WO1993007548A1

WO1993007548A1 - Control device for controlling the energizing of an oscillating motor for driving a compressor unit of the resonant-piston type, and compressor unit comprising such a control device

Info

Publication number: WO1993007548A1
Application number: PCT/NL1992/000180
Authority: WO
Inventors: Antonius Adrianus Johannes Benschop; Fidelus Adrianus Boon
Original assignee: N.V. Philips' Gloeilampenfabrieken
Priority date: 1991-10-09
Filing date: 1992-10-09
Publication date: 1993-04-15

Abstract

A compressor unit (1) comprises a compressor (2) of the resonant-piston type and an oscillating motor (3) for driving a compressor (2). Furthermore, the compressor unit (1) includes a control device (30) for energizing the oscillating motor (3) with an AC voltage having a constant amplitude. The control device (30) derives from the motor voltage and the motor current a measuring signal which is indicative of the ratio of the power consumption of the motor (3) to the squared amplitude of the motor current. The efficiency of the compressor unit is maximized for a frequency for which said ratio reaches an extreme value. The optimum frequency may thus be found on the bases of said ratio.

Description

Control device for controlling the energizing of an oscillating motor for driving a compressor unit of the resonant-piston type, and compressor unit comprising such a control device.

The invention relates to a control device for controlling the energizing of an oscillating motor for driving a compressor unit of the resonant-piston type, comprising a energizing circuit for supplying the motor with an AC voltage from an electric power source, and a detection circuit for deriving at least on the basis of the motor current a measuring signal which contains information about the efficiency with which the compressor unit is operated.

The invention further relates to a compressor unit comprising such a control device.

A control device as well as a compressor unit of the type mentioned hereinbefore are known from EP-A-0 266 835 (PHN 11.926). The control device described therein energizes the oscillating motor by means of a DC voltage-to-AC current converter, while the measuring signal is then determined as the electric power consumption of the motor. With a maximum efficiency of the compressor unit this motor power consumption is maximum. The DC voltage-to-AC current converter comprises a feedback current control in which the current flowing through the motor is measured and compared with the reference value and, subsequently, the energizing is adapted in response to the difference between the measured current and the reference value. For a proper operation of the current control the current control is to be realised with high-speed power electronics which is relatively expensive.

It is an object of the invention to provide a control device in which the measuring signal containing information about the efficiency of the compressor unit can be derived in a simple and reliable fashion without the need for a control of the motor current.

This object is achieved in a control device according to the opening paragraph, which is characterized in that the energizing circuit is arranged for energizing the motor with an AC voltage having an essentially constant amplitude, in that the detection circuit comprises deriving means for deriving a signal as a measuring signal that has a predetermined relation to the ratio of the electrical power consumption of the motor to a function F(I) of the amplitude I of the current consumption of the motor, so that the maximum P/F(I) ratio occurs at an optimum frequency which essentially coincides with the frequency at which there is maximum motor efficiency, and in that the P/F(J) ratio monotonously increases for frequency values below said optimum frequency and the P/F(I) ratio monotonously decreases for frequency values above said optimum frequency. When the control device according to the invention is used, the power supplied to the motor may simply be controlled by energizing the motor with a pulse- width modulated voltage from a DC voltage source. The AC voltage supplied to the motor is then directly determined by the DC voltage level of the DC voltage source and the duty cycle of the pulse-width modulated voltage.

Further embodiments as well advantages thereof will be described in detail hereinafter while reference is made to the drawing Figures 1 to 7, in which: Fig. 1 shows a compressor unit,

Figs. 2a, 2b, 2c and 2d show a number of different parameters plotted against frequency,

Fig. 3 shows a control device to be used in the compressor unit, Figs. 4, 5 and 6 show flow charts of embodiments of control programs, and

Fig. 7 shows an embodiment of an energizing circuit to be used in the control device.

Fig. 1 shows a compressor unit 1, comprising a compressor 2 of the resonant-piston type and an oscillating motor 3 for driving the compressor 2. Such a compressor 2 and oscillating motor 3 are described in detail in European Patent Application EP 0 155 057 (PHN 10.965C). This compressor 2 has a housing 4 which bounds a cylindrical cavity 5 in which the piston 6 is movable. At one end the cavity 5 is closed by a plate 7 in which, a gas inlet valve and a gas outlet valve 9 are mounted in a customary manner. The other end of the cavity 5 is closed by a plate 10, forming a variable-volume chamber with one end of the piston 6, which chamber contains an amount of gas. This variable-volume chamber forms a gas spring. The other closed end of the piston 6 constitutes one of the walls of a variable-volume chamber in which, for example, the refrigerant of a refrigerating system is compressed, the refrigerant being supplied and discharged through the valves 8 and 9. By means of a lever 11 the piston is coupled to an armature 12 of the oscillating motor 3. The armature 12 is rotatably mounted on a motor shaft 13. The oscillating motor further comprises a stator 14 (14A, 14B) and a core 15 with two coils 16A and 16B arranged opposite one another. Two magnets 17 and 18 are arranged between the core 15 and the stator 14. At the end faces 19 of the coils 16 which faces are remote from each other, circularly curved air gaps 20, 21, 22 and 23 are formed in which sliding elements 24 and 25 of the armature 12 are movable. If in such an oscillating motor 3 the coils 16 are energized with an AC current, the magnetic field generated in the air gaps (20, 21, 22, 23) by the magnets 18 and 19 and the coils 16 will cause the sliding elements 24 and 25 of the armature to perform an oscillating motion in the circular air gaps 20, 21, 22 and 23 with a frequency corresponding to the frequency of the energizing current through the coils 16. This oscillating motion of the armature 12 is transferred to the piston 6 via the lever 11.

The efficiency of the compressor unit reaches its maximum value in the neighbourhood of the resonance frequency of the compressor unit. This efficiency of the compressor unit may be expressed as follows in the following relation:

p ^ - ccaomm __ p ^••'^c• CCoott:: - P ^cil.e - p ^cl>m ι ■_{* \} H .Λ ^{β "}p— * p ^{( 1 )} toe tot

where Tjφ represents the efficiency of the compressor;

P_com represents the power created in the refrigerant by the compressor;

P_tot represents the electronic power supplied to the motor;

P_le represent the electric losses in the motor;

P_jjjj represent the mechanical losses in the compressor unit; f represents the frequency of the energizing current of the motor. The electric losses P_Ie are constituted by what is commonly referred to as copper losses P^- and what is commonly referred to as iron losses P_fe.

The copper losses P_cu are given by the following relationship: P_cu = I² x R (2) where R represents the electrical resistance of the motor windings and where I represents the effective current flowing through the motor windings.

The what is commonly referred to as iron losses Pf_e are the result of the alternating magnetic flux through the iron material of the motor. These losses are a function of the frequency (the number of flux alternations per time unit and the motor ' current). The iron losses strongly depend on the material used in the motor and the geometry of the motor. The iron losses, however, may always be approximated well by the following function:

Fd, I², f , ft (3) A combination of the relationships 1, 2 and 3 leads to the following relationship for the efficiency:

_{η (f)} _. ^l-JPta _ I² X R + F (I, , I² , , f^ f²) tot tot

As has already been published in said Patent Application EP-A-0, 155,057, a frequency f_opt for which the efficiency of the motor has been found to be optimum may be determined by supplying an AC current having constant amplitude and variable frequency f to the motor and then determining the power consumption of the motor. Due to the fact that the amplitude of the current is constant at the frequency f while there is a maximum motor power consumption, there is also maximum efficiency of the motor. In that case the- electrical losses are to be considered substantially constant. For this matter, the effect of the frequency f on the electrical losses P._e is small compared with the effect of the current on these losses.

By way of illustration the power consumption P for a motor of the type described with reference to Fig. 1 is plotted against the frequency f (in Fig. 2a) for the case where the motor is energized with an AC current having constant amplitude. Although the optimum frequency f_opt may very well be determined on the basis of the variation of the power consumption P, this method has the drawback that the motor is to be energized with an AC current having constant amplitude. For:

In order to maintain the amplitude of the motor current at a constant level, the motor current is measured and the energizing of the motor is adjusted in response to this measurement. Such a motor current measurement and adjustment of the energizing operation demands additional components and is therefore disadvantageous for reasons of cost price;

The motor compressor unit presents its highest efficiency when driven by a sinusoidal current. In that case there are hardly any higher harmonic distortions which contribute to losses, but hardly contribute to the power supplied to the compressor.

Energizing the motor with a sinusoidal AC current also demands additional components for current control through the motor. It is simpler to energize the motor with a square-wave AC voltage, it is true. But this leads to a loss of efficiency due to the high.:: harmonic current components. In addition, high voltage peaks across the motor occur in a square-wave motor current, which is undesirable for the electronic components.

Energizing the motor from an AC voltage source requires considerably fewer components. In addition, it is not necessary for the energizing voltage to be sinusoidal to realise a sinusoidal motor current. For that matter, due to the low-pass characteristic of the motor, higher harmonic current components are strongly suppressed.

When the motor is energized with an AC voltage having constant amplitude, there is a problem that the optimum frequency can no longer be determined on the basis of the variation of the power consumption as a function of the frequency. Fig. 2b shows by way of illustration the motor power consumption P plotted against frequency f for the case where the motor is energized with an AC voltage having constant amplitude.

As appears from Fig. 2b the position of the maximum of the power consumption P does not coincide with the optimum frequency as shown in Fig. 2a. In addition, determining the maximum is not a simple matter because in the low frequency zone of the graph (below 50 Hz) the power consumption P increases with a decreasing frequency, so that on the basis of the derived (dP/df) it is not possible to determine unambiguously a direction in which the frequency is to be changed from the maximum. All this means that many attractive search algorithms which utilize this derived value cannot be used.

Another manner of determining the optimum efficiency frequency is deteπiiining the frequency for which the P_le/P ratio is minimum or for which the P/P*_e ratio is maximum. Since P_je is, in essence, a function of I, a value for P_le may be derived from the amplitude of the motor current I.

Fig. 2c shows, plotted against frequency, the variation of the P/F(I) ratio of the motor power consumption P to the function F(T) for the case where F(I) = I. As appears from the drawing Figure, the frequency for which there is a maximum P/I ratio substantially coincides with the optimum frequency f_opt, so that the optimum frequency f_opt may be found in a simple manner on the basis of the P/I ratio.

Fig. 2d shows, plotted against frequency, the variation of the P/F(I) ratio for the case where F( ) = I².

As appears from Fig. 2d the frequency for which there is a maximum P/I² ratio substantially coincides with the optimum frequency f_opt. From the drawing Figs. 2c and 2d it will be obvious to the expert that the optimum frequency can be determined not only on the basis of P/I and P I², but that, in principle, the P/P^ ratio may be selected for the optimum frequency with k being situated between the limits 1 and m, where 1 is essentially equal to 1 and m is essentially equal to 2. It will be obvious that the various functions F(I) can be used. It is always essential that for the selected value function F(I) the maximum of the ratio in essence correspond to the optimum frequency f_opt and that the P/F(I) ratio be a monotonously increasing function for frequency values smaller than the frequency at which the maximum occurs and that the P/F(I) ratio be a monotonously decreasing function for the frequency values higher than the frequencies at which the maximum occurs.

Furthermore, it will no doubt be obvious when searching for the optimum frequency, the minimum of the F(I)/P ratio may equally well be determined instead of the maximum P/F(J).

An embodiment of a control device 30 according to the invention by which the optimum frequency is determined and, subsequent thereto, the frequency of the energizing motor is adjusted to the determined optimum frequency f_opt, is represented in Fig. 3. The control device 30 comprises a programmable control unit 40 of a customary type, comprising a central processing unit 31 (CPU), a read-only memory 32 (ROM), a random access memory 33 (RAM), analog-to-digital converters 34, 35 and 3 as well as an interface circuit 37. The memories 32 and 33 and the analog-to-digital converters 34, 35 and 36 and the interface circuit 37 are coupled in a customary manne to the central processing unit 31 by way of an address bus 38, a data bus 39 and a control bus 41. The control device further includes an energizing circuit 44 inserted between a DC voltage source 45 and the motor coils 16. The energizing circuit 44 is of a type energizing the coils 16 from the DC voltage source 45 with an AC voltage having an amplitude and frequency that can be adjusted via signal lines 42 and 43. For the purpose of this adjustment the signal lines 42 and 43 are connected to the interface circuit 37. Energizing circuit 44 further includes a measuring circuit for measuring the energizing current and generating a measuring signal I_j-. which is proportional to the measured energizing current. Energizing circuit 44 will be described in more detail hereinafter.

The output terminals of the coils 16 are connected to the inverting and to th non-inverting input of a differential amplifier 46 for measuring the voltage across the motor. The differential amplifier 46 then generates a signal U_m which is proportional to the voltage across the motor. The signal U_m is applied to a first input of a multiplier 47 and the signal I,., is applied to a second input of this multiplier 47, so that the output signal of multiplier 47 is representative of the instantaneous electric power consumption of the motor 3. The output signal of the multiplier 47 is applied to a low-pass filter 48, so that the output signal P-^ of the low-pass filter 48 is representative of the average power consumption of the motor 3. The signal V_m is applied to the input of an analog- to-digital converter 34. A measuring circuit 46 of a customary type derives the amplitude I of the energizing current L^ from the instantaneous value I_m of the energizing current, and applies a signal indicative of this amplitude I to the input of the analog-to-digital converter 46a. Further, a signal T_m is applied to the inverting input and a signal T_g to the non-inverting input of a differential amplifier 49. The signal T_m represents the temperature of the space to be cooled measured by means of the sensor (not shown). The signal T_g represents the desired temperature. The output signal T_in of the differential amplifier, which is representative of the difference between T_m and T_g, is applied to the input of an analog-to-digital converter 35. According to a customary control algorithm the control device 40 derives an adjusting value for the amplitude of the energizing voltage from the difference between T_m and T_g. A signal V_om indicative of the desired amplitude is applied to the energizing circuit 44 over the signal line 43.

The central processing unit 31 carries out a program stored in the read-only memory 32. While this program is being carried out, the optimum frequency of the energizing current is determined at regular intervals. Each time the optimum frequency has been determined, the frequency of the energizing circuit is adapted. A flow chart of such a program is shown in Figs. 4, 5 and 6.

Fig. 4 shows the flow chart of the main program which is carried out continuously by the central processing unit 31. During the step S10 a subroutine F_opt is called to determine the optimum frequency f_opt and the corresponding value of O^_t = P_tot I², and the frequency of the energizing voltage of the motor is adjusted to the optimum frequency f_opt thus determined. Subsequently, in the step Sll it is ascertained whether a flag FL has been set to the logic value "1". This flag indicates whether the desired value of the amplitude of the energizing voltage must be adapted. Should this be the case, the amplitude of the energizing voltage is adjusted once again over the signal line 43 during the execution of step S12. Furthermore, the flag FL is reset to the logic value "0" during the step S12. Once the step S12 has been executed, the programm proceeds with the step S10. However, if during the step Sll it has been found that the flag FL has not yet been set to the logic value "1", the step Sll is followed by the step S13. During the step S13 the digital value of P^ on the output of the analog-to-digital converter 34 and the digital value I on the output of the analog-to-digital converter 36 are read by the central processing unit 31. From this the ratio Q_m = P_m/I is determined.

The thus determined ratio is compared with the optimum value Q^. Then the absolute value ΔQ of the difference between Q^ and Q_m is determined. Subsequently, during the step S14 it is ascertained whether the absolute value ΔQ of the difference is smaller than a predetermined small value ε. If it is smaller, this means that the motor 3 is still energized with an AC voltage whose frequency is near the optimum value f_opt. In that case the program proceeds with the step S14. However, if the value ΔQ is larger than e, this means that as a result of varying operating parameters, such as for example the temperature of the refrigerant in the compressor 2, the frequency f_opt of the energizing voltage no longer corresponds to the actual optimum value. In that case the program proceeds with the step S10, after which the optimum frequency f_opt is determined again during the subroutine F_opt. Figs. 5A and 5B are flow charts of the subroutine F_opt. After the subroutine F_opt has been called, the frequency of the energizing voltage is first of all set to an initial value fO via the signal line 42 during the step S20. The central processing unit 31 then applies a signal V_f to the energizing circuit 44. Subsequently, during step S21 a waiting time is observed to allow the resonant system to adjust to the new frequency. Subsequently, during the step S22 the power consumption value P_m and the amplitude I of the motor current associated to the newly adjusted frequency are read in by means of the analog-to-digital converters 34 and 36. On the basis thereof, the ratio Q_m = P_in/I² is determined and stored as a variable P3 in the random access memory 33. In addition, the output signal Fr is incremented by a constant value ΔF, which corresponds to a frequency change Δf (see Fig. 2a). As a result of this, the frequency of the energizing current is adjusted to a new value equal to fO + Δf.

During the steps S23, S24, S25 and S26 the ratio Q_m for the frequencies fO + Δf and fO + 2Δf is determined and stored as the variable P2, PI respectively. During S27 an estimate of the derivative Afg (= dP/df) for the last frequency setting is derived from the values of PI, P2, P3 and ΔF in accordance with the following relationship:

_{Λ f}„ 3P1 - 4P2 + P3 ^M*⁼ ΔF

During the step S28 it is ascertained whether the derivative Afg is smaller than zero. If it is not smaller, this means that the maximum of P has not yet been reached. During the steps S29, S30 and S31 the frequency of the AC voltage is then again incremented by the value Δf and a corresponding value of the power consumption is determined and stored in PI. Moreover, the values Q_m of P2 and P3 which represent the values for the two preceding frequency settings are adjusted. Subsequently, during the step S27, the derivative for the newly set frequency is determined. The program loop formed by the steps S27, S28, S29, S30 and S31 is repeated until the derivative Afg thus determined is smaller than zero. This indicates that the optimum frequency is situated between the last and penultimate frequency setting. During the step S32 the value of Fr is then decremented by a value ΔF', which is substantially smaller than the value ΔF, so that the frequency of the energizing current is reduced by a small value Δf . After a waiting step S33 the ratio Q_m for the newly set frequency is determined in the step S34. Subsequently, during the step S35 it is ascertained whether the last ratio Q_m determined, represented by P2, is smaller than the ratio Q_m determined penultimately, represented by PI. If it is not smaller, the program proceeds with the step S35. The program loop comprising the steps S32, S33, S34 and S35 is repeated until the ratio Q_m determined last is smaller than the ratio Q_m determined penultimately. In that case, the power determined last is substantially equal to the maximum power value, and thus the associated frequency is substantially equal to the optimum frequency f_opt of the energizing current, so that the subroutine F_opt is terminated. In practice, it has been found that for reasons of stability it may be desirable not to set the frequency of the AC current to a value with a maximum ratio Q_m, but to a frequency slightly lower than this frequency. After the optimum frequency has been determined in accordance with said subroutine F_opt, the frequency of the AC voltage may then be set to the optimum frequency thus determined minus a predetermined value. It will be obvious to those skilled in the art that the program, described by way of example with reference to Fig. 5, for determining a maximum in the relationship between the ratio Q_m and the frequency setting may readily be replaced by any other known program for determining the maximum.

The main program Main is interrupted at constant intervals. During these interruptions a control program Control for determining the desired amplitude of the energizing voltage is carried out. A flow chart of an example of this control program is given in Fig. 6. During the first step SI of this control program Control the digitized value T^ on the outputs of the analog-to-digital converter 35 is read by the central processing unit 31. Subsequently, during the step S2 a new desired value for the amplitude of the energizing voltage is determined in accordance with a customary control algorithm of a kind which adjusts the desired value to a value at which the desired temperature T_g and the measured temperature in the space to be cooled are maintained substantially equal to each other. Further, during the step S2, the flag FL is set to the logic value "1" to indicate that the setting of the amplitude of the energizing voltage must be adjusted the next time that the steps Sll and S12 of the main program are carried out. After termination of the control program Control the main program Main is resumed.

Fig. 7 shows an embodiment of the energizing circuit 44. One terminal 72 of the series arrangement of the coils 16A and 16B of the motor 3 is coupled to a positive power supply terminal 70 and a negative power supply terminal 71 of the DC voltage source 44 via electronic switches 74 and 75 respectively, comprising for example power transistors of the FET type. The other terminal 73 of the series arrangement of the coils 16A and 16B is coupled to the positive power supply terminal 70 and the negative power supply terminal 71 via similar electronic switches 76 and 77 respectively. A low-impedance resistor 78 is arranged in the connection between the switch 75 and the power supply terminal 71 and a low-impedance resistor 79 is arranged in the connection between the switch 77 and the power supply terminal 71. The switches 74, ..., 77 are controlled by means of control circuits 80, ...,

83. The control circuits 80, ..., 83 are of a customary type which generate a control signal for opening the switch in response to a pulse of an "R" input and which generate a control signal for closing the switch in response to a pulse on an ^HS input. The "S" inputs of the control circuits 80 and 83 of the two diagonally opposed switches 74 and 77 are connected to the output of a pulse generator 85 by way of a delay circuit 84. The "S" inputs of the control circuits 81 and 82 of the other two diagonally opposed switches 75 and 76 are similarly connected to the output of a pulse generator 87 by way of a delay circuit 86. The "R" inputs of control circuits 81 and 82 are connected to the outputs of the respective pulse generators 87 and 85. The pulse generator 85 is of a type which generates a pulse in response to a logic 0-1 transition on its input, for example, a monostable multivibrator. Pulse generator 87 is of the type which generates a pulse in response to a logic 1-0 transition on its input. The inputs of the pulse generators 85 and 87 are connected to the output of a pulse-width modulator 88 of the type whose duty cycle can be set via the signal line 43. The frequency of the pulse-width modulator is set via the signal line 42. By means of the control system described herein, in response to a 0-1 transition of the output signal of the pulse-width modulator 88, the diagonally opposed switches are closed and a short time afterwards, determined by the delay switch 84, the two other diagonally opposed switches 80 and 83 are opened. The coils 16A and 16b are then connected to the power supply terminals 70 and 71 by means of the switches 80 and 83. Upon the next logic 1-0 transition of the output signal of the pulse-width modulator 88 the switches 72 and 77 are closed and a short time afterwards, this the switches 75 and 76 are opened. The coils 16A and 16B are then connected to the power supply terminals 70 and 71 by way of the switches 75 and 76. Since the motor operates as a filter, the current consumed by the motor is in essence determined by the fundamental harmonic of the applied AC voltage. This fundamental harmonic is directly determined by the duty cycle of the pulse-width modulated signal. Thus, the motor is as it were energized by a sinusoidal AC voltage whose amplitude is determined by the duty cycle. The energizing current flowing through the motor is determined by means of a differential amplifier 89, whose inverting input is connected to the junction point between the resistor 78 and the switch 75, and whose non-inverting input is connected to the junction point between the switch 77 and the resistor 79. The output signal I_m of the differential amplifier 89, which is proportional to the voltage difference between - said junction points, is thus proportional to the energizing current flowing through the coils 16A and 16B.

Claims

CLAIMS:

1. Control device for controlling the energizing of an oscillating motor for driving a compressor unit of the resonant-piston type, comprising a energizing circuit for supplying the motor with an AC voltage from an electric power source, and a detection circuit for deriving at least on the basis of the motor current a measuring signal which contains information about the efficiency with which the compressor unit is operated, characterized in that the energizing circuit is arranged for energizing the motor with an AC voltage having an essentially constant amplitude, in that the detection circuit comprises deriving means for deriving a signal as a measuring signal that has a predetermined relation to the ratio of the electrical power consumption of the motor to a function F(I) of the amplitude I of the current consumption of the motor, so that the maximum P/F(I) ratio occurs at an optimum frequency which essentially coincides with the frequency at which there is maximum motor efficiency, and in that the P/F(I) ratio monotonously increases for frequency values below said optimum frequency and the P/F(I) ratio monotonously decreases for frequency values above said optimum frequency.

2. Control device as claimed in Claim 1, characterized in that F(I) is essentially equal to I , where k is situated between the limits 1 and m, with 1 essentially being equal to 1 and m essentially being equal to 2.

3. Control device as claimed in Claim 1, characterized in that the device comprises measuring means for performing a measuring cycle in which the frequency of the AC voltage is varied and in which in response to the measured signal the frequency is determined for which the measured signal adopts an extreme value, and comprises adjusting means for adjusting the frequency to a value which is essentially equal to the determined frequency.

4. Compressor unit comprising a compressor for the resonant-piston type and an oscillating motor for driving the compressor, characterized in that the compressor unit comprises a control device as defined in Claim 1, 2 or 3.