EP1228573A1

EP1228573A1 - Noise management technique for switched voltage supplies

Info

Publication number: EP1228573A1
Application number: EP00993830A
Authority: EP
Inventors: Christopher Koszarsky; Joel James Walukas
Original assignee: Ericsson Inc
Current assignee: Ericsson Inc
Priority date: 2000-02-08
Filing date: 2000-12-27
Publication date: 2002-08-07
Also published as: WO2001059944A8; CN1354910A; WO2001059944A2; AU2001229132A1

Abstract

A noise management method and apparatus for electronic devices using switch voltage supplies includes producing a noise signal at one or more noise frequencies and moving the noise signal out of a frequency band of interest during processing of a signal in the frequency band of interest. In one embodiment, a multiplier and divider shift a stable reference frequency out of the receive signal band of a radiotelephone. The shift is programmable and is controlled in relation to the channel currently being received by the radiotelephone.

Description

NOISE MANAGEMENT TECHNIQUE FOR SWITCHED VOLTAGE SUPPLIES

CROSS REFERENCE TO RELATED APPLICATION

This application is related to U. S. patent application number 08/956,947, filed on October 23, 1998 by Ronald D. Boesch and Christopher Koszarsky, and published as publication number WO9921282 A on April 29, 1999.

BACKGROUND OF THE INVENTION

The present invention relates generally to electronic devices employing a switched power supply. More particularly, the present invention relates to a noise management technique for a switched voltage supply.

Recently, battery powered electronic devices such as cellular radiotelephones, have tended toward multiple operating voltages over recent years. Current generations of digital logic circuits use lower operating voltages in order to meet a need for cost reduction, semiconductor process improvements, physical size constraints and power consumption constraints. As a result, circuits which three to five years ago operated at an operating voltage of 5 volts or greater now must perform at voltages below 2.0 volts. Further, digital logic circuits currently being marketed require 1.8 volt operating voltages and future generations will be lowered to 1.5 volts.

Concurrent with this trend toward lower supply voltages is a reduction in the nominal voltage of the battery pack used in such electronic devices. Previous generations of cellular telephones used 5-cell battery packs with nominal voltages of 6.5 volts. In future generations, it is expected that 1 to 3 -cell battery packs will have a nominal battery voltage of 3.6 volts.

However, in many electronic devices, some functional blocks are not easily implemented with low supply voltages and require operating conditions which run counter to the trend toward lower battery and supply voltages. For example, in a cellular telephone, functional blocks such as the power amplifier and synthesizer continue to require a relatively large supply voltage, such as 5.0 volts. For a digital logic circuit, power conversion and supply may require dropping a 3.6 volt nominal battery voltage to the required 1.5 volts. Accordingly, such electronic devices incorporate a DC-to-DC converter whose sole purpose is to supply an elevated or reduced voltage to these functional blocks in a power efficient manner. Such a DC-to-DC converter may include capacitive charge pumps as well as inductive switching converters.

FIG. 1 is a block diagram of a prior art electronic device 100 using a switched power supply. The electronic device 100 includes a battery 102, a switch stimulus 104, a DC-to-DC converter 106 and operating circuits which, in the embodiment of FIG. 1, are transceiver circuits 108. In the illustrated embodiment, the electronic device 100 forms part of a radiotelephone such as a cellular or personal communication system (PCS) telephone.

The battery 102 is a depletable energy source which may be recharged and provides a substantially stable output voltage at a predetermined voltage value, such as 3.6 volts. The switch stimulus 104 provides a switching signal, such as a square wave at a predetermined frequency, labeled F_switch in FIG. 1. The DC-to-DC converter 106 receives the battery voltage from the battery 102 and a switching signal from the switch stimulus 104 and produces an output voltage, labeled N_output in FIG. 1. This output voltage may be stepped up in voltage value from the battery voltage or may be reduced in voltage value from the battery voltage. The output voltage is provided to the transceiver circuits 108 to serve as the operating supply voltage for the transceiver circuits 108. The DC-to-DC converter 106 may produce other output voltages in addition to V_ou(put.

The DC-to-DC converter implementation for a switched voltage supply is effective but has some disadvantages. The most notable disadvantage is the amount of noise generated by the switching operation of the DC-to-DC converter. The spectral noise generated by the DC- to-DC converter 106 is most apparent at the switching frequency and harmonics of the switching frequency. Thus, in the embodiment of FIG. 1, noise generated by the switching voltage supply follows Equation 1.

F ^A no •ise = m ^»«. F-. swi •tch - In Equation (1), m is any integer value and F_switch is the frequency of the switching stimulus 104 applied to the DC-to-DC converter 106. In the frequency domain, this noise spectrum is represented by FIG. 2. In FIG. 2, it can be seen that the noise spectrum includes a plurality of spikes at frequencies corresponding to the harmonic frequencies of the noise signal. A harmonic occurs for every integer value of m and each harmonic is separated by -^switch- This switching noise illustrated in FIG.2 will affect any electronic device incorporating a switched voltage supply. However, the effect of the switching noise is more apparent in an electronic device such as a cellular telephone, where highly sensitive receiver circuits must operate in the same environment as the switched voltage supply. For example, in one application, a switching voltage supply with a switching frequency F_switch of 540 kHz is used in a cellular telephone. The cellular telephone operates in a radiotelephone system having a receive band of 869.04 MHz to 893.97 MHz, with channel spacing of 30 kHz. The supply will generate noise components according to Equation (2).

F_noise = m.540kHz (2) However, for values of m between 1610 and 1655, the spectral components of the switching noise fall within the receive band and may interfere with the overall operation of the receiver. Similarly, where a cellular telephone operates in a transmit band different from the receive band, certain harmonics corresponding to predetermined values of m in Equation (2) may cause spurious response within the transmitter. Further, the locally generated switching stimulus is generally not controlled and is likely to vary over operating and environmental conditions of the cellular telephone. In some cases, it can vary as much as +/- 5%. This implies a further refinement to the definition for noise occurring in the electronic device 100 of FIG. 1 shown in Equation (3).

F_noise = w.{540-5%, 540 + 5%} (3) In previous electronic devices, the switching noise generated by the switching power supply has been suppressed by decoupling operating circuitry from the noise signal. For example, capacitors have been added to printed circuit boards to shunt the conducted noise signal to ground potential. For radio noise signals, metal shields have been added around the receiver and transmitter portions of the radiotelephone. These previous techniques have been successful only at reducing noise, not eliminating noise. Further, they add to the size and expense of an electronic device such as a radiotelephone incorporating the switching power supply and they are not guaranteed to produce a solution. The metal shields, capacitors and other components all require additional real estate on the printed circuit board and can add substantially to the cost of the completed product. Accordingly, there is a need for an improved method and apparatus for accommodating noise in electronic devices using switched voltage power supplies. BRIEF SUMMARY OF THE INVENTION

In accordance with the present invention, a method and apparatus for accommodating noise in an electronic device comprises managing, rather than suppressing, the noise in the device. In particular, a noise signal produced at one or more noise frequencies is moved out of a frequency band of interest. This is done by shifting the frequencies of the noise signal to one or more shifted frequencies outside the frequency band of interest. Thus, in a radiotelephone, the frequency of the generated noise signal is shifted above or below the specified receive band or transmit band for the radiotelephone. In addition, substantially all harmonics of the noise signal are shifted as well to ensure that the noise in the frequency band of interest is minimized.

The foregoing discussion of the preferred embodiments has been provided only by way of introduction. Nothing in this section should be taken as a limitation on the following claims, which define the scope of the invention.

BRΓEF DESCRIPTION OF SEVERAL VIEWS OF THE DRAWINGS FIG. 1 is a block diagram of a prior art electronic device using a switched power supply;

FIG. 2 illustrates a noise spectrum for a switched power supply; FIG. 3 is a block diagram of a radiotelephone operable in a radiotelephone system; FIG. 4 is a block diagram of a portion of the radiotelephone of FIG. 3; and FIG. 5 is a block diagram of a second embodiment of a portion of the radiotelephone of FIG. 3.

FIG. 3 is a block diagram of a radiotelephone system 300 including a mobile station or radiotelephone 302 and a base station 304. The base station 304 in the illustrated embodiment is one of a plurality of base stations in the system 300 configured for radio communication with mobile stations such as the radiotelephone 300 in a fixed geographic region surrounding the base station 304. Examples of such a system include cellular and personal communication systems as well as fixed and trunked radio systems. One particular example is North American Digital Cellular service according to J-STD-009, PCS IS-136 Based Mobile Station Minimum Performance 1900 MHz Standard and J-STD-010, PCS IS-136 Base Station Minimum Performance 1900 MHz Standard ("IS- 136"). The base station 304 communicates radio frequency (RF) signals with the radiotelephone 302 to communicate data representative of voice and other data with the radiotelephone 302.

The radiotelephone 302 includes an antenna 306, a transceiver circuit 308, a controller 310, a memory 312, a clock circuit 314, a power control circuit 316, and a user interface 318. The transceiver circuit 308 includes a receiver 320, a transmitter 322, a synthesizer 324 and an oscillator 326. For receiving radio frequency signals from the base station 304, such signals are detected at the antenna 306 and converted to digital data in the receiver 320 by demodulating a received signal. The digital data are conveyed to the controller 310 for further processing. For transmission of speech and other data from the radiotelephone 302 to the base station 304, data are conveyed from the controller 310 to the transmitter 322. The transmitter modulates a carrier signal provided by the synthesizer 324. The modulated carrier is conveyed to the antenna 306 for transmission of radio signals to the base station 304.

The synthesizer 324, as noted, provides a carrier signal to the receiver 320 and the transmitter 322 for modulation and demodulation of radio signals. By varying the frequency of this carrier signal, the receiver 320 and the transmitter 322 may be tuned to different channels as required in the system 300. The synthesizer 324 may be configured in any conventional manner, for example, using a phase locked loop (PLL). The oscillator 326 is preferably a temperature compensated crystal oscillator (TCXO), stable to within +/- 2.5 ppm over all environmental and operating conditions, including humidity and aging. Any other highly stable source of a switching signal may be substituted.

The controller 310 controls operation of the radiotelephone 302. The controller 310 is preferably implemented as a microcontroller, microprocessor or other digital logic device. As such, the controller 310 operates in conjunction with data and instructions stored in the memory 312. Timing of operations within the radiotelephone is controlled in response to clocking signals provided by the clock circuit 314. The power control circuit 316 preferably includes a battery or other depletable energy storage device, a DC-to-DC converter and other associated circuitry for forming a switched voltage supply. The power control circuit 316 provides operating voltages required for other circuits of the radiotelephone 302. For digital logic circuits, such as the controller 310, the operating voltages may be on the order of, for example, 1.5 or 1.8 volts. For other circuits, such as the transceiver circuits 308, the operating voltages may be 5 volts or greater. The user interface 318 provides user control of the radiotelephone 302. In one embodiment, the user interface 318 includes a keypad, a display, a microphone and a speaker.

FIG. 4 is a block diagram showing a portion 400 of the radiotelephone 300 of FIG. 3 in accordance with the present embodiment. In FIG. 4, the radiotelephone portion 400 includes transceiver circuits 308, a battery 402, a DC-to-DC converter 404 and a temperature compensated crystal oscillator (TCXO) 406. As noted above in connection with FIG. 3, the transceiver circuits 308 provide radio frequency communication with a remote radio such as a base station of a cellular or PCS radiotelephone system. The battery 402 provides operating power for the radiotelephone portion 400 and other circuits of the radiotelephone 300. The battery 402 provides a battery voltage at an output 410. TheDC-to-DC converter

404 provides at least one output voltage, labeled V_output, in response to the battery voltage and switching signals received from the TCXO 406. The TCXO 406 is preferably an oscillator which provides a switching signal at an output 412, the switching signal having a very stable frequency over a wide range of conditions. For example, the TCXO output signal is preferably stable within +/- 2.5 ppm over all environmental and operating conditions, including operating voltage, temperature, humidity and aging. In one embodiment, for operation of the radiotelephone 302 in accordance with the IS-136 digital radiotelephone standard, the TCXO 406 uses a reference frequency of 19.44 MHz +/- 2.5 ppm.

Thus, in one embodiment of the present invention, the switching signal provided by the oscillator 406 is used as both a switching stimulus for the DC-to-DC converter 404 and for providing the oscillator signal used by the transceiver circuits 308. The transceiver circuits 308 are responsive to the oscillator signal for modulating and demodulating RF signals. The transceiver 308 may include a synthesizer or other circuit for channel selection. The DC-to- DC converter forms a switched power supply responsive to the switching signal for providing operating power to circuits of the radiotelephone 300. The TCXO 406 forms a switch stimulus providing highly stable output signals including the oscillator signal and the switching signal. One or more frequency shifting circuits may be added in the path of the oscillator signal or the switching signal to adapt the frequencies of these signals to requirements of the circuits which receive them. The use of the TCXO advantageously improves the stability of the switching stimulus provided to the DC-to-DC converter to improve the predictability of the switching frequency (^_swi_tch) of the switching signal provided to the DC-to-DC converter 404. This reduces the bandwidth of the noise produced by the switching signal. Further, this allows the oscillator used in conventional switching supplies to be removed.

To improve the usability of the circuit of FIG. 4, it may be altered as illustrated in FIG. 5. FIG. 5 is a block diagram of a portion 500 of the radiotelephone 302 of FIG. 3.

Similar to the embodiment illustrated in FIG. 4, the radiotelephone portion 500 includes transceiver circuits 308, a battery 402, a DC-to-DC converter 404 and a TCXO 406. In addition, the radiotelephone portion 500 includes a frequency shifting circuit 502. The battery 402 forms a direct current (DC) power source for powering the radiotelephone portion 500 and other circuits of the radiotelephone 302. The TCXO 406 forms a switch stimulus, providing a switching signal at an output 412. This switching signal is provided to the frequency shifting circuit 502 as well as to the transceiver circuit 308. Not all connections are shown in the drawing so as to not unduly complicate the drawing figures. The frequency of the switching signal is F_switch. The DC-to-DC converter 404 produces an output DC voltage, labeled V_output. The transceiver circuits 308 form operating circuitry responsive to the output

DC voltage for processing a signal having a signal frequency within a frequency band of interest.

The frequency shifting circuit 502 includes a multiplier 504 and a divider 506 in the illustrated embodiment. The multiplier 504 operates to increase the frequency of the switching circuit received from the TCXO 406. A divider 506 operates to divide down or reduce the frequency of the signal provided to the DC-to-DC converter 404. Preferably, both the multiplier 504 and the divider 506 are programmable. That is, a multiply value (A) may be stored in the multiplier 504 to control the multiplication value applied by the multiplier 504. Similarly, a division value (B) may be stored in the divider 506 to control the frequency division value applied by the divider 506. For this purpose, the frequency shifting circuit includes control inputs 510 for receiving the clock signal, the programmable data values and a strobe signal. Preferably, these same signals are used for providing programmable data to the transceiver circuits 308. As will be described below, when data are stored in the transceiver circuits for controlling the channel to which the transceiver circuits 308 are tuned, data may likewise be stored in the frequency shifting circuit 502 for controlling the frequency shift applied by the multiplier 504 and the divider 506. The multiplier 504 and the divider 506 provide two advantages. The divider 506 reduces the switching stimulus frequency that is presented to the DC-to-DC converter 404. This allows the DC-to-DC converter to be traditional in design. As noted above, conventional DC-to-DC converters require a switching stimulus generally less than 1 MHz in frequency. Thus, the 19.44 MHz stable reference provided by the TCXO 406 in the illustrated embodiment can be used by the DC-to-DC converter 404, retaining the benefits of the stable signal provided thereby. Secondly, the combination of the multiplier 504 and the divider 506 provide controlling parameters that govern the frequency of the unwanted noise produced by the switching signal provided to the DC-to-DC converter 404. The switching noise is described by Equation 4.

F_noise = (A/B) . m . F_switch (4)

Thus, if there exists integer values of A and B that satisfy Equation 5,

F_noiSe ≠ (F_κ) for (5) then the noise generated by the switching supply can effectively be moved to a portion of the spectrum that will not effect sensitivity of the transceiver circuits 308.

Accordingly, values of A and B may be chosen to shift the one or more frequencies of the noise signal to one or more shifted frequencies outside a frequency band of interest. For the transceiver circuits 308 and the radiotelephone 302 of the illustrated embodiment, the frequency band of interest includes a receive band and a transmit band as specified by the radiotelephone system. For example, a cellular telephone in the United States has a receive band of 869.04 MHz to 893.97 MHz, with channel spacing of 30 kHz. After the noise signal has been moved outside the specific channel of interest, a signal in the specific channel of interest can be processed. In this example, the channel of interest is defined by the radiotelephone system. The transceiver circuits 308 may then receive or transmit radio signals relatively free of interference from the noise produced by the switching power supply of the radiotelephone 302. After processing the signal, the frequency shift may be changed or removed so that the noise signal is returned to the original noise frequencies. This shifting of the noise frequencies can be done on a channel-by-channel basis. New values of A and/or B can be loaded with each new channel assignment. Thus, the switching frequency provided to the DC-to-DC converter 404 varies with channel selection. In a further enhancement, values of A and B can be chosen so that a guardband G of frequencies around the desired receiver frequency is unoccupied by the noise generated by the switching power supply. In this case, the controlling equation becomes

F_noise ≠ (F^±G/2) for (6) A conservative receiver noise guardband for narrowband cellular systems such as IS-136 may be on the order of +/- 8 channels or 480 kHz. Other guardband values may be chosen and variable guardband values may be selected as well.

Any suitable value for A and B may be chosen for controlling the multiplier 504 and the divider 506, respectively. In most systems, A cannot equal 1 for all cases. The switching frequency F_{swit l} cannot be derived from integer division of the reference clock frequency,

19.44 MHz in the illustrated example, when the protected guardband (Frx ± G/2 contains a harmonic of the reference clock. In the cellular radio system mentioned above, the 45th harmonic of the 19.44 MHz signal falls within the receive band at 874.8 MHz (corresponding to channel 160). Any switching frequency derived from the 19.44 MHz signal through integer division would also produce harmonics at 874.8 MHz.

The objective is to keep noise out of the guard band around the receive frequency Frx (i.e., the band of interest). Then:

Fnoise = (MB), m .Fref ≠ (Frx ±G/2) for Assume that A=l (i.e., there is no multiplier between the reference clock and the switching stimulus). Furthermore, assume that the receive frequency is a harmonic (integer multiple) of the reference frequencey. Then:

Frx = n . Fref for Substituting into the previous equation,

(1/B) . m . Fref ≠ n . (Fref±G/2n) for and or m ≠ nB. (l+G/(2n.Fref)) for and (7)

However, since n and B are integers, then nB is an integer and this last equation is violated. However, if A is not equal to 1, then m ≠ ri /A . (l±G/(2n.Fref)) for and (8) The values of n, B and A can be chosen so as to guarantee nB/A is a non-integer. In the present example, n equals 45. Since this number is odd, setting B to an odd value guarantees an odd numerator. Setting A to 2 guarantees fractional values of «B/A. Setting A equal to 3 however will not work since, in this example, wB/A equals 45B/3=15B, which will be an 5 integer.

The value of guardband size, G, determines the minimum frequency of the switching stimulus. For a switching frequency F_switch < G, harmonic noise cannot be kept out of the guarded band.

Fnoise = m . Fswitch ≠ (F_re +G/2) (9)

ι° m ≠ F_K /F_switch ± G/(2 . F_switch))

If F_switch is less than G, then +/- G/(2F_switch) is greater than 0.5. This guarantees that the righthand side of the above equation contains an integer value.

In choosing multiplier and divider values as well as frequency values, larger values of the switching frequency F_switch and A enable more consistent switching operation (i.e., less is variation in the frequency of the switch stimulus). As shown in Equation (10), the switching frequency F_switch (which is also the fundamental frequency of the noise) is equal to the reference frequency Fref times the multiplication value A and divided by the division value B.

Fswitch = (A/B).Fref

20 Fnoise = m. Fswitch = (A/B).m.Fref ≠ (Frx ± G/2) for

m ≠ (B.Frx/(A.Fref)) ± B.G/(2A.Fref) = Frx/Fswitch ± G/(2Fswitch) for

(10) This final equation allows several rules and observations. First, the guardband G can

25 be maximized if Frx/Fswitch is selected to be halfway between integer values. That is, if

2. (Frx/Fswitch) is an odd integer, G/(2F switch) can be as large as 0.5 before violating the inequality in equation 10. For example, consider Frx Fswitch = 1, 124.1, the guardband teπn G/(2A.Fswitch) < 0.1. If Frx/Fswitch = 1,124.5, the guardband term G/(2.Fswitch) < 0.5 allowing G to be increased by a factor of 5 for the same system parameters. However, if

30 Frx/Fswitch = 1,124.9, the guardband term G/(2Fswitch) < 0.1.

Second, increasing Fswitch allows the guardband G to increase by the same ratio. For example, let Frx = 875.7 MHz and Fswitch = 500 kHz . Then Frx/Fswitch = 1751.4 and

Claims

G/(2.500 kHz) < 0.4. Solving the inequality, G < 400 kHz. Increasing, Fswitch to 1 MHz changes the inequality to G/(2. 1 MHz) < 0.4. Solving this new inequality yields G < 800 kHz.Third, the size of the guardband G determines the minimum frequency of the switching supply stimulus. If G > Fswitch, then G/(2F switch) > 0.5. This guarantees the band ± G/(2Fswitch) contains an integer regardless of the values of Frx and Fswitch violating the inequality in equation 10.Fourth, larger values of A provide finer resolution in the adjustment of the switching frequency. For example:IfA = 2, Fswitch = Fref, 2.Fref/3, Fref/2, ... for B = [2,3,4, ]If A = 4, Fswitch = Fref, 4.Fref/5, 2.Fref/3, 4.Fref/7, Fref/2 , ... for B = [ 4, 5, 6, 7, 8... ]From the foregoing, it can be seen that the present embodiments provide an improved method and apparatus for accommodating noise in switched voltage supplies used in electronic devices. Rather than suppressing the noise, the noise is actually managed by shifting the frequency of the noise outside of a frequency band of interest. While the frequency shift occurs, further signal processing within the band of interest can occur without the interference introduced by the noise. In this manner, the large and expensive and marginally effective components formerly used to try to suppress noise may be eliminated from the design of the electronic device. They are replaced instead with a relatively simple multiplier and divider for shifting the noise frequencies.While a particular embodiment of the present invention has been shown and described, modifications may be made. For example, in place of separate multiplier M divider circuits, the frequency shifting circuit may use any suitable digital logic circuit or analog circuit such as a mixer for adjusting the frequency of the noise signal. It is therefore intended in the appended claims to cover all such changes and modifications which follow in the true spirit and scope of the invention. CLAIMS

1. A noise management method comprising: producing a noise signal at one or more noise frequencies; and moving the noise signal out of a frequency band of interest.

2. The noise management method of claim 1 wherein moving the noise signal comprises: shifting the one or more frequencies of the noise signal to one or more shifted frequencies outside the frequency band of interest.

3. The noise management method of claim 1 further comprising: after moving the noise signal, processing a signal in the frequency band of interest; and then, returning the noise signal to the one or more noise frequencies.

4. The noise management method of claim 1 further comprising: generating a switching signal; in response to the switching signal and a reference voltage, generating an output voltage; in response to the output voltage, after moving the noise signal, processing a signal in the frequency band of interest; and

5. A noise management method comprising: processing a signal within a frequency band of interest; and while processing the signal, shifting a noise signal to one or more frequencies outside the frequency band of interest.

6. The noise management method of claim 5 further comprising: producing the noise signal ancillary to producing a switching signal and producing an output voltage responsive to the switching signal and a reference voltage.

7. An electronic device comprising: a direct current (DC) power source; a switch stimulus configured to produce a switching signal at one or more noise frequencies; a frequency shifting circuit coupled to the switch stimulus and configured to the one or more noise frequencies of the switching signal outside a frequency band of interest; a DC/DC converter coupled to the DC power source and the frequency shifting circuit and configured to produce an output DC voltage; and operating circuitry responsive to the output DC voltage for processing a signal having a signal frequency within the frequency band of interest.

8. The electronic device of claim 7 wherein the switch stimulus comprises an oscillator circuit.

9. The electronic device of claim 8 wherein the switch stimulus comprises temperature compensated crystal oscillator.

10. The electronic device of claim 7 wherein the frequency shifting circuit comprises a frequency multiplier.

11. The electronic device of claim 10 wherein the frequency shifting circuit further comprises a frequency divider.

12. The electronic device of claim 11 wherein at least one of the frequency multiplier and the frequency divider is programmable for selecting the one or more selected frequencies.

13. The electronic device of claim 7 further comprising a control circuit for selecting the one or more selected frequencies in response to the signal frequency.

14. A radiotelephone comprising: a battery; a crystal oscillator; a frequency shifting circuit coupled to the crystal oscillator and producing a switching signal having a selected frequency outside a frequency band of interest; a voltage converter coupled to the battery and the frequency shifting circuit to produce an operating voltage; and a transceiver circuit operative in response to the operating voltage to communicate radio frequency (RF) signals within the frequency band of interest.

15. The radiotelephone of claim 14 further comprising: a synthesizer coupled to the transceiver and producing a local oscillator signal for channel selection in the transceiver circuit, the frequency shifting circuit coupled to the synthesizer and producing the local oscillator signal in response to the switching signal.

16. A radio comprising: transceiver circuits responsive to the oscillator signal for modulating and demodulating RF signals; a switched power supply responsive to a switching signal to provide operating power to circuits of the radiotelephone; and a switch stimulus providing one or more highly stable output signals including the oscillator signal and the switching signal.

17. The radio of claim 16 wherein the switch stimulus comprises a temperature compensated crystal oscillator.

18. The radio of claim 16 further comprising: a frequency shifting circuit for adjusting frequency of at least one of the oscillator signal and the switching signal.