Audio Texas Instruments Incorporated Audio measurements, Part 2

By Richard Palmer Application Specialist, Audio Amplifiers Introduction real-world layout constraints. The measurements of a par- This article is a continuation of “ amplifier ticular audio circuit may vary from the typical specifications. measurements,” which first appeared in the July 2001 A large variance is usually indicative of a PCB layout or issue of Analog Applications Journal (see Reference 1) measurement system issue. Reference 2 provides more and which contains guidelines for measuring the following details about the measurements in this article. three parameters: Supply rejection • power supply rejection ratio (PSRR), Two types of supply rejection specifications exist: power- • supply ripple voltage rejection ratio (kSVR), and supply rejection ratio (PSRR) and supply ripple voltage • efficiency. rejection ratio (kSVR). The only difference between them All measurements were made by using TI Plug-n-Play is that PSRR is a dc specification and kSVR is an ac specifi- APA evaluation modules (EVMs) for the TPA2001D1 filter- cation that measures the ability of the APA to reject ac- free class-D and the TPA731 class-AB mono devices. Note ripple voltage on the power supply bus. All power supply that the graphs in the data sheets reflect typical specifi- decoupling capacitors are removed from class-AB circuits, cations and were measured on test boards specifically and class-D circuits have a small 0.1-µF decoupling capacitor, designed to allow accuracy and ease of measurement. Board C, placed close to the APA power pins to provide a reverse space, layout, and components were not constrained by path for recovery switching currents. It is recommended size/cost requirements. The measurements in this article, that the designer use equal decoupling capacitance values however, were taken by using circuits on EVMs that reflect when comparing devices from different manufacturers to

Figure 1. PSRR and kSVR measurement circuit

Audio Power Amplifier AP Analyzer In VOUT CIN (DMM2)** + IN+ Diff Input OUT+ RC Filter for Filter-Free BTL RL + Channel B CIN – Class-D Output IN– OUT– Measurements –

VS GND + Channel A – C* AP Generator Out C ** Inputs SVR + – + Balanced, ac-coupled VS ZIN = 100 kW /185 pF Channel A RSVR** (DMM1)** Set Load Reference = RL – Internal Filter£ 10 Hz – 80 kHz Reading Meter = Crosstalk Outputs Unbalanced–Float Data 1 = Analyzer Crosstalk Ch A ON V+ GND Source 1 = Generator Frequency Ch B Track Ch A Z= 20W OUT Regulated Set Load Reference = RL Power Supply Sweep 20 kHz – 20 Hz

* The 0.1-µF capacitor, C, is required for class-D operation.

** The PSRR measurement uses the DMMs only since it is a dc value. kSVR measurements use either the analyzer, , or DMMs since they are ac values. RSVR and CSVR are used for kSVR measurements only.

26

Analog and Mixed-Signal Products www.ti.com/sc/analogapps 1Q 2002 Analog Applications Journal Texas Instruments Incorporated Audio Amplifiers get a valid comparison of the performance, since a higher Figure 2. kSVR filter circuit capacitance equates to a better kSVR. PSRR is the ratio of the change in the output voltage, VOUT(dc), to a change in the power supply voltage, VS, CSVR expressed in dB as shown in Equation 1. RGEN RSVR

 ∆VOUT() dc  PSRR = 20log  (1)  ∆VS  V For example, the output voltage of an GEN RAPA RS that has a PSRR equal to –70 dB would change by 31.6 µV if the supply voltage changed by 0.1 V. kSVR is the ratio of the output ripple voltage, VOUT(ac), to a change in the supply ripple voltage, VS, expressed in dB as shown in Equation 2.

 VOUT() ac  kSVR = 20log  (2) output. Here the analyzer, an Audio Precision System–II,  VS  is configured for a crosstalk measurement1, 2; it sweeps This parameter is normally listed as a typical value in the the ac voltage at constant over the audio band, data sheet tables at a specified frequency and temperature measuring and presenting a graph of the data points in dB. of 1 kHz and 25°C, respectively. A graph is provided in the The kSVR filter circuit is shown in Figure 2. The dc power supply output impedance, RS, is very low (milli- data sheet of the typical values of kSVR over the audio bandwidth because it is a frequency-dependent term. ); and the impedance of the APA to ground, RAPA, as seen by the power supply or signal generator, is very high The PSRR and kSVR measurement circuit is shown in Figure 1. All inputs are ac-coupled to ground. The PSRR (hundreds of ohms). The value of RSVR is added to the measurement requires only the two DMMs. The power circuit to increase the equivalent impedance of the power supply and is chosen to be approximately equal to the ac supply voltage, VS, is initially set, then read from the meter on the power supply. When the power supply meter signal source generator output impedance, RGEN. A voltage does not have the desired resolution, DMM1 is used to divider is formed between RSVR and RGEN that provides a reasonable amplitude ac signal at the APA power pin. measure VS. DMM2 then measures VOUT across the load. CSVR is added to ac-couple the signal generator to the VS is then stepped up or down by a specific amount, and APA. The filter cutoff frequency, fC, should be set 3 dB the corresponding value of VOUT is measured. The differ- ences of the two measurements are then substituted into below the lowest frequency of the audio band, fMIN, which Equation 1, and the PSRR is calculated for that specific in this case is 20 Hz. Equation 3 provides the value for fC, change in supply voltage. PSRR is specified as a typical which is ~14 Hz. value that is valid for a given supply voltage range at 25°C. fMIN fC = (3) The kSVR measurement requires the signal generator, 2 analyzer, a DMM, and the kSVR filter components RSVR 1, 2 and CSVR. The RC measurement filter is used when the The equivalent resistance is then calculated with Equation 4, analyzer cannot accurately process the square wave output where RAPA is the supply voltage divided by the quiescent of the filter-free class-D APAs. DMM1 is used to measure current of the device (VS/IQ). VS at the APA power pin. The generator injects a small sine wave signal onto the power bus, and the audio analyzer RREQ=+ GEN R APA!() R SVR +≈+ RR S GEN R SVR (4) measures this ac voltage at the APA power pin and at the Continued on next page

Figure 3. kSVR of the TPA2001D1 and the TPA731

0

– 20 VS = 3.3 V AV = 12 dB Class-AB W R=L 8 AV = 6 dB Class-D – 40 C=B 1Fm BTL

SVR – 60

k(dB) Class D – 80 Class AB –100 20 200 2000 20,000 Frequency (Hz)

27

Analog Applications Journal 1Q 2002 www.ti.com/sc/analogapps Analog and Mixed-Signal Products Audio Amplifiers Texas Instruments Incorporated

Continued from previous page Those devices with BYPASS pins will have improved kSVR as the capacitance on the pin is increased. Those The value for CSVR is then calculated with Equation 5. operated SE have lower kSVR, particularly at the extreme low and high ranges of the audio band. This is primarily 1 CSVR = (5) due to the resonance of the output ac coupling capacitor. 2π××fRCEQ The kSVR graphs of the TPA2001D1 and the TPA731 are shown in Figure 3. Both of these devices are differential The capacitor will most likely be electrolytic due to the input and BTL output. The TPA731 was measured with value required. It will have some reactance that will vary the inputs floating. Newer devices are typically measured with frequency as shown by Equation 6. with the inputs ac-grounded. 1 X = (6) Efficiency measurements CSVR 2π××fCCSVR Efficiency is the measure of the amount of power that is delivered to a load for a given input power provided by the At 20 Hz the impedance will be quite high—approximately supply. A class-AB APA acts like a variable resistor net- the value of RGEN and RSVR; and at 20 kHz the value will work between the power supply and the load, with the be in the milliohms. output operating in the linear region. They The actual values for the measurement circuit were dissipate quite a bit of power because of this mode of RGEN = 20 Ω, RS ≈ 0, RAPA = 5 V/6 mA = 833 Ω, CSVR = operation and are therefore inefficient. The output stage 330 µF, RSVR = 20 Ω, and fC = 12 Hz. This yields a capaci- in class-D amplifiers acts as a switch that has a small tive reactance of 24 Ω at 20 Hz, and 24 mΩ at 20 kHz. The resistance when operated in the saturation region, which value of the ac signal must be adjusted at low frequencies provides a much higher efficiency. so that the desired voltage is applied to the APA power A circuit for measuring the efficiency of a class-AB or pin. The value of the dc voltage from the power supply class-D system is shown in Figure 4. The simplest setup must also be adjusted, since IQ will create a small voltage occurs when the power supply voltage and current meters drop across RSVR. are accurate and have the resolution required. When the

Figure 4. Efficiency circuit, BTL

Oscilloscope or AP Generator Out Audio Power Amplifier V2 Analyzer CIN (DMM2) + IN+ Diff Input OUT+ RC Filter for + + + Filter-Free Channel A BTL R2 V Channel 1 (A) CIN – Class-D OUT Output – – IN– OUT– Measurements –

Outputs Balanced ZL* – + VS GND Z=OUT 40W V3 Set Load Reference = RL (DMM3) Set Frequency of Signal + + + – R1** VS Channel 2 (B) V1 – – (DMM1)**

Inputs Balanced, dc-coupled V+ GND ZIN = 100 kW /185 pF Set Load Reference = RL Regulated Set dBrB Ref to Ch A Power Supply

* Load ZL is a speaker for class-D APAs and is a purely resistive load for class-AB APAs. ** DMM1 and Channel 2 of the AP/oscilloscope (or a third DMM) are used to measure the average power supply current and voltage when power supply meters are not accurate. Resistor R1 can be removed when power supply meters are used.

28

Analog and Mixed-Signal Products www.ti.com/sc/analogapps 1Q 2002 Analog Applications Journal Texas Instruments Incorporated Audio Amplifiers

supply current meter is not sufficient, R1 is placed in the R2, provides a means of measuring the current that is used circuit. It should be a small value (0.1 Ω) and able to handle in the efficiency calculation (IOUT ≈ V2/R2). the power dissipated. The average voltage, V1, across R1 Equation 7 provides the efficiency of the class-AB APA, provides the average supply current (IS = V1/R1) that is and Equation 8 provides the efficiency of the class-D APA. used to calculate the average power provided by the supply. The input power of both equations is just the average The true rms DMMs and the audio analyzer provide an voltage applied to the power pin of the APA, multiplied by rms value of both the voltage and the current, which, the average value of the power supply current. Average when multiplied together, provide the average power. value is used for the power supply measurements, since When used, the power supply meters provide the average the voltage and current have dc and ac components and value of the supply voltage and current. The oscilloscope are nonsinusoidal. The output power is also an average can measure the average or rms values of the power sup- that results from the multiplication of two rms terms. ply and output voltage. Some even have The RC measurement filter is used for making filter-free current probes that can be used to measure the current class-D output measurements when the analyzer or DMM 1, 2 through a wire, in which case R1 is not needed. cannot accurately process the switching waveform. The The load measurement is different for class-AB and filter resistance must be large enough to minimize current class-D APAs. Two elements are shown: one is the actual flow through the filter; while the capacitance must be load, ZL, and the other is resistor R2. The Class-AB load is sized to achieve the desired cutoff frequency, which a non-inductive power resistor, ZL = RL. R2 is not required should be just above the audio band. If the filter resistor is for class-AB efficiency measurements since the load is not large enough, the filter current must be accounted for purely resistive. in the efficiency equation. The recommended values of The filter-free class-D load is a speaker, so the impedance RFILT and CFILT are 1 kΩ and 5.6 nF, respectively. This will vary with frequency. A speaker is used because it has provides a filter cutoff frequency of ~28 kHz. This filter is the inductance that helps provide the high class-D effi- not used with traditional class-D devices since they ciency. A purely resistive load is not a true indicator of the already have LC filters in the output circuit. operating environment of the filter-free class-D, and will The efficiency was measured with a 3.3-V supply; the not provide accurate efficiency numbers. The power must results are shown in Figure 5. Figure 5 provides the be calculated independently of the speaker impedance measured efficiency from the power supply meter and a since it varies with frequency. The small power resistor, Fluke 87III DMM measuring the voltage across the load. Continued on next page

 V 2   OUT() RMS   RL  POUT   ηClass− AB == (7) PS VISAV()× SAV ()

V  RRMS2 () VOUT() RMS ××  P VI×  R2  OUT OUT() RMS OUT () RMS   (8) ηClass− D == = PS VISAV()× SAV () VISAV()× SAV ()

Figure 5. Efficiency of the TPA731 class-AB APA and the TPA2001D1 class-D APA

Efficiency Measurements 100 80 Class-D 60

40 VS = 3.3 V Class-AB Z=L 8W

Efficiency (%) 20 BTL 0 0 100 200 300 400 500

POUT (mW)

29

Analog Applications Journal 1Q 2002 www.ti.com/sc/analogapps Analog and Mixed-Signal Products Audio Amplifiers Texas Instruments Incorporated

Continued from previous page power and is in units of dB. For example, the of a sine wave is 3 dB. Sine waves are used in the charac- The DMM class-AB data was in close agreement with terization of APA performance but do not give a clear idea measurements made with the AP-II Analyzer or a TDS 754 of what the performance will be with music. The CF of oscilloscope. The class-D DMM and AP data were similar, music may vary between 6 dB and 24 dB and directly but the oscilloscope measured 5–10% higher due to its impacts the amount of heat dissipated in the device—the averaging, which introduced a somewhat large margin of higher the crest factor, the lower the heat dissipated and error, particularly at high power output. The DMM reading the higher the ambient operating temperature may be. is more reliable, since it filters out the high-frequency har- The PD data discussed in the previous two paragraphs can monics of the switching waveform to provide a more stable, be used to determine the CF of the device. low-frequency value. Tables of measurement values for Equation 10 may be used to calculate CF. Since a sine both amplifiers are available in Reference 2. wave was used for the measurements, the crest factor is 3 dB, and the average output power, POUT(AV), is known. Power dissipated versus power to the load The peak output power, POUT(PK), is calculated by manip- The efficiency measurements provide the information ulating Equation 10 into Equation 11, where POUT(PK) required to calculate the amount of power dissipated, PD, and POUT(AV) are expressed in and CF is expressed in the amplifier. PD provides some insight into the supply in dB. currents and power that will be required. PD is calculated  POUT() PK  by using Equation 9 and the measured values of supply CF() dB = 10 log  (10) and output power from the efficiency measurements. It is    POUT() AV  assumed that the power dissipated in the RC filter, used for the filter-free class-D APA measurements, is negligible; P therefore it is not included in the calculation. When it is P = OUT() PK (11) significant, it must be included as part of P . OUT() AV  CF  OUT   10 10  PPPDSOUT=− (9) Figure 6 shows graphs of dissipated power versus the For example, the maximum peak output power is 500 mW output power calculated with Equation 9. The data was for the TPA731. This is calculated by using 250 mW as measured up to the maximum output power, which occurs POUT(AV) and a CF of 3 dB for the output sinusoid. The just prior to , and can easily be discerned from a peak will not change throughout the calculations, as it is graph of THD+N versus output power.2 The designer can the maximum output power possible and is independent of choose the percent (level of clipping) that is the output waveform. The CF is then increased in 3-dB acceptable for a system and test the device through that steps up to 18 dB, and the corresponding POUT(AV) is cal- power level. culated for each step. The PD in the device is measured for each value of POUT(AV) with the efficiency measure- Crest factor and output power ment circuit. The crest factor (CF) is the ratio of the peak output to the The efficiency data and CF calculations can help the average output. It is typically graphed in terms of output designer approximate the power that must be provided by

Figure 6. Power dissipated vs. output power

Power Dissipated vs. Power Out

300

250 VS = 3.3 V Class-AB Z= 8W 200 L BTL 150

D P(mW) 100 Class-D 50 0 0 50 100 150 200 250 300 350 400 450 POUT (mW)

30

Analog and Mixed-Signal Products www.ti.com/sc/analogapps 1Q 2002 Analog Applications Journal Texas Instruments Incorporated Audio Amplifiers the power supply. Figure 7 shows the graph of Figure 7. Supply and output power vs. CF PS and POUT versus crest factor. The graph allows easy comparison of the devices, and it is clear that the class-D APA provides more POUT with less power from the supply than the class- Supply and Output Power vs. CF AB APA. The difference between PS and POUT is the dissipated power, P . 525 D PO (Class-AB) Measurement pitfalls 420 PS (Class-AB) P (Class-D) The following is a list of common pitfalls, or mis- 315 O takes, that can be encountered and will produce PS (Class-D) measurement errors. 210 • The signal generator should be balanced for

Output Power (mW) 105 differential inputs, unbalanced for SE inputs. • Wires should be twisted together to minimize 0 magnetic coupling into loops and ground loops. 3691215 18 • Use the RC measurement filter for the filter- Crest Factor (dB) free class-D outputs when using an analyzer or oscilloscope. The capacitor should be connected to either the power supply or the APA power ground. • The leads of filter components should be kept short. References • The DMM must be “true rms” to get accurate For more information related to this article, you can down- measurements. load an Acrobat Reader file at www-s.ti.com/sc/techlit/ • The wires from the power supply and from the APA to litnumber and replace “litnumber” with the TI Lit. # for the load should be at least 18 AWG for 2-W APAs.2 the materials listed below. Supply rejection ratio measurements Document Title TI Lit. # • A 0.1-µF decoupling capacitor is required for class-D 1. Richard Palmer, “Audio power amplifier operation. All other capacitors should be removed. All measurements,” Analog Applications decoupling capacitors should be removed for class-AB Journal (July 2001), pp. 40-46...... slyt135 measurements. 2. “Guidelines for Measuring Audio Power Amplifier Performance,” Application Report . . .sloa068 Efficiency measurement • The filter-free class-D RC measurement filter should Related Web sites have a high resistance for RFILT, with a value of 1 kΩ www.ti.com/sc/docs/products/analog/tpa731.html recommended in conjunction with a 5.6-nF capacitor. www.ti.com/sc/docs/products/analog/tpa2001d1.html The current through the filter must be considered when RFILT is small. • Check to make sure that the DMMs used at the load are set to measure ac volts and to measure dc volts at the power supply.

31

Analog Applications Journal 1Q 2002 www.ti.com/sc/analogapps Analog and Mixed-Signal Products IMPORTANT NOTICE TI Worldwide Technical Support Texas Instruments Incorporated and its subsidiaries (TI) reserve the right to make corrections, modifications, enhancements, Internet improvements, and other changes to its products and services at TI Semiconductor Product Information Center Home Page any time and to discontinue any product or service without notice. support.ti.com Customers should obtain the latest relevant information before TI Semiconductor KnowledgeBase Home Page placing orders and should verify that such information is current support.ti.com/sc/knowledgebase and complete. All products are sold subject to TI's terms and conditions of sale supplied at the time of order acknowledgment. Product Information Centers Americas TI warrants performance of its hardware products to the Phone +1(972) 644-5580 Fax +1(972) 927-6377 specifications applicable at the time of sale in accordance with TI's Internet/Email support.ti.com/sc/pic/americas.htm standard warranty. Testing and other quality control techniques are Europe, Middle East, and Africa used to the extent TI deems necessary to support this warranty. Phone Except where mandated by government requirements, testing of Belgium (English) +32 (0) 27 45 54 32 Netherlands (English) +31 (0) 546 87 95 45 all parameters of each product is not necessarily performed. Finland (English) +358 (0) 9 25173948 Russia +7 (0) 95 7850415 France +33 (0) 1 30 70 11 64 Spain +34 902 35 40 28 TI assumes no liability for applications assistance or customer Germany +49 (0) 8161 80 33 11 Sweden (English) +46 (0) 8587 555 22 product design. Customers are responsible for their products and Israel (English) 1800 949 0107 United Kingdom +44 (0) 1604 66 33 99 applications using TI components. To minimize the risks Italy 800 79 11 37 associated with customer products and applications, customers Fax +(49) (0) 8161 80 2045 should provide adequate design and operating safeguards. Internet support.ti.com/sc/pic/euro.htm Japan TI does not warrant or represent that any license, either express or Fax implied, is granted under any TI patent right, copyright, mask work International +81-3-3344-5317 Domestic 0120-81-0036 right, or other TI intellectual property right relating to any Internet/Email combination, machine, or process in which TI products or services International support.ti.com/sc/pic/japan.htm are used. Information published by TI regarding third-party Domestic www.tij.co.jp/pic products or services does not constitute a license from TI to use Asia such products or services or a warranty or endorsement thereof. Phone Use of such information may require a license from a third party International +886-2-23786800 under the patents or other intellectual property of the third party, or a Domestic Toll-Free Number Toll-Free Number license from TI under the patents or other intellectual property of TI. Australia 1-800-999-084 New Zealand 0800-446-934 China 800-820-8682 Philippines 1-800-765-7404 Reproduction of information in TI data books or data sheets is Hong Kong 800-96-5941 Singapore 800-886-1028 permissible only if reproduction is without alteration and is Indonesia 001-803-8861-1006 Taiwan 0800-006800 accompanied by all associated warranties, conditions, limitations, Korea 080-551-2804 Thailand 001-800-886-0010 and notices. Reproduction of this information with alteration is an Malaysia 1-800-80-3973 Fax 886-2-2378-6808 Email [email protected] unfair and deceptive business practice. TI is not responsible or Internet support.ti.com/sc/pic/asia.htm [email protected] liable for such altered documentation. C011905 Resale of TI products or services with statements different from or Safe Harbor Statement: This publication may contain forward- beyond the parameters stated by TI for that product or service looking statements that involve a number of risks and voids all express and any implied warranties for the associated TI uncertainties. These “forward-looking statements” are intended product or service and is an unfair and deceptive business to qualify for the safe harbor from liability established by the practice. TI is not responsible or liable for any such statements. Private Securities Litigation Reform Act of 1995. These forward- Following are URLs where you can obtain information on other looking statements generally can be identified by phrases such Texas Instruments products and application solutions: as TI or its management “believes,” “expects,” “anticipates,” “foresees,” “forecasts,” “estimates” or other words or phrases Products of similar import. Similarly, such statements herein that describe Amplifiers amplifier.ti.com the company's products, business strategy, outlook, objectives, Data Converters dataconverter.ti.com plans, intentions or goals also are forward-looking statements. DSP dsp.ti.com All such forward-looking statements are subject to certain risks and uncertainties that could cause actual results to differ Interface interface.ti.com materially from those in forward-looking statements. Please Logic logic.ti.com refer to TI's most recent Form 10-K for more information on the Power Mgmt power.ti.com risks and uncertainties that could materially affect future results Microcontrollers microcontroller.ti.com of operations. We disclaim any intention or obligation to update any forward-looking statements as a result of developments Applications occurring after the date of this publication. Audio www.ti.com/audio Trademarks: All trademarks are the property of their Automotive www.ti.com/automotive respective owners. Broadband www.ti.com/broadband Digital control www.ti.com/digitalcontrol Mailing Address: Texas Instruments Military www.ti.com/military Post Office Box 655303 Optical Networking www.ti.com/opticalnetwork Dallas, Texas 75265 Security www.ti.com/security Telephony www.ti.com/telephony © 2005 Texas Instruments Incorporated Video & Imaging www.ti.com/video Wireless www.ti.com/wireless SLYT128