or alternatively to J4 pin 8, the dynamic illumination stabilization input of the photo control. When the switch 134 is in the former position, either the photo control module or the remote control module can dominate the internal control bus 142. With the switch 134 in the latter position, the control signal from the remote control is removed from the output bus and sent to the photo control as an illumination stabilization reference. The photo control then uses this input as its internal illumination level reference set point, allowing the remote control to adjust the regulation point within the photo control.
Instead of using the output from the infrared remote receiver to set the reference voltage during dynamic illumination stabilization, such a reference voltage can be set using a manual dimmer. The connector J8 provides for a direct connection with a manual dimmer, and its signal line, on connector J8, line 7 can substitute for the remote control voltage signal.
Pin 4 on the connectors J1, J2 and J3 and on output connector J7 are pass-through connectors for polling wires, when polling logic control is desired for multiple occupancy sensors. A plurality of switches S2F and S3F are provided to separate the local motion control from the polling influence of other controls. When switch S2F is closed, motion sensors with polling wires connected to output connector J7 participate in polling with the motion sensor on the local controller. If the switch S3F is also closed, motion controllers connected to input/pass-through connectors J1, J2 and J3 participate in polling with the motion sensor on the local controller. If both switches S2F and S3F are closed, motion sensors connected with the input connector J7 can participate in polling with any motion controllers connected to input/pass-through connectors J1, J2 and J3. In all cases, the output control signal is influenced by a motion control, if present, located in the local uni-controller.
A power-indicator LED D1F in series with current-limiting resistor R1F provides local indication of that power is present.
The unidirectional signal device 130 is preferentially a unidirectional current buffer of the types described in Fig, 13b not representing the invention through Fig. 13g not representing the invention, so that it provides current amplification from pin 3 (input control bus 144) of the connectors to pin 7 (internal control bus 142) of the connectors, such that signals arriving at pin 3 are not burdened by the current requirement of a substantial number of other controls and ballasts resident in the system on pin 7.
On night, weekends, and other times, building-wide EMS control, external to the uni-controller may be used to regulate the maximum level that a significant-area of the building can achieve. However, there may be office workers, janitors, and others with need to work in areas of the building during these periods, and their work may depend on sufficient illumination. Since the building-wide EMS control may fix the maximum illumination at a low level, it is useful to have a method of disconnecting the local lighting controller from controls higher in the hierarchy, which will frequently include building-wide EMS control.
Pin 9 on the remote control connector J5 and the manual control connector J8 accepts override commands. When signals arrive at pin 9, a disconnect timer 196 is activated, which opens an electronic network disconnect switch 194. The switch 194 disconnects control line input on pin 3 from input/pass-through connectors J1, J2 and J3 from influencing the voltage on the internal control bus 142, thus preventing control from controllers attached to these connectors, including EMS control from influencing the output. The timer 196 allows this local override of network control to be of a limited duration, so that proper network control can be re-established within a reasonable period of time, which will be generally between 10 and 120 minutes. Both the timer 196 and the switch 194 are located on the motherboard so that these components can be shared by both the remote control and the manual control inputs.
Fig. 15 is a schematic exploded top-view of the physical layout of a typical uni-controller. Three boards, the photosensor board 184, the motion sensor board 186 and the combined infrared remote receiver and infrared control board 182, fit onto the three corners of the rounded triangular motherboard 168. Vertically separating the boards 182, 184 and 186 from the motherboard 168 are a plurality of standoffs 170. The electronic components of each control are not shown, except for the photodiode D1A and manual switch S2K on the photosensor board 184, and a photodiode D1H on the infrared board 182. While the infrared manual remote receiver 59 and infrared control 61 are shown to both occupy board 182, separated by line 183, these components may be placed on different boards with no change of function. Furthermore, the arrangement of controls on the different boards has no effect on the operation of the lighting control network, and some or all of the components shown on separate boards could be placed on the motherboard 168 instead.
In addition to the standoffs 170 that physically connect the motherboard 168 to the control boards 182, 184, and 186, electrical connections are made through a plurality of male connectors J4, J5 and J6 Incated on the motherboard 168 to a plurality of female connectors 174 located on the individual control boards 182, 184, and 186.
In the center of the motherboard 168 are located input/pass-through connectors J1, J2, and J3, manual input connector J8. and output connector J7. Additional input/pass-through or output connectors could also be placed on the motherboard 168, as described above. In addition, a power-indicator LED D1F is located centrally on the motherboard 168.
Fig. 16 not representing the invention is a cross-sectional schematic view of the uni-controller of Fig. 15 not according to the invention, taken along line 16-16. A uni-controller housing 178 is an external shell, preferably made of plastic, containing the electronic components, as well as means for attaching the uni-controller to a wall or ceiling (not shown). The upper surface of the uni-controller is open to allow maintenance of the uni-controller electronic components, as well as to permit the electrical connections with the lighting system network. The lower surface of the uni-controller is generally continuous, but has a number of orifices that allow sensors from inside the uni-controller to sense the environment, as well as switches and adjustments to allow user adjustment of uni-controller operation without having to remove the uni-controller from its site of installation or to remove uni-controller components. In the region of the cross-section, orifices in the external housing 172 permit photodiodes D1A and D1H from the infrared receiver board 182 and the photosensor board 184, respectively, to protrude into the space below the uni-controller, where they can see downward into the work area. In addition, the switch S2K protrudes from the lower surface of the uni-controller housing 172, permitting its adjustment without moving or otherwise disturbing the uni-controller.
The infrared board 182, the photosensor board 184 and the motion sensor board (not shown are attached to the uni-controller motherboard 168 with standoffs 170. In addition, each control board is electrically connected to the motherboard 168. For the infrared control 61, the connector J5 on the motherboard 168 is connected with connector 174 on the infrared board 182. The multiplicity of connections between the motherboard 168 and the control boards 182, 184, and 186 allow the four boards to be removed intact from the uni-controller housing 178.
The motherboard 168 normally rests on a housing ridge 178, which extends around the entire housing 172. and is secured at a plurality of points by securing extensions 180, which protrude from the housing and prevent the motherboard 168 from detaching from the housing 172. When the securing extensions 180 are pressed, possibly in conjunction with the outward flexing of uni-controller housing 172, the motherboard 168 can be removed from the housing 172.
When the uni-controller is installed, output connector J7 protrudes upwards from the motherboard 168, allowing network connections to be made from the motherboard 168 into the celling or other structure on which the uni-controller is mounted. This connector J7 is conveniently a telephone connector, allowing connections to be created or removed easily.

Methods To Power The Control Network

Power Means Overview

The control system as described above, populated with intelligent dimmers (possibly augmented by night lights), occupancy sensors, light sensors, and potentially the end-nodes of intelligent, digitally-addressed building-wide energy management systems, will require sources of low-voltage power. Under prior art, this power comes directly from the mains, and requires separate step-down transformers mounted in junction boxes within the walls or ceiling for each control. Beyond the obvious expense of the transformers, there are additional expenses of installing high-voltage conduit bringing power to the transformers.
The distributed network control of this invention involves the use of many more sensors than are typically installed in commercial buildings, perhaps by a factor of two to five from current usage. Thus, the means of powering these controls becomes un integral aspect of the feasibility of economically incorporating distributed network control of lighting systems.
The present invention teaches two different means for powering the control circuitry. The first means uses the ballast to supply power to the controls, wherein the line-isolation link of the dimming ballast not only receives dimming information from the control network, but also supplies power to the network for operating the control devices. In general, this method of powering network control devices is preferable, since it requires no more than the normal installation of ballasts and controls and the connections between them.
The block diagram in Fig. 12 not representing the invention depicts the use of these special ballasts to power a network. Ballasts with the given power supply are the special power-generating ballasts 146 of a construction to be described below. Each power-generating ballast 146 supplies enough power for one or more controls. In Fig. 12, there are 18 power-generating ballasts supplying power to 13 controls and 7 infrared remote receivers (the computer controller 148 will generally not require power from the network). The power is transferred within the network through a power distribution wire contained within the network cable 166.
Note, however, that power from the power-generating ballasts 146 is generally available in the network. For instance, the Zone 2 uni-controller 152 has no ballasts attached directly to either its input connectors J1, J2 or J3, nor its output connector J7, yet it receives power for operation from the network.
The use of power-generating ballasts 146, however, requires the use of special ballasts whose line-isolation link can supply power, and in general, there is only a limited amount of power available from each ballast to power attached controls. Thus, if the number of controls exceeds the power-generating ability of the available power-generating ballasts 146, the network will fail to regulate properly without auxiliary help.
Therefore, a second means is provided for certain applications where the use of special ballasts is inconvenient, or where there is not enough power available from the special ballasts. With the second means, special compliant power supplies are used at the most accessible locations in the network. This means requires additional components other than ballasts and network controls, and involves the installation of additional components, as well as additional connections to the mains. However, once installed, this means can handle any arrangement and combination of ballasts and controls. In addition, networks employing this means can operate using a variety of widely-commercially-available dimmable ballasts, whose low-cost may compensate for the cost of additional components and installation.
Fig. 17 not representing the invention is a block diagram depicting the use of compliant power supplies to power a network not according to the invention. The network shown is in its topological connections, configuration of controls, and designation of internal components roughly equivalent to the network of Fig. 12. As in Fig. 11 and in Fig. 12 both not representing the invention, the uni-controllers are depicted only with those internal connections related to controls while power, return and polling connections are not shown.
In this example, the power-genereting ballasts 146 depicted in Fig. 12 not representing the invention are replaced with normal dimmable ballasts 52 that do not supply power for the network controls. Instead, a plurality of network power sources 188 incorporating compliant supplies are provided at dispersed locations in the network to power network controls. As shown, these power sources 188 can be connected either to the output connector J7, as with Zone 1, Area 2 uni-controller 154, or to the input connectors J1, J2 or J3, as shown for Zone 1 and Zone 2 uni-controllers 150 and 152. The uni-controller has the power and power return pins of the connectors J1, J2, J3, and J7 in parallel. Therefore a power supply connected to any connector of a uni-controllor distributes power throughout the network. A wire for transmitting this power resides in the network cables 166.
The network shown has excess power source capacity, since each power source 188 can supply enough power each for generally for 30-60 controls. Thus, any one of the power sources shown could power the network as shown, especially if the controls are power supplies are located within a reasonable distance of one another. Depending on the current draw of each control, however, the distance between controls and the power sources may become a significant factor, and additional power sources may be needed for controls or ballasts that are distantly located.
The following sections will describe the construction of power-generating means for the network controls using the methods described above.

Line-Isolated Analog Control System

The design of the power source for the control circuitry is not completely separate from issues involving the design and characteristics of the analog control. In the first means, the control protocol characteristics are intrinsically linked with the power-generation means. In the field of control electronics, the circuit to be controlled is frequently connected to, and not isolated from, the mains. There therefore may exist a need to transmit control data across a voltage isolation boundary between low voltage control circuitry and the mains. This need arises out of the expense in material, construction, and installation costs associated with control networks that are not isolated from the source of high voltage power (mains). Such non-isolated networks are burdened with safety and fire code restrictions as may be imposed by local and national laws. The material and construction requirements thus present a cost burden that could be vastly reduced if the control network were isolated from the mains.
There are many techniques which may be employed to provide this necessary isolation. Technologies currently exploited include light or infrared light transmission, magnetic coupling, radio transmission and others. There are also many variations in topology, construction, execution, and control protocol possible within each technology.
The present invention teaches a control protocol which is a variable DC voltage with an absolute magnitude of less than that which is considered a safety hazard, and isolated from the mains. An isolation link interprets this voltage across the voltage isolation boundary to the controllable element (receiver) associated with the mains.
While such voltage control systems exist at present, several versions exhibit drawbacks which limit their usefulness. In one version of a voltage control signal system, the mains-related receiver generates a mains-isolated current source with a voltage compliance greater than 10 volts DC. The controlling element (transmitter) pulls down on the current source, thus adjusting the voltage. The resultant voltage is then sensed back on the mains side of the voltage isolation boundary. Because the isolation link itself generates the current, the current which the controlling transmitter must sink may not be insignificant and is also incrementally increased every time an additional controlled element (receiver with its isolation link) is added to the system. If enough receivers are added to the system, the current sink required can be larger than that which the controller can manage, resulting in loss of control. Even if there is sufficient current capability within the controller, the voltage drop caused by the magnitude of current across what may be a significant length of control system wire with its attendant resistance may cause undesirable inaccuracies across the network.
One version of a voltage control system employs the current source technique described above with a control voltage origin offset to two (2) volts from zero. This system does not respond to control voltages below two volts, and by doing so, affords a minimum of two volts in order to provide power for controlling circuitry. Control circuits, thus engaged, pull down the control line voltage using a shunt mode of control and circuitry simultaneously powered therefrom. This requires that the control circuit be designed to function properly on as little as two (2) volts, and notably with less current than is offered by the number of receivers in the system, for if the circuitry drew more current then the receivers offered, the control line would be pulled down by just the circuit quiescent current. Another issue in this system is that each controlled element must accurately discern that the control voltage has reached two volts. This may not easily be achieved without the expense of an accurate internal reference in the receiver. If the two volt level is not accurately interpreted by the controlled element, or the receiver-to-receiver accuracy is not consistent, system uniformity will suffer.
The limiting boundaries for the size of this system are:

1. The control transmitting circuitry in the system must not draw more current than can be supplied by the receiver current source.
2. The current sink required causes undesirable voltage drops along the network.
3. The controlling transmitter's ability to sink sufficient current to provide control.

This system may also suffer from insufficient operating voltage when operated from the remaining two volts. It will be appreciated to those skilled in the art that such a low voltage limits the type and sophistication of modern silicon based circuitry that can be employed.
It is an important aspect of an improved system that the magnitude of the sink current required by the isolation link be made as small as is practical. It is an additional important aspect of an improved system that sufficient voltage and current be made available as a source of power for complex control circuitry.

Methods for Powering Control Networks from Ballasts

Because the control network is connected to the ballasts that it controls, and the ballasts are supplied with sufficient energy to power the control network, it is convenient to power the network from the ballasts. This method of powering the network saves both in the need to provide separate power supply for the network, and in the cost of installation, given that the control network does not need to be separately wired for power.
Fig. 18A not representing the invention is a block diagram of a power-supplying dimmable ballast 146 which is designed to vary the power to gas discharge lamps in response to an input signal, while additionally supplying power to energize controls. A pair of nodes E1C and E2C are power input lines receiving power from a common power source such as AC line voltage. The dimmable ballast 146 includes a line-isolated control interface 147, which is further subdivided into a power means 248 for providing power to external controls and a control means 250 for reception of control signals from the external controls. A node E5C receives power-line-isolated control signals from an external control. A node E7C provides a source of line isolated, low voltage power for use by external controls. A node E6C represents the return for both the node E5C and the node E7C, it should be noted that in certain embodiments taught by the current invention, certain electronic components may be shared between both the power means 248 and the control means 250. Upon command from control signals received by the control means 250, variable power is provided to lamps through a high frequency AC lamp output E3C and a corresponding return E4C.
It can be seen that several elements, notably the line-isolated control interface 147, the power means 248, and the control means 250, are required to both generate line-isolated power for external controls, and to receive a line-isolated signal to control the ballast. Fig. 18B not representing the invention is a block diagram of a high-voltage enclosure (light fixture) 56 showing an alternative embodiment to Fig. 18A, wherein the line-isolated control interface elements do not reside within the ballast proper. By placing these elements at the line-isolation boundary of the high-voltage fixture 56, the rating of wires connecting to the control node E5C, the return node E6C and the line-isolated power for controls node E7C need not meet electrical requirements generally established for wires entering a high-voltage enclosure. Further, placing these elements at the line-isolation boundary obviates the need to replicate said circuitry within each ballast. Instead, a single remote line-isolated control interface 55 may service a single ballast 53 or a plurality of ballasts 53 that do not reside in the associated fixture 56. It should be noted that the ballasts 53 which interact with the remote-line-isolated control interface 55 in this embodiment are distinguished from previously described ballasts 146 in that the ballasts 53 do not have an integral line-isolated control interface 147, but use that of the separate remote interface 55 shared with other ballasts 53. In effect, the ballast 53 comprises those parts of the power-supplying ballast 146, shown in Fig. 18A, which do not include the integral line-isolated control interface 147. Remote line-isolated control interface 55 contains most of the functions of the interface 147, but may be slightly altered in components and configuration in order to perform these functions remotely.
There are a number of methods for obtaining control power from the ballast. These various methods not according to the invention are discussed in the following sections according tu the general architectures that may employed.

Methods Using Existing Magnetic Components

Fig. 19 not representing the invention is a schematic of a typical electronic ballast which is designed to provide power for gas discharge lamps and simultaneously provide line isolated, low voltage power for external use. The method employed to obtain control power involves the specialized use of magnetic components that are simultaneously powering lamps. The input terminals E1C and E2C receive common AC power. A rectifier 198 rectifies the AC power into DC at the input of a boost inductor L1M. A power factor correction circuit 202, in conjunction with the Boost inductor L1M and a diode D1M draws power from the power line in such a fashion as to reflect to the power line, a nearly resistive load. A capacitor C1M stores energy from the boost inductor L1M to provide a DC power source for a high frequency power oscillator 204. The output of the high frequency power oscillator 204 is connected to a resonant inductor L2M. The output of the resonant inductor L2M is connected to a capacitor C2M which has a value which is chosen to be near resonance with the inductor L2M at the operating frequency of the high frequency power oscillator 204. The junction of the inductor L2M and the capacitor C2M is further connected to a coupling capacitor C3M which blocks the DC component of the output of the high frequency power oscillator 204 and passes the high frequency AC component to the primary (T1M:A) of an output transformer T1M. The output transformer T1M has a plurality of secondaries T1M:B. T1M:C and T1M:D, which are representative of windings providing power to a singular or plurality of gas discharge lamps. The description given here is representative of conventional electronic ballasts, and will hereafter be referred to as the "electronic ballast lamp power circuitry".
T1M:E is a further secondary of the transformer T1M. It is the objective of the secondary T1M:E to provide a source of power line isolated power for external use. As such, the secondary T1M:E sources power by magnetic excitation from the magnetic field of the transformer T1M, but is so arranged as to be galvanically isolated from the transformer T1M and all other windings thereof, sufficient to provide a line-isolated source of power.
A diode D2M rectifies the AC voltage from the secondary T1M:E and stores the energy in a smoothing, capacitor C4M. A plurality of transistors Q1M and Q2M and associated circuitry comprise an active current limiter. The potential on the capacitor C4M is seen at the emitter of the transistor Q1M through a resistor R1M. A resistor R2M forward biases the transistor Q1M base/emitter and holds the transistor Q1M in saturation for normal loads attached to the output terminals E7C and E5C. The magnitude of external load current is reflected in the voltage drop across the resistor R1M. The value of the resistor R1M is chosen such that the transistor Q2M base/emitter will be forward biased at the desired current limit point. As the transistor Q2M is biased on, the transistor Q2M collector diverts current flowing into the base of the transistor Q1M, reducing the transistor Q1M conductivity. If an aridition external load is applied, the transistor Q2M is biased on still further, reducing the conductivity of the transistor Q1M still further. Thus, the current made available to external loads is limited to that which forward biases the transistor Q2M. A zener diode D3M regulates the output voltage to a desired level. Excess current not drawn by the load is diverted through the diode D3M in order to hold the output voltage to a desired level. Other methods of current limiting are well-known in the art and may be alternatively used. The description of the control power conditioning described here, beginning with the discussion of the use to the diode D2M to rectify the AC voltage will be hereinafter referred to as the "power conditioning circuitry." Together, the secondary T1M:E and the power conditioning circuitry constitute the power means 248 indicated in Fig. 18 not representing the invention.

Methods Using Added Magnetic Components

Fig. 20 not representing the invention is a schematic of a typical electronic ballast which is designed to provide power for gas discharge lamps and simultaneously provide line isolated, low voltage power for external use. The method employed to obtain control power involves the use of added magnetic components that are not used to power lamps. The electronic ballast lamp power circuitry is similar to that described for Fig. 19.
The secondary T1M:E is a further secondary of the transformer T1M. It is the objective of the secondary T1M:E to provide a source of power for external use. As such, the secondary T1M:E sources power by magnetic excitation from the magnetic field of the transformer T1M. The arrangement of the secondary T1M:E is such that it does not provide for sufficient galvanic isolation from the power line. A transformer T2M is designed to provide galvanic isolation between the secondary T1M:E and the line-isolated outputs E7C and E5C. A pair of secondary windings T2M:A and T2M:B of the transformer T2M are so arranged to provide sufficient insulation to provide galvanic isolation across the device. The power conditioning circuitry is the same as described for Fig. 19.

Methods Using Added Magnetic and Oscillating Components

Fig. 21 not representing the invention is a schematic of a typical electronic ballast which is designed to provide power for gas discharge lamps and simultaneously provide line isolated, low voltage power for external use. The method employed to obtain control power involves the use of added magnetic components that are not used to power lamps, as well as an added oscillating component. The electronic ballast lamp power circuitry is similar to that described for Fig. 19.
The power factor correction circuit 202 contains active circuitry which requires low voltage power. As such, the power factor correction circuit 202 generally includes circuitry for providing low voltage power that is power-line related. An oscillator 208 makes use of some of the low voltage line isolated power from the power factor correction circuit 202. The output of oscillator 208 is coupled to the isolation transformer winding T2M:A. The windings T2M:A and T2M:B of the transformer T2M are so arranged to provide sufficient insulation to provide galvanic isolation across the device. The power conditioning circuitry is the same as described for Fig. 19. A detailed example of this embodiment will be described later.

Methods Using Capacitive Components

Fig. 22 not representing the invention is a schematic of a typical electronic ballast which is designed to provide power for gas discharge lamps and simultaneously provide line isolated, low voltage power for external use. The method employed to obtain control power involves the use of added capacitive components. The electronic ballast lamp power circuitry is similar to that described for Fig. 19.
Some of the energy present at the junction of the resonant inductor L2M and the resonant capacitor C2M is coupled to a capacitor C5M. AC energy coupled through the capacitor C5M is rectified by the diode D2M and stored in the smoothing capacitor C4M. A diode D4M restores the charge on the capacitor C5M. The AC reference return E5C may be attached to earth ground or any potential with a sufficiently low AC impedance path returning to the AC line inputs E1C and E2C to provide a current path providing energy for outputs E7C and E5C. The value of the capacitor C5M is chosen such that the impedance of the capacitor C5M is sufficiently low as to provide a current path at high frequency, whilst having a sufficiently high impedance at power line frequency, such that only negligible power line frequency current will flow. Thus, the capacitor C5M may be regarded as a galvanic line isolation component. The rest of the power conditioning circuitry is the same as described for Fig. 19.

Separating Control Power from the Control Signal

The seemingly opposing requirements of low control current and sufficient current for circuit power, and sufficient voltage at all times for complex circuitry, can be achieved by removing the requirement that the control voltage also power the control circuitry. With this accomplished, the current required for control may be reduced to a level only bounded by the circuitry of the isolation link, and a now separate voltage can be made available for the control circuitry at all times. Further, if the two volt minimum specification is deleted, the origin of slope for the control voltage can be set at zero (0) volts, eliminating the receiver internal reference voltage requirement. The current invention includes a control system with the characteristics just described. This novel control system employs magnetic coupling within a single transformer to simultaneously send control information from transmitter to receiver while sending substantial low voltage power from receiver to transmitter. Further, this novel system requires only extremely small control currents and responds properly to a control signal with an origin of slope of zero (0) volts.
The circuit functions as a specialized oscillator which is powered from the receiver side. The oscillator has two distinct states a first state which is fixed by timing elements during which time energy is accumulated in a transformer core, and a second state during which time this energy is released equally through two of the transformer windings. These equal energies are stored as equal voltages on two capacitors which reside on opposite sides of the voltage isolation boundary and represent the control voltage as transmitted and as received. The time to release this energy is variable and dependent on the control voltage being transmitted.
Operation is best understood by following an embodiment, depicted in Fig. 23 not representing the invention, a circuit diagram of a line-isolation control and power interface, through a cycle of operation. It should be noted that this embodiment corresponds in part to the method for powering controls using added magnetic and oscillating elements, as depicted in Fig. 21 not representing the invention.
Central to the circuit is a transformer T1P, which is a 4 winding ferrite cure transformer with equal turns an all windings and a specified inductance. The transformer windings T1P:A & B are closely coupled but voltage isolated from the windings T1P:C and D. Assume that an integrated circuit U1P has been triggered, turning on a transistor Q1P and impressing Vcc across the winding T1P:A. The transistor Q1P remains on for a time controlled by a resistor R1P and a capacitor C1P. The "on" time created by the integrated circuit U1P is stable for changes in supply voltage and temperature, owing to its ratiometric operation (similar to an industry-standard LM555N manufactured by National Semiconductor). The polarity of winding T1P:D is such that a diode D2P is now conducting and energy to operate an integrated circuit U8P and to offer, to external control circuitry is being stored in a capacitor C6P The transformer windings T1P:B & C polarity are negative and a plurality of diodes D4P and D5P are not conducting. Also during this time, a transistor Q4P is on. The transistor Q4P conduction disables an integrated circuit U7P, a controllable current mirror. During this time, a predictable amount of energy is stored in the core of the transformer T1P -- predictable because the transformer T1P inductance is known, as is the charge time and voltage across the winding T1:A.
The integrated circuit U1P now times out and the transistor Q1P shuts off. The energy stored in the transformer T1P's core now reverses the polarity of all windings. The diodes D4P and D5P now conduct into a plurality of capacitors C5P and C7P, respectively. The capacitors C5P and C7P are made large enough that no appreciable voltage changes across them due to an individual cycle. Current continues to flow as the transformer T1P discharges. An integrated circuit U5P comparator senses the polarity across the diode D5P. When all energy in the transformer T1P is discharged, the voltage on all windings begin to collapse. The voltage across the diode D5P starts to reverse causing the integrated circuit U5P output to go low. The integrated circuit U5P going low sets an integrated circuit U2P NOR gate output high and triggers the integrated circuit U1P beginning the next cycle. During this portion of the cycle, the transistor Q4P has been off, enabling the integrated circuit U7P current mirror.
The relationship of the current in the integrated circuit current mirror U7P and the energy discharged by the transformer T1P is an important factor in providing control signal replication accuracy. An integrated circuit U8P is a current buffer which clamps the voltage on the capacitor C5P to the value of the control voltage fed into a control line E2P. Since this voltage is clamped, the voltage across the transformer winding T1:C and thus across the winding T1:B will also be clamped. It is an objective of this circuit to replicate, as nearly as possible, the voltage found on the capacitor C5P, onto the capacitor C7P. This can only be accomplished if the current in both circuits are equal. With equal loads, the voltage across the winding series resistance and across the junctions of the diodes D4P and D5P will be very closely aligned. As for matching the diodes: this can be done to enhance accuracy, but if the proper type is specified, and the circuit is built from the stock of one manufacturer, there is no need. What is important is to load both sides of the circuit equally. This implies that 1/2 of the average current available from the transformer T1P must flow in each leg of the circuit. Since the average total current available per transformer T1P discharge cycle is equal to 1/2 the peak current stored in the transformer T1P, the value of current in the integrated circuit U7P should be set to Ipeak/4 by a resistor R6P. Thus 1/2 of all the available current is sunk by the integrated circuit U7P and the balance, (1/2) must be sunk by the output of an integrated circuit U8P in order to achieve clamping. The capacitors C5P and C7P provide the high frequency clamping effects required. The net result is a very accurate relationship between the voltage on the capacitors C5P and C7P. The integrated circuit U7P is disabled during the charge cycle of the transformer T1P by the transistor Q4P. If it were not, the current would not properly average Ipeak/4 over the range of voltages and subsequent charge/discharge duty cycles under which the circuit operates.
It is possible to "stall" the oscillator 208. Sometimes at startup, or when the control voltage signal slews from high to low quickly, the oscillator 208 may stop. This is similar to startup issues encountered with chips such as the Linfinity LX1562 (a PFC controller) and others that rely on "end of current" sensing. This is circumvented by averaging the voltage at the output of the integrated circuit U5P. The integrated circuit U5P has a duty cycle that varies with the control voltage. At low control voltages the average output voltage is high, at high control voltages, the average falls almost to 50% of Vcc. A voltage below 50% of Vcc indicates a halted oscillator 208. An integrated circuit U3P output now goes high, biasing the integrated circuit U5P non-inverting input up through the diode D1P. The result is a flip of the integrated circuit U5P' output which triggers the integrated circuit U1P and restores normal operation. A monolithic solution would perhaps be a missing pulse detector with internal timing components.
An integrated circuit U4P is a noise blanking one shot which triggers just as the transistor Q1 turns off and the integrated circuit U5P output goes high. When the control voltage is very low, and/or when there is a significant load reflected to the transformer T1P by the load at a voltage node E1P, leakage inductance in the transformer T1 may allow the voltage on the diode D5P anode to "ping". This can cause multiple triggers from the integrated circuit U5P output unless inhibited. The integrated circuit U4P need not be externally programmable. About 1-2 µs is all that is required.
A diode D3P in conjunction with a resistor R5P prevent negative voltages on the transformer winding T1:B from forward biasing the substrate at the integrated circuit U5P non-inverting input.
Low voltage is offered to the external control circuitry via the power source E1P with return on a voltage return E4P. The voltage on a capacitor C6P is current limited by the combination of a plurality of transistors Q2P and Q3P, and an integrated circuit U9P. A resistor R3P sets the current limit. Another resistor R4P sets the pull down current for controls attached to the control line E2P at roughly 10 µa at 1 volt, while a resistor R7P and a capacitor C4P provide input filtering and reverse polarity protection for an integrated circuit U8P.
One important facet of the design is the inherent power supply rejection. Once the ratio of inductance, integrated circuit U1P "on" time, and integrated circuit U7P current is set, Vcc may be varied widely without appreciable impact on accuracy. This is because the energy stored in the transformer T1 and the current sunk by the integrated circuit U7P are both directly proportional to Vcc.
Another important aspect of the circuit is that it may be converted to emulate the input current of the present de facto standard by decreasing the value of the resistor R4P until a short circuit value of about 0.5 ma is attained on the control line E2P. Alternately, the resistor R3 can be increased such that the current limit is 0.5 ma and voltage source E1P may be shorted to the control line E2P. This provides a true current source and removes the availability of an external power supply.

Methods for Powering Control Networks with Compliant Power Sources

The use of the auxiliary power obtained from the ballasts, as described above, has the potential disadvantage that the circuitry for generating the line-isolated power must be built into the ballasts. Because this implies a significant number of these ballasts, the cost of including this circuitry within each ballast may be significant.
Alternately, the current invention alternatively teaches the use of one or more power modules placed throughout the network, providing distributed power. The use of compliant, current-limited power sources allows the cohabitation of many of these power sources in the network. Because each of these sources can provide the power for numerous controls (up to 50), the cost of the power source is spread among a large number of controls.
Fig. 24 not representing the invention is a simplified schematic of a compliant, current-limited power supply for use as a distributed power source in a lighting control network. The power supply converts AC line voltage to line-isolated low voltage DC. A switch 190 resident in the power supply allows the use of a common transformer T1E with multiple input voltages. A pair of diodes D1E and D2E rectify the low voltage output of the transformer T1E for filtering by a capacitor C1E. A current-limited voltage regulator 192 affords a compliant source of current at the voltage regulation point to an output E3E so that multiple power supplies may have their outputs tied together and contribute with nominal equality to the power required by the system regardless of their physical location Alternately, the current limited voltage regulator 192 may be replaced with a constant current regulator.

Self-Energizing Contactors

In the field of electrical power, and more specifically in the field of switched electrical systems, it is often more advantageous to switch electrical power to a load with a power relay or contactor rather than with a mechanical power switch. The use of a contactor or power relay may be advantageous in the following situations:

1. When the electrical load imposed upon a switch is rather high, perhaps exceeding the current rating of a switch that is practical, for the chosen location.
2. When such a switch would share space with low voltage or unrelated wiring and circuitry.
3. When the switching control element is not a mechanical switch, but a low current, low voltage circuit.

The relay may be fitted to the system in a location more appropriate to the flow of power than the position of the actuating control, which may be located some distance away, in a location convenient to system control.
Because contactors or relays are electromagnetic devices, a low voltage power source is often used for activation. This voltage is usually line voltage isolated, local earth ground referenced, and produced by an isolating step-down transformer. The contactor electromagnetic coil is energized by connecting it to the isolated low voltage source by a switch. Thus the switch needs carry only the current necessary to activate the contactor coil and not the load current.
It is obvious that the source of power for the isolating step-down transformer must be the power line in order to provide turn on energy. This is disadvantageous because it draws some magnetization current (power). Furthermore, there is also the potential for the annoyance of audible hum coming from the transformer laminations at the power line frequency. Further, the switch must either be located at the primary, where it must be rated for the voltages encountered at the mains, or located at the secondary, where it must be rated for the current of the contactor coil. The transformer also takes up space and can be expensive.
A novel means has been devised that is capable of activating the coil of a power relay or contactor from a source of energy not derived from a transformer, yet low enough in voltage and current to be considered non-power-line related.
Regardless of the source of line power, or its voltage or format, all circuits have one thing in common: all types avoid using earth ground as a current carrying conductor. Indeed, for safety and fire reasons, special ground fault detection circuitry is frequently employed to shut off power if ground currents beyond a specified amount are detected.
Although ground currents are undesirable, it is not uncommon for many electrical circuits and line powered products to leak small amounts of current to earth ground. These currents may be a deliberate side effect of components within some products. For example, products which contain circuitry that produces conducted noise at high frequency may employ capacitors tied to earth ground as a noise suppression measure. While small in value, these components still represent some conductance to ground at the power line frequency. Small leakage currents may also be attributable to the imperfect nature of insulators used in construction of the product. The upper limit for this leakage current is generally governed by local or national safety codes and insurance underwriting organizations.
A novel approach to the activation of an electromagnetic coil such as is employed by a relay or contactor is to make use of an acceptably small leakage current tn provide the activation signal. Because the leakage current in this novel system flows to earth ground, the voltage relative to earth ground can also be limited to a low and safe value. Further, the energy employed by this novel circuit means may be diverted to an indicator when not used for activation. Thus, the circuit may also indicate the state of activation, or provide a pilot light or a means to locate the switch during low ambient light conditions.
Fig. 1 employs such a self-energizing relay. A self-energizing relay 80 receives power from an input line power wire 76. In the manner mentioned above, and described in more detail below, the self-energizing relay 80 derives sufficient line-isolated power in order to energize a pilot light on the manual dimmer with pilot light and on/off switch 60, whose switch is connected to the self-energizing relay 80 by line-isolated control wire 78. The pilot light serves to indicate to the user the location of the manual dimmer 60 in low light conditions. Normally, since the local control network is supplied power from the dimmable ballasts 52, the network at this time nominally has no source of power. However, the self-energizing relay provides power for its own energization, under control of the on/off switch of the manual dimmer 60. Once the manual dimmer 60 activates the relay 80, power is provided to the ballasts 52 through a plurality of line-related power wires 82.
The self-energized contactor enables closure of an AC power relay or contactor by a small leakage current accumulated between the AC power line voltage and earth ground during each half cycle of line voltage. Fig. 25 is a simplified schematic of a self-powered contactor energizing circuit. Current to charge a capacitor C1T is drawn from an input E1T AC high, through a resistor R1T with a return path from the other side of the capacitor C1T to an earth ground E2T. The current is small, but sufficient to charge the capacitor C1T to the trigger voltage of a diac Q1T, once each half cycle of the AC line. With a switch S1T open, the diac Q1T discharges the capacitor C1T through the bridge of a plurality of diodes D1T, D2T, D3T, and D4T depending on polarity to the LED inside an opto-isolator U1T. The pulse of current into the opto-isolator U1T activates an isolated triac Q2T at the output of the opto-isolator U1T. This in turn triggers the gate of the triac Q2T, which completes the conduction path for the armature of a relay K1T to an AC low E3T. The resistor R2T limits the current into the gate of the triac Q2T. With the triac Q2T in conduction the relay closes, making power available to an external load.
The bi-directional nature of the active components provides for triggering and conduction for both polarities of line voltage. When the switch S1T is closed, a second conduction path for the capacitor C1T discharge is provided. This path, through the switch S1T and a pair of diodes D5T or D6T depending on voltage polarity, is lower in voltage drop than the voltage required for the diode bridge and the LED inside the opto-isolator U1T to conduct owing to the number of junction drops. Thus, the opto-isolator U1T will not trigger. The current discharged from the diac Q1T forward biases the LED's D5T and D6T and provides an indication when the switch S1T is closed.

Benefits And Advantages Of the Present Invention

The invention and the illustrative examples shown provide a number of benefits and advantages in the design and implementation of dimmable fluorescent lighting systems:

The coordinated response of multiple controls all controlling the same ballasts, as allowed by the present invention, provides the capability to tailor the system response with great precision and flexibility. As detailed above, the combination of photosensor, occupancy sensor and infrared remote receiver allows control which can respond appropriately to a wide number of input stimuli normally observed in work environments. Local networks can be coupled with hierarchical control to permit even a wider range of responses.
The use of ballasts to power control networks avoids the need to provide additional power source components and wiring. This decreases the cost of materials, but perhaps more importantly, it substantially reduces the costs of lighting system installation since additional wiring and labor are not necessary. This is particularly important in retrofit situations, since the infrared remote control makes wiring inside walls unnecessary. The reduced costs of installation in retrofit situations means considerably greater financial incentives to install control networks, with their possibilities of significant energy savings. The use of network power with compliant power sources, strategically located in the network, has comparable economic benefits.
Lower cost of installation is also supported by the distributed intelligence of network control according to the methods of the present invention. Central lighting management units have the disadvantage that all controls must be directly connected to the central management unit, irrespective of the distance of the controls from the unit. Furthermore, such central management units generally have fixed limits on the size and complexity of networks allowed. The present invention, however, has no central management unit, and provides for a variety of configurations, including star and daisy-chain networks, or combinations of these and other configurations. Adding new units requires wiring only to the closest member of the existing network. Furthermore, the networks are configured from the topology of connections according to simple rules, simplifying the jobs of lighting engineers.
The methods of the present invention enable and encourage the controls to be housed in single enclosures, which we have named uni-controllers. These uni-controllers may share common electronics, such as power management, network communication and connectors, and interfacing with infrared remote receivers, as well as sharing the mechanical housing substantially reducing the cost of manufacturing network controls in comparison with producing separate controls. In addition, the cost of lighting control network installation is substantially reduced, since only a single housing must be affixed to the ceiling and electrically connected to the network.
The network wiring of remote control receivers with other controls, such as photosensor and occupancy sensor, enables a variety of novel calibration methods not according to the invention. With prior art, the labor cost of calibrating controls contributed to a large overall cost and reduced the financial intentives of installing lighting controls. The network design of the present invention provides for calibration that is either automatic or substantially facilitated, with the potential of reducing calibration labor by a factor of many fold.
The combination of infrared remote control and photocontrol enables a novel means of lighting control not according to the invention. Current photocontrol (daylight harvesting) usually allows the user only a single lighting set point (that is, artificial and natural light combined). Dynamic illumination stabilization not according to the invention, combined with the use of Total Effective Illumination (TEI) allows the user to variably set the total amount of lighting, providing improved personal control over lighting. In addition, because the user will in general be manually lowering the lighting from the maximal illumination found in conventional daylight harvesting, dynamic illumination stabilization not according to the invention will in general enhance energy savings.
Occupancy sensors were at the forefront of the lighting energy-savings movement of the 1980's. Their wider use, however, has been limited by a number of drawbacks. Firstly, turning lights on and off repeatedly as is generally done, substantially reduces lamp life, and therefore reduces the financial incentives of installing occupancy sensors. In addition, in large or topologically complex areas, single occupancy controls were inadequate. For example, bathrooms are frequent sites for occupancy sensor installation, but single sensors are unable to scan beyond the partitions of a multitude of stalls. The methods of the present invention include the ability to modulate lights rather than to turn them on and off, affording significantly longer lamp lives with comparable energy savings. The present invention furthermore teaches dual timing mechanisms to make occupancy sensors more pleasant and safe. Furthermore, the polling mechanisms of the present invention permit occupancy sensors to intelligently communicate with one another so as to allow multiple occupancy sensors to scan a largo or complex area as if they were a single sensor. Together, these improvements on the functionality of occupancy sensors should enhance their presence in lighting control, affording large energy savings.
While the financial benefits accruing to the energy-savings afforded by lighting control networks is large and easily quantifiable, the benefits due to increased worker productivity may be far greater. Giving workers more ergonomic lighting and personal control may decrease worker sick days, increase satisfaction, and reduce eye strain. Over the lifetime of a lighting control system, these more subtle benefits may significantly outweigh the financial returns through energy-savings.

In a typical commercial office building, lighting comprises approximately half of the total energy usage. In a study carried out by the Lighting Research Center at the Rensselaer Polytechnic Institute in Troy. N.Y., a controlled test of the energy-savings possible by the use of lighting controls was performed. This test used unsophisticated conventional lighting controls, and did not involve the use of energy management systems, from which substantial energy savings may be projected. Nonetheless, this test indicated that perimeter offices with lighting control systems in perimeter offices required 40 to 75 percent less energy for lighting. Thus, it appears possible that the widespread use af the methods of the present invention may provide the potential for significant savings, even on a national scale.
It should be apparent to one skilled in the art that the above-mentioned embodiment according to the invention are merely illustrations of a few of the many possible specific embodiments of the present invention. Numerous and varied other embodiments can be readily devised by those skilled in the art without departing from the scope of the present invention as defined by the following claims.

Claims

A lighting control system for controlling the amounts of illumination emanating from each of a plurality of artificial lights (94) for illuminating corresponding areas, the lighting control system comprising:
a) a plurality of lighting controllers (84),

b) a plurality of lighting modulators (88), each of which provides a variable amount of power to the corresponding light (94) and is adapted to vary the power to the corresponding light (94); characterised by further comprising

c) an electrically-conductive control wire (58) connecting the controllers (84) and the lighting modulators (88), in which a control signal is shared between the controllers (84) and the modulator (88), this control signal having a default value; wherein each lighting controller (84) comprises in uni-directional signal device (130);
wherein each of the uni-directional signal devices (130) in the controller (84) is adapted to change the level of the control signal uni-directionally from said default value, if the controller (84) takes control;
wherein further the control signal is fixed at the level of the particular lighting controller (84) attempting to set the level of the control signal furthest from the default value, and
wherein the lighting modulator (88) varies the power to the corresponding light (94) in response to the value of the control signal.
The system according to claim 1, wherein the uni-directional signal device (130) is a pull-down diode (86), a unidirectional current gate or a current buffer.
The system according to claim 1 or 2, wherein at least two controllers (84) are poll-responsive controllers (124), for example occupancy controls (62), the system further comprising:
h) an electrically-conductive polling wire (126), which connects at least two poll-responsive controllers (124) and carries a polling wire signal value, wherein the poll-responsive controllers (124) are adapted to switch the polling wire signal value from an inactivated value, for example representing the absence of people in an area under surveillance, to an activated value, for example representing the presence of people in the area under surveillance. and wherein when at least one poll-responsive controller (124) connected to the polling wire (126) has switched to the activated value, for example after having detected people within the area under surveillance, the activated value of the polling signal forces all poll-responsive controllers (124) connected to the polling wire (126) to attempt to modulate the signal in the control wire (58) to a lighting level suitable for the illuminated areas.
The system according to anyone of the claims 1 to 3, wherein a light (92.) is a gas discharge lamp, for example a fluorescent lamp (50).
The system according to anyone of the claims 1 to 4, wherein the controllers (84) are chosen from the set comprising occupancy controls (62), photocontrols (64), computer controls and remote control receivers (59).
The system according to anyone of the claims 1 to 5, further comprising
e) a regulated power supply; and

f) an electrically conductive power wire (66) connecting the power supply and at least one control (84), wherein regulated DC voltage power flows in the power wire (66) for energizing the controllers (84) connected to the power wire (66).
The system according to anyone of the claims 1 to 6, further comprising
g) an infrared control receiver (59) adapted to receive signals transmitted from an infrared transmitter and to communicate these signals to a control (61, 62, 64) through a separate infrared control receiver signal wire (69, 67, 65), wherein the signals include both mode-changing signals and remote lighting level requests.
The system according to anyone of the claims 1 to 7, further comprising
i) a power wire, connected to AC line voltage, supplying power to a lighting modulator (88); and

j) a self-energizing line contactor (80) used to switch AC line power to a modulator (88), wherein energy used in controlling the contactor (80) is stored transiently from a line-related AC source (E1T) within an earth-ground referenced energy-storage device (C1T), which is released through an opto-isolator (U1T) when the contactor (80) is in a switched state, wherein the opto-isolator (U1T) triggers an electric switch (Q2T) so as to energize a relay coil (K1T) using a source of line-related AC voltage used in powering the lighting modulator (88).
A method for controlling lighting in an area using a plurality of lighting modulators (88) and a corresponding plurality of artificial lights (94), for example gas discharge lamps (50), wherein the amount of illumination emanating from each light (94) is controlled by the corresponding modulators (88), the method comprising
a) providing at least one lighting controller (84) chosen from the set comprising occupancy controls (62), photocontrols (64), computer controls and remote control receivers (59); characterized by further comprising the steps of

b) providing common electrical connection between said plurality of lighting modulators (88) and said at least one lighting controller (84) using an electrically-conductive control wire (58), in which an analog electrical control signal with a default value is established;

c) regulating the electrical control signal with the lighting controllers (84), wherein all controllers (84) regulate the level of the control signal uni-directionally from the default value;

d) fixing the control signal to a value at the level of the lighting controller (84) that is attempting to set the level at a value that is furthest from the default value; and

e) affecting the illumination in the area by means of the lighting modulators (88), wherein the lighting modulators (88) are responsive to the level of the electric control signal in the control wire (58).
The method of claim 9, wherein the controllers (84) each comprise a uni-directional signal device (130) and wherein at least two controllers (84) are poll-responsive controllers (124), the method further comprising the steps of:
f) providing common electrical connection between said at least two poll-responsive lighting controllers (124) using an electrically-conductive polling wire (126) in which an analog electrical polling wire control signal with a polling value is established;

g) making an electrical connection between the control wire (58) and the polling wire (126) mediated by said uni-directional signal devices (130) of the poll-responsive controllers (124), in which analog electrical control signals are uni-directionally transferred from the polling wire (126) to the control wire (58);

h) adjusting the polling control signal in the polling wire (126) by means of the poll-responsive controllers (124), for example occupancy controls (62), connected to the polling wire (126) from an inactivated value, for example representing the absence of people in an area under surveillance, to an activated value, for example representing the presence of people in the area under surveillance, and when at least one poll-responsive controller (124) connected to the polling wire (126) has switched to the activated value, for example after having detected people within the area under surveillance, in response to the activated value of the polling control signal forcing all poll-responsive controllers (124) connected to the polling wire (126) to attempt to modulate the signal in the control wire (58) to a lighting level suitable for the illuminated areas; and

i) fixing the control signal in the control wire (58) to a value at the level of a device that is attempting to set the level at a value that is furthest from the default value, wherein said device is a member of the set comprising the lighting controllers (84) connected to the control wire (58) and the poll-responsive controller (124) connected to the polling wire (126).