Texas Instruments Incorporated

Signal Conditioning: Op Amps

An audio circuit collection, Part 2

By Bruce Carter
Advanced Linear Products, Op Amp Applications

Introduction
This is the second in a series of articles on single-supply audio circuits. The reader is encouraged to review Part 1*, in the November 2000 issue of Analog Applications Journal, which concentrated on low-pass and high-pass filters. Part 2 concentrates on audio-notch-filter applications and curve-fitting filters.

60-Hz hum filter

Audio notch filters

There are audio applications where a single frequency is undesirable and needs to be rejected. The frequency to be rejected is not part of the original audio and is annoying to the listener. Frequencies on either side of the rejected frequency, however, contain useful audio content. If they are rejected as well, there will be an equally annoying hole in the audio. Notch filters are used to reject very narrow frequency bands with minimal attenuation on either side of the notch. Notch filters, compared to low-pass and high-pass filters, are hard to implement. Components that would cause only a slight ripple or washout in a low-pass or high-pass filter often have a dramatic effect on the depth of the notch. A slight mistuning of a low-pass or high-pass filter is inaudible, but mistuning a notch filter may cause it to miss the interfering frequency altogether. Following are some hints about how to implement a notch with a reasonable degree of confidence.
* The original Part 1 release of this series had op amp polarity symbols reversed in several figures. Please download the corrected version at: www-s.ti.com/sc/techlit/slyt155

One of the most common problems with audio is the presence of a 60-Hz hum. Since 60 Hz is the frequency used for ac-power distribution, it is one of the most prevalent interference sources. Usually, a 60-Hz hum is the result of poor grounding practices. It is better to attack the problem at the source than to filter the audio. Nevertheless, there may be situations where the grounding of a particular system may not be accessible. In that case, an add-on filter may be appropriate. The 60-Hz hum filter shown in Figure 1 is based on twin-T configuration. This topology is very effective but can be temperamental. The circuit response is very dependent on the actual values of R1, R2, R3, C1, C2, and C3. All of these tuning components should be 1%, but that may not be enough. They should all be taken from the same lot, because parts manufactured at the same time tend to have the same characteristics. If components are matched properly, performance can be very good. Mismatched components will seriously degrade the response. In the twin-T configuration, R3 is half the value of R1 and R2. The best way of making the resistance value for R3 is to use two of the resistors used for R1 and R2 in parallel. Similarly, taking two capacitors of the same value as C1 and C2 in parallel forms C3. This increases the component count of the circuit by two (one additional resistor and capacitor) but greatly increases the Continued on next page

Figure 1. Configuration of a 60-Hz notch filter

VCC VCC R6 100 k C4 4.7 F IN + R7 100 k C3 24 nF R1 221 k R2 221 k

U1A

C5 4.7 F + OUT

U1B R3 110.5 k

+
C2 12 nF

R4 4.7 k

C1 12 nF

R5 100 k

41 Analog Applications Journal February 2001 Analog and Mixed-Signal Products

Signal Conditioning: Op Amps

Texas Instruments Incorporated

Continued from previous page

Figure 2. Response of 60-Hz notch filter

matching, because the designer can take easy steps to insure that the parts are 0 from the same batch (off the same reel, out of the same box, etc.). Note in Figure 1 that no half-supply 20 reference is required. R6 and R7 generate half supply after the input capacitor. The theoretical response of this filter is shown in Figure 2. The ideal values produce a 40 notch as shown. If, however, the RC combination is 1% off, the notch will be shifted left or right by 0.6 Hz, and rejection of the 60-Hz frequency could be less than 60 20 dB. 55 60 65 70 50 A way around that problem is to put a Frequency (Hz) small potentiometer in series with R3, which should be reduced one or two standard E-96 values. Figure 3 shows circuit Figure 3. Varying notch depth response with R3 varied from 107 k to 113 k. Varying R3 over even this small range 0 produces a tremendous variation in the depth of the notch as well as the Q of the circuit. In every case, though, more than 20 20 dB is achievable, and most of the time as much as 30 or 40 dB. If a deeper null is needed, R1 and R2 need to be adjustable as well. They can be adjusted in a dual 40 potentiometer to reduce the number of 70 55 60 65 50 potentiometers in the circuit to 2. If the Frequency (Hz) capacitors are matched (which is possible if they are from a single batch), 60-dB rejection is possible. The circuit can be modified to reject Figure 4. Response of a 50-Hz notch filter hum from the 50-Hz line frequency of ac power in Europe and other parts of the world, as follows: 0 Change R1 and R2 to 42.4 k. Change R3 to 21.2 k (two 42.4-k resistors in parallel). 20 Change C1 and C2 to 75 nF. Change C3 to 150 nF (two 75-nF capacitors in parallel). 40 The previous comments regarding the 60-Hz notch filter apply to the 50-Hz version as well. Theoretical response of the 50-Hz notch filter is shown in Figure 4. 60 60 55 40 45 50 Standard E-24 capacitor values and stanFrequency (Hz) dard E-96 resistor values do not produce a combination that is as close to perfect as in the 60-Hz case, so trimming may be even more necessary for a 50-Hz notch filter. The audio modulation from adjacent channels is usually not a problem on strong local stations, but the carrier is. It Medium-wave whistle filter shows up in the audio as a 10-kHz tone. This tone can be In North America, medium-wave (AM) stations are quite loudespecially at night, even on local stations. For separated by 10-kHz channels from 540 to 1700 kHz. those who can hear it, the pitch is extremely annoying. AM frequency response is unlimited compared to FM, Making the problem even worse, there are channels above which is severely rolled off above 15 kHz. Unfortunately, and below those used by most stations, and they are not there will be interference from the adjacent channels, exactly at 10 kHz. The FCC allows a tolerance of 20 Hz especially at night. from the assigned frequency, and that will make the two
42 Analog and Mixed-Signal Products February 2001 Analog Applications Journal
VOUT (dB) VOUT (dB) VOUT (dB)

Texas Instruments Incorporated

Signal Conditioning: Op Amps

Figure 5. 10-kHz notch filter

VCC VCC R6 100 k C4 4.7 F IN + R7 100 k C3 720 pF R1 44.2 k R2 44.2 k

U1A

C5 4.7 F +

OUT

U1B R3 22.1 k

+
C2 360 pF

R4 4.7 k

C1 360 pF

R5 100 k

adjacent channels modulate each other and create beat-frequencies, adding to the annoying aspect of the tone. To eliminate the 10-kHz tone, a notch filter is needed that eliminates a narrow band around 10 kHz, while leaving other frequencies untouched. Many years ago, high-priced AM receivers used a high-Q LC filter, but tuning was so critical that it was of limited use. The op amp approach shown in Figure 5 is extremely stable and never requires additional adjustments once the initial center frequency is set. The twin-T notch filter topology is used again, due to its ability to provide large attenuations with only two op amps. All of the comments about the 60-Hz notch filter apply to the 10-kHz notch filter. Even with simple tuning, however, the improvement should be dramatic. The response of this filter is shown in Figure 6. In Europe and much of the rest of the world, the medium-wave transmission uses channels from 531 to 1611 kHz, separated by 9 kHz. This will cause a 9-kHz (instead of 10-kHz) tone in the received audio. To reject the 9-kHz tone resulting from 9-kHz channel spacing: Change R1 and R2 to 45.3 k Change R3 to two 45.3-k resistors in parallel (22.65 k). Change C1 and C2 to 390 pF. Change C3 to two 390-pF capacitors in parallel (780 pF). The response of a well-tuned 9-kHz notch filter is shown in Figure 7. Much of the world relies on short-wave radio stations for news and entertainment. Short-wave radio is transmitted on Continued on next page

Figure 6. Response of 10-kHz notch filter

0

VOUT (dB)

20

40

60 9.0

9.5

10.0 Frequency (kHz)

10.5

11.0

Figure 7. Response of 9-kHz notch filter

0

VOUT (dB)

20

40

60 8.0

8.5

9.0 Frequency (kHz)

9.5

10.0

43 Analog Applications Journal February 2001 Analog and Mixed-Signal Products

Signal Conditioning: Op Amps

Texas Instruments Incorporated

Continued from previous page several bands, with stations separated by only 5 kHz. To reject the 5-kHz tone resulting from 5-kHz channel spacing: Change R1 and R2 to 42.4 k Change R3 to two 42.4-k resistors in parallel (21.2 k). Change C1 and C2 to 750 pF. Change C3 to two 750-pF capacitors in parallel (1500 pF). The response of a well-tuned 5-kHz notch filter is shown in Figure 8. This notch filter topology can be retuned to reject almost any audio frequency that poses a problem. Areas of the world served by both 10-kHz-spaced and 9-kHzspaced medium-wave stations may experience objectionable tones at any frequency from 1 kHz to 10 kHz and above.

Figure 8. Response of 5-kHz notch filter

0

VOUT (dB)

20

40

60 4.0

4.5

5.0 Frequency (kHz)

5.5

6.0

Curve-fitting filters
Analog designers are often asked to design low-pass and high-pass filter stages for maximum rejection of frequencies that are out of band. This is not always the case, however. Sometimes the designer is asked to design a circuit that will conform to a specified frequency response curve. This can be a challenging task, particularly if all the designer knows is that a single-pole filter rolls off 20 dB per decade; and a double-pole filter, 40 dB per decade. How does the designer implement a different roll-off? It is not possible to get more out of a filter than it is designed to produce. A single pole will give no more than 20 dB per decadeand cannot be increased or decreased. More roll-off demands a double-pole filter with 40 dB per decade. Obtaining different roll-off characteristics is done by allowing filters at closely spaced frequencies to overlap. One popular curve-fitting application is the RIAA equalization (see Figure 9), which compensates for equalization applied to vinyl record albums during manufacture. Designers of many newer pieces of audio gear have omitted the RIAA equalization circuit completely, assuming that the majority of users will not desire the function. In spite of the enormous popularity of audio CDs, there are still dedicated audiophiles who have a large library of record albumstitles that are not available on CDs or are out of print. RIAA has the following response: 17 dB from 20 to 50 Hz, 0 dB from 500 to 2120 Hz, and 13.7 dB at 10 kHz.
44 Analog and Mixed-Signal Products

Figure 9. The RIAA equalization curve

20

10 VOUT (dB)

10

20

10

100

1k Frequency (Hz)

10 k

100 k

Figure 10. RIAA equalization circuit

VCC C1 3.3 F IN + 8 2 4 R1 47 k R2 510 R3 360 k + C2 33 F C3 9.1 nF VCC /2 C4 1.5 nF R4 51 k U1A

February 2001

TLC2272
1 OUT

Analog Applications Journal

Texas Instruments Incorporated

Signal Conditioning: Op Amps

Figure 11. RIAA equalizer resonse

x1 y1 501.1872 2.5989 AC Analysis
1 2

x2 y2

2.2132k 1.4036

20

10 VOUT (dB)

10

20 1 10 100 1k Frequency (Hz) 10k 100k

RIAA equalization curves often include another breakpoint at 10 Hz to limit low-frequency rumble effects that could resonate with the turntables tone arm. The standard input impedance in the circuits shown in Figure 10 is 47 k. This impedance makes a convenient place to inject dc offset into single-supply circuits, so it is isolated from the phonograph cartridge by an input capacitance. The phonograph cartridge output is assumed to be 12 mV. Application circuits were evaluated from many sources in print and on the Web. Many of these did not work at all, did not easily translate to single-supply operation, or deviated markedly from the RIAA specification. The circuit topology presented in Figure 10 was one of the most common, appearing in several sources. This circuit was tweaked manually to produce the closest possible conformance to the RIAA curve. A small additional gain resistor was sometimes added between the junction of R3 and C3 and the inverting input. It did not seem to be necessary, and this implementation contains the smallest number of passive components. There is even a Web page that contains a Java-based calculator dedicated to this topology (see www.vwlowen.demon.co.uk/java/riaa.htm). The implementation of the circuit in Figure 10 yields the curve shown in Figure 11. Several things are troublesome with this topology: No matter how much the circuit is optimized, the section from 500 Hz to 2120 Hz is not simulated well. The first-order breakpoints that are possible with the single op amp create only a slight ripple on the characteristic curve in Figure 11. These breakpoints require a secondorder filter. This is very near the region where human hearing is the most sensitive and errors will be the most audible. The musical content immediately below 1 kHz will be too loud, and that immediately above 1 kHz will be too soft. Aesthetically, this will make the sound muddy, lacking brilliance and tonal clarity. C2 is a large capacitance value that happens to be in the highest-gain network in the circuit. Power-on transients will cause large, unexpected voltage swingspossibly overloading the input to the next stage. They could also create loud, possibly destructive transients in

loudspeakers. Further, the difficulty of getting precision values of electrolytic capacitors will lead to wide variations in responseboth of the amplitude and the lowfrequency roll-off breakpoint. Fine-tuning this circuit is difficult; virtually all components interact. The procedure for tuning is: 1. Set the low-frequency gain (LFG) with R2 and R3: LFG = R3 = 16.97 dB 100 R2 R4 = 0 dB 100 R2 1 = 9.46 Hz 2 R2 C2 1 = 48.6 Hz 2 R3 C3

2. Set the mid-frequency gain (MFG) with R4: MFG =

3. Set the low-frequency roll-off (LFR) with C2: LFR =

4. Set the low-frequency breakpoint (LFB) with C3: LFB =

5. The mid-frequency breakpoint (MFB) is already determined by the values of R4 and C3: MFB = 1 = 342 Hz 2 R4 C3 1 = 2080 Hz 2 R4 C4

6. Set the high-frequency breakpoint (HFB) with C4: HFB =

These steps must be followed in order. The initial selection of R2 determines the other components. It is unfortunate that there is no control over the mid-frequency breakpoint, which probably accounts for the error in the response of the curve (Figure 11). The mid-frequency breakpoint is constrained to 342 Hz when it should be 500 Hz. Continued on next page
45

Analog Applications Journal

February 2001

Analog and Mixed-Signal Products

Signal Conditioning: Op Amps

Texas Instruments Incorporated

Continued from previous page Fine-tuning can be improved by splitting the implementation into two op amps (see Figure 12): Set the low-frequency roll-off with R1 and C1: 1 LFR = = 10.3 Hz 2 R1 C1 Set the low-frequency gain with R3 and R2: LFG = R3 = 16.9 dB R2

Figure 12. Dual op amp implementation of equalizer

VCC C1 330 nF IN U1A 8 2 R1 47 k VCC /2 4 U1B

6 C3 2.2 nF VCC /2

R3 330 k

Set the low-frequency breakpoint with C2: 1 LFB = = 48.2 Hz 2 R3 C2 Set the high-frequency breakpoint with R4 and C3: HFB = 1 = 723 Hz 2 R4 C3
20

R2 47 k VCC /2

C2 10 nF

Figure 13. Response of split op amp circuit

2 R1 47 k
VCC/2

6 C4 330 pF
VCC/2

R3 330 k

R2 47 k
VCC/2

C2 10 nF

46 Analog and Mixed-Signal Products February 2001 Analog Applications Journal

IN

The response of this circuit is shown in Figure 13. The circuit will be the starting point for simulation of the RIAA curve. The response from 500 to 2120 Hz should be flat at 0 dB. This first-order circuit is 1.8 dB too high at 500 Hz, and 2.4 dB too low at 2120 Hz. Selecting the HFB at 723 is a trick that shifts the response at 1 kHz down to 0 dB. This is a fairly drastic change, though. The first step in improving the RIAA characteristic is to change the 2120-Hz portion to second-order. A unity-gain Sallen-Key stage is selected, as shown in Figure 14. R4, R5, C3, and C4 control the 2120-Hz breakpoint. The response of the circuit changes to that shown in Figure 15. The 2120-Hz response has improved from 2.4-dB to 0.8-dB deviation from the curve. Unfortunately, there is less interaction with the 50-Hz low-pass filter; and the 500-Hz response is now 2 dB, instead of 1.8 dB, above ideal. Clearly, another second-order filter is required. Accomplishing this requires a change in firstorder-stage topology and an increase in complexity to four op amps, as shown in Figure 16.

x1 y1

495.4830 1.8427

AC Analysis
1 2

x2 y2

10 VOUT (dB)

10

20 1 10 100 1k Frequency (Hz) 10k 100k

Figure 14. Unity-gain Sallen-Key equalizer

VCC C1 330 nF U1A 8 3

C3 470 pF R4 110 k R5 68 k U1B 5

TLC2272
1

TLC2272
1

R4 100 k

TLC2272
5 7 OUT

2.1380k 2.3838

TLC2272
7 OUT

Texas Instruments Incorporated

Signal Conditioning: Op Amps

12

VCC /2

References
1. Audio Circuits Using the NE5532/34, Philips Semiconductor (October 1984). 2. Audio Radio Handbook, National Semiconductor (1980). 3. Op Amp Circuit Collection, National Semiconductor AN-031 (1978).

C4 22 nF C5 22 nF C3 22 nF

R7 91 k

U1C

10 R6 22 k VCC /2 VCC /2

Figure 17. Response of filter with four op amps

x1 y1 501.1872 815.1290m AC Analysis
1 2

Related Web sites

www-s.ti.com/sc/techlit/slyt023 amplifier.ti.com www.ti.com/sc/device/tlc2274 www.vwlowen.demon.co.uk/java/ riaa.htm
20

10 VOUT (dB)

10

20 1 10 100 1k Frequency (Hz) 10 k 100 k

Analog Applications Journal

February 2001

TLC2274
8

R9 220 k

x2 y2

2.1380k 1.2776

Analog and Mixed-Signal Products

This circuit topology is very flexible. Figure 15. Response of the Sallen-Key stage Most of the RIAA breakpoints are independently adjustable, as follows: x2 x1 495.4830 AC Analysis R1 and C1 set the LFR as before. y2 y1 1.9747 20 U1A, R2, and R3 control the overall gain of the circuit. R4 and R5 control the LFG. 10 R5 and C2 control the 50-Hz LFB. C3, C4, C5, R6, R7, and U1C form a 0 500-Hz high-pass filter that reverses the effect of the 50-Hz low-pass filter and flattens the response through 10 1 kHz until the 2120-Hz low-pass filter begins to affect the response. 20 R8, R9, R10, C6, C7, and U1D form the 1 10 100 1k 2120-Hz low-pass filter as before, but Frequency (Hz) the input resistor has been split into a summing resistor. The overall response Figure 16. Filter with four op amps of the filter is shown in Figure 17. The 500-Hz response is VCC above the ideal curve by U1A 0.8 dB, and the 2120-Hz C2 C1 4 TLC2274 27 nF 330 nF 3 response is below the ideal IN 1 C6 2 curve by 1.3 dB. This cir470 pF R5 100 k cuit is about the best that 11 R1 47 k can be created without U1B R4 R10 R8 many more op amps and TLC2274 11 k 68 k 220 k 6 VCC /2 7 complex design techR3 5 R2 100 k niques. It should produce 110 k C7 330 PF VCC /2 very aesthetically pleasing VCC /2 sound reproduction.
1 2

2.1380k 823.0408m

VOUT (dB)

10 k

100 k

U1D 13

TLC2274
14 OUT

47

IMPORTANT NOTICE Texas Instruments Incorporated and its subsidiaries (TI) reserve the right to make corrections, modifications, enhancements, improvements, and other changes to its products and services at any time and to discontinue any product or service without notice. Customers should obtain the latest relevant information before placing orders and should verify that such information is current and complete. All products are sold subject to TI's terms and conditions of sale supplied at the time of order acknowledgment. TI warrants performance of its hardware products to the specifications applicable at the time of sale in accordance with TI's standard warranty. Testing and other quality control techniques are used to the extent TI deems necessary to support this warranty. Except where mandated by government requirements, testing of all parameters of each product is not necessarily performed. TI assumes no liability for applications assistance or customer product design. Customers are responsible for their products and applications using TI components. To minimize the risks associated with customer products and applications, customers should provide adequate design and operating safeguards. TI does not warrant or represent that any license, either express or implied, is granted under any TI patent right, copyright, mask work right, or other TI intellectual property right relating to any combination, machine, or process in which TI products or services are used. Information published by TI regarding third-party products or services does not constitute a license from TI to use such products or services or a warranty or endorsement thereof. Use of such information may require a license from a third party under the patents or other intellectual property of the third party, or a license from TI under the patents or other intellectual property of TI. Reproduction of information in TI data books or data sheets is permissible only if reproduction is without alteration and is accompanied by all associated warranties, conditions, limitations, and notices. Reproduction of this information with alteration is an unfair and deceptive business practice. TI is not responsible or liable for such altered documentation. Resale of TI products or services with statements different from or beyond the parameters stated by TI for that product or service voids all express and any implied warranties for the associated TI product or service and is an unfair and deceptive business practice. TI is not responsible or liable for any such statements. Following are URLs where you can obtain information on other Texas Instruments products and application solutions: Products Amplifiers Data Converters DSP Interface Logic Power Mgmt Microcontrollers Applications Audio Automotive Broadband Digital control Military Optical Networking Security Telephony Video & Imaging Wireless amplifier.ti.com dataconverter.ti.com dsp.ti.com interface.ti.com logic.ti.com power.ti.com microcontroller.ti.com www.ti.com/audio www.ti.com/automotive www.ti.com/broadband www.ti.com/digitalcontrol www.ti.com/military www.ti.com/opticalnetwork www.ti.com/security www.ti.com/telephony www.ti.com/video www.ti.com/wireless

TI Worldwide Technical Support

Internet
TI Semiconductor Product Information Center Home Page support.ti.com TI Semiconductor KnowledgeBase Home Page support.ti.com/sc/knowledgebase

Product Information Centers

Americas Phone Internet/Email +1(972) 644-5580 Fax support.ti.com/sc/pic/americas.htm +1(972) 927-6377

Europe, Middle East, and Africa Phone Belgium (English) +32 (0) 27 45 54 32 Netherlands (English) +31 (0) 546 87 95 45 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 Germany +49 (0) 8161 80 33 11 Sweden (English) +46 (0) 8587 555 22 Israel (English) 1800 949 0107 United Kingdom +44 (0) 1604 66 33 99 Italy 800 79 11 37 Fax +(49) (0) 8161 80 2045 Internet support.ti.com/sc/pic/euro.htm Japan Fax International Internet/Email International Domestic Asia Phone International Domestic Australia China Hong Kong Indonesia Korea Malaysia Fax Internet

+81-3-3344-5317

Domestic

0120-81-0036

support.ti.com/sc/pic/japan.htm www.tij.co.jp/pic

+886-2-23786800 Toll-Free Number 1-800-999-084 800-820-8682 800-96-5941 001-803-8861-1006 080-551-2804 1-800-80-3973 886-2-2378-6808 support.ti.com/sc/pic/asia.htm

New Zealand Philippines Singapore Taiwan Thailand Email

Toll-Free Number 0800-446-934 1-800-765-7404 800-886-1028 0800-006800 001-800-886-0010 [email protected] [email protected]

C011905

Safe Harbor Statement: This publication may contain forwardlooking statements that involve a number of risks and uncertainties. These forward-looking statements are intended to qualify for the safe harbor from liability established by the Private Securities Litigation Reform Act of 1995. These forwardlooking statements generally can be identified by phrases such as TI or its management believes, expects, anticipates, foresees, forecasts, estimates or other words or phrases of similar import. Similarly, such statements herein that describe the company's products, business strategy, outlook, objectives, plans, intentions or goals also are forward-looking statements. All such forward-looking statements are subject to certain risks and uncertainties that could cause actual results to differ materially from those in forward-looking statements. Please refer to TI's most recent Form 10-K for more information on the risks and uncertainties that could materially affect future results of operations. We disclaim any intention or obligation to update any forward-looking statements as a result of developments occurring after the date of this publication. Trademarks: All trademarks are the property of their respective owners. Mailing Address: Texas Instruments Post Office Box 655303 Dallas, Texas 75265

2005 Texas Instruments Incorporated

SLYT145