# Setting the bias for the power amp section of an audio amplifier?

I've been working on the internal amplifier of SW-12 active subwoofer. I started suspecting its power amp section may not be working properly. (I get musically coherent sound out of the attached woofer, but the volume seems to be lower than before I started fooling around with the amp.).

Since I have changed R88, R89, and R90 (with the resistors of the same resistance rating but with 7W power rating), I think I'd better check if the bias for the power amp section is correct, at the least.

How do I do that? Can somebody guide me through this?

What mode do I set my multimeter to and measure where?

I have numbered various points in the schematic of the power amp section:

Oh, and what should the value/s be?

Do the LEVEL, PHASE, and LOW PASS settings matter when I do this?

How about signal going into the amp's LINE in? Is no signal fine?

= = = = = = = = = = = = = = = = = = = = = = = = =

The service manual for SW-12 is here:

http://www.audiolabga.com/pdf/SW12-15%20I.pdf

As far as I can tell, it doesn't mention anything about setting the bias for the power amp section of the amplifier.

= = = = = = = = = = = = = = = = = = = = = = = = =

Update (3/6/17):

Per jonk's guidance, I have checked the following:

+/- 81 volt rails: about +/- 91 volts;

+/- 15 volt rails: about +/- 15 volts;

voltage across two zener diodes, D5 & D6: +33.6V and -33.6V, respectively

Per Tonny Elliot's guidance, I had checked/adjusted R50 to make the DC offset at the speaker output 0V +/- 10mV or so (within spec) before starting the thread here.

= = = = = = = = = = = = = = = = = = = = = = = = =

R66, R67, R139 (brown resistor), and C55. I thought some of you might find this configuration interesting.

= = = = = = = = = = = = = = = = = = = = = = = = =

2 pictures showing speaker output V, A, and V*A added (3/11/17)

The load attached is a space heater with 17.2 ohms (not exactly 16 ohms as required for the alignment procedure described in Service Manual), but good enough for taking these readings, isn't it?

Adjusting R84 doesn't seem to affect these numbers significantly, by the way.

= = = = = = = = = = = = = = = = = = = = = = = = =

After looking at the numbers (V, A, and V*A of the speaker output) and the accompanying charts above last night, I started thinking maybe the amplifier is OK. I don't like abrupt changes I see around LEVEL 6~7, but that how the circuit for this amplifier may have been designed. When this subwoofer was working fine, LEVEL 7 (1 o'colock) was about the highest I used. If this is the case, I don't know why I feel the sound lacks energy today. (I can't rule out the possibility of the woofer being damaged.)

Can somebody tell whether the numbers above look at least reasonably OK or absolutely not?

According to the specs for my subwoofer, SW=12, the maximum rated output is 150W into 8 ohm. (The input sensitivity for max rated output is listed as 60mVrms for LINE IN.) If my reasoning is correct, the voltage required to deliver 150W into 8 ohms is 34.6V. So, the amplifier seems to be amplifying the signal enough. or does it?

Several hours later. . .

I just remembered this comment of mine that I posted on March 9:

jonk: "Also, consider examining R34, too, which sets the gain." I made a test CD with 30 Hz. sine signal. Per instructions in the service manual, I set the output of the receiver to 60mV and measured the output of the subwoofer's amplifier with a 17.2 ohm dummy load connected. The voltage was about 20V. I increased the gain to make the output voltage about 33V (per spec). I found the sound quality got noticeably worse.

I hadn't touched R34 until then, so a part or parts of amplification may be missing. As to R30, the output voltage is noticeably lower than the spec's value, but I can't increase it because R30 had been rotated CW (to increase) fully.

• $R_{84}$ is part of a $V_{BE}$ multiplier. So that's the resistor to adjust to set "bias." You don't mention that fact, but I hope you already realize it. Do you know why it's there? (It doesn't have Early Effect compensation. But that's another issue.) – jonk Mar 2 '17 at 19:12
• I guessed R84 is the one to adjust, but I don't know why it's there. I'm a novice in electronics. So, you adjust R84, but measure what where? – zeron Mar 2 '17 at 19:26
• I'll write a little. Your question is too big to consider writing a full manual for. Before I do, though, I assume you've already checked the voltage rails?? What do they look like? – jonk Mar 2 '17 at 19:28
• If it's not clear, you should definitely check out the $\pm 15\:\textrm{V}$ voltage rails (against ground) and also the $\pm 81\:\textrm{V}$ rails. There are some $34\:\textrm{V}$ zeners in there on page 2 of the schematic. You might also want to check the voltages on those, as well. For starters. I don't know how old your unit is. But replacing all the electrolytics comes to mind, as well. The do get old. So how old is this unit? – jonk Mar 2 '17 at 19:35
• Over 20 years old. No, I haven't measured +/-15V rails or +/-81V rails yet since I'm new to this and don't feel comfortable taking live measurements. For 81V rails, do I set my multimeter to DC mode and attach its negative probe to L and the positive probe to D or E? The heat sink of the amp seems to be grounded. Should I use the heat sink instead of L? Thanks for your help. – zeron Mar 2 '17 at 19:48

the power amp section is really the output stage (OPS). The input and vas stages are in the driver circuit.

the ops is quite innovative. to check bias of it, measure voltage across j/m -> should be around 4v -> 6 pn junctions.

or you can check the voltage across emitter resistors on Q16/17/19/20/22/23/25/26. they should be roughly the same.

but I doubt they are the issues. your issues sound like gain setting / protection. check r66/67/167 and c55, the input circuitry. having a signal generator + scope would be helpful.

• dannyf: "but I doubt they are the issues. your issues sound like gain setting / protection. check r66/67/167 and c55, the input circuitry." When I was touching up some solder joints, I inadvertently created a short around R88 and R90. I've been thinking that running the amp with the short could have damaged some components. I'll check the components you mention, but I can't seem to locate R167. Is the number correct? Thanks. – zeron Mar 7 '17 at 20:55
• dannyf: "check r66/67/167 and c55, the input circuitry." The following measurements were taken with the components in circuit: R66 = ~20.8 ohms. R67 = ~0.98Kohms. R139 = ~ 20.7 ohms. I don't have the capability to measure capacitance at the moment. The schematic indicates that R66 is supposed to be 39.2K ohm and R139 39K ohm (and their color codes agree with these numbers). Do you think it's worth the trouble to take their measurements off the circuit? – zeron Mar 8 '17 at 5:26
• R139/R66 = 20ohm sounds way too long to be correct. if that's measured correctly, it can definitely contribute to low gain. but how did you measure it? – dannyf Mar 8 '17 at 11:13
• dannyf: "but how did you measure it?" All in the circuit. I can understand R139's reading being affected by C55 or others, but I expected R66 to be not affected by others. I measured their resistance by touching their two leads with my multimeter's probes. Interestingly, about a month ago, R66 read about 28K ohms, according to my note. Again in circuit. I'll see if I can get a picture of this area. – zeron Mar 8 '17 at 16:37
• i just took readings of R66 and R139. They both read about 25 ohms now. (About 10 hours ago, I turned on the amplifier for testing.) – zeron Mar 8 '17 at 17:06

Because I got busy with other things I had to take care of, I stopped on restoring the subwoofer around March, 2017. (The original problem was making a crackling noise after warming up.)

Because the area of the circuit board where R87, R88, R89, and R90 were was getting hot and one of the resistors actually burned out, I replaced three of them with resistors of the same ohm (750 ohm) but with higher wattage (7W instead of 5W). (One of the four resistors was hard to remove, so I left the original.) While this replacement of the three resistors may have helped to reduce the the overheating problem, the original noise remained. This was all I did before stopping on the subwoofer last year.

I resumed working on the subwoofer about two weeks ago. I replaced on original remaining resistors of the four and changed C55 this time, but the noise remained.

I decided to go back to square one two nights ago and started looking for bad solder joints on the circuit board. I noticed that if I wiggle the heat sink of Q9, the noise gets affected. I tested further with cooling the heat sink with a Q-tip moistened with alcohol and confirmed that the temperature on the heat sink (= the temperature of Q9) was the cause of the noise. I touched up the solder joints of pins of Q9 and pins of several other components near it. On the first try, I tried to do this minimally on the pins of Q9 fearing damaging it with too much heat. Upon testing, I found the noise was still there. Deciding that I'll just replace the transistor if damaged, I did resoldering the pins of Q9 thoroughly with more solder. After this, the noise disappeared completely. I listened to music for several hours last night. The subwoofer sounds more robust than before. (After replacing the four resistors and one capacitor, I recalibrated R34, R30, and VDC offset. Maybe this helped. Also my impression that the sound level was lower than before may have been purely subjective. After putting the woofer back into its cabinet and sealing it as it should be, it sounds robust again.)

I thank everyone who posted on my threads about the subwoofer. I started knowing almost nothing about electronics and that's is still true today, but I have learned a lot from people here. Thanks again.

"You always find what you're looking for at the last place you look."

(The first thing I did was actually touching up all suspicious looking solder joints, but Q9's solder joints looked OK even through a magnifying glass. Trying other harder things and failing to produce the desired result was a necessary step to go look harder all over.)