Acoustic 125 Low Output

[potatofarmer]

Schematic: http://music-electronics-forum.com/a...schematics.pdf
For reference, here's a similar topic: http://music-electronics-forum.com/t36001/

The power PCB is 170080.

I'm working on an Acoustic 125 for a friend which has low output. The schematic calls for injecting 140mV into the power amp and getting 10V out. I was getting closer to 2V out. (Test signal is 1kHz and the load is a pair of industrial heating pads that come to ~15 ohms.)

First I checked the DC on the board - the voltages on Q1 and Q2 were bad - wish I'd written them down, but I think the emitters and bases were measuring the same. All the resistors around them measured OK, so I replaced those with 2N3906 and the DC measurements fell in line. Output is now 3.6V with 140mV in.

The DC measurements around Q3 and Q4 seem fine. Q6's emitter is measuring 39V DC, while the schematic calls for 36V DC.

This amp was worked on about a year and a half ago (the previous tech left a sticker) and Q7 was replaced with a 2N5416 and clip-on heatsink. Q5 is still the original RCA 40409 with the original heat sink. Seems odd he'd only replace one. I'm tempted to contact him, but I'm guessing the work was done for a previous owner and it probably worked fine for some length of time after it left his shop so it doesn't really seem like it's his responsibility.

I'm assuming I should see something like 10VAC on Q5 and Q7, but it's just the 3.6V that's also on the output.

I don't know much about solid state amps, so I don't know what I should be looking for AC-wise. Clearly something's not making enough voltage gain, but I don't understand the schematic well enough to pick a culprit.

Any suggestions as to where I should start looking? Thanks.

[Enzo]

The signal level ought to be about the same on Q4,5,6,7,8. Q203 is the voltage amp, once through it, all the rest of the stuff to the right is a current amp for the speaker.

SO go explore that voltage amp. It is a 40408. Is it amplifying?

[potatofarmer]

Sorry it's been a long time between replies; haven't had a chance to look at the amp until today.

Q3 was barely amplifying - 3.6V on the collector - and the DC test point labeled 37.5V was at 43V. Not sure if I just failed to notice that earlier or if something changed. I have a 2N3440 on hand so I swapped that in after verifying it was an acceptable substitute. Pretty sure I bought it to work on a blown Crate amp I picked up to learn on, so it's serving its purpose.

Anyway, now I'm getting 4V out; Q3 collector is still at 43V, base is at 0.6V, emitter is grounded. Does have continuity to ground.

I'm assuming the DC voltage reading high is a flag of some sort. Lifted 1 leg of R9 and it's reading fine (6k8 from base to ground).
R18 and R19 are reading okay in-circuit, 1k3 so I didn't bother to lift one leg first.
D3 and D4 are both reading okay; 0.5V forward and over-limit reverse.
I pulled C8 to measure it (27pF from Q3 collector to base) and my DMM is awful for reading capacitors but it measured around 70pF. Was not shorted. I didn't want to risk oscillation by powering up without it installed, so I kludged together three 100pF caps in series. Didn't have an effect, so the original cap was probably still OK. I'll replace it with a new 27pF.

Should I start looking at Q4? Measure the voltage drop across R19 and R18? The only thing I can guess is that Q3's collector is high because Q4 isn't drawing enough current.


Possibly unrelated, but what's the purpose of R4? It's supposed to be selected from beween 68k to open. There was a 260k resistor installed there - it looked really rough so I pulled it. No noticeable effect on output, so I'll probably replace it with a less-eroded resistor. I'll get another 260k unless there's some target I should be shooting for- I'm guessing it's biasing Q1?

[g1]

All DC voltages shown are contingent on that B+ of 76V. Is that what you have there?
If not, you want the midpoint at Q6 to be roughly half your B+. I expect that is what R4 adjusts.
Q3 bias is set by current through Q1 collector across R9.

[potatofarmer]

Yeah, I assumed that. Before I started the supply voltage was a couple volts higher than on the schematic, but with the new Q3 the supply voltage is getting pulled down.
This is probably easier to read than the wall of text:

       Target  Actual
Supply: 76       72
Q6-E:    36      39.8
R17:     72       68.6
Q3-C:   37.5    42.1
Q3-B:   0.6      0.6
Q2-E:   18.6    18.75
Q2-B:   18       18

This is with the 260K R4 re-installed. With it pulled, the voltages on Q1 and Q2 were around a volt high. Though now I realize that with the supply lower than 76, the voltages on Q1&2 should be lower as well.

Can/should I just install a 250k pot as a variable resistor in place of R4 to find a value where that mid-point is pulled to half of the supply voltage?

[J M Fahey]

The power PCB is 170080.
The schematic calls for injecting 140mV into the power amp and getting 10V out. I was getting closer to 2V out. (Test signal is 1kHz and the load is a pair of industrial heating pads that come to ~15 ohms.)

Please retest with NO load .
I want to separate the "not enough voltage gain" and "not enough current into a load" problems.

Q6's emitter is measuring 39V DC, while the schematic calls for 36V DC.

Will be adjusted later, but no big deal.

This amp was worked on about a year and a half ago (the previous tech left a sticker) and Q7 was replaced with a 2N5416 and clip-on heatsink.

Your problem may lie just there.
40409/40410 had fixed heat sinks, not removable, which to boot had 2 mounting pins on opposite sides which were soldered to the board into their own holes and solder pads for mechanical stability.
They also electrically joined (shorted together) such pads through heatsink metal, and in fact acting like a big "hidden in plain sight" jumper between them.
So many designers who had the eternal problem of drawing a track from "here" to "there" without crossing others gratefully used that "undocumented feature" and drew a track into one heatsink pad, continued from the other.
Run the Time Machine forward, a Tech replaces an unobtanium 40409/40410 with a different TO39 transistor, with a floating heat sink (your case) or a plastic TO126 cased one (what I did) and amp stops working or works poorly because it has been lobotomized, part of the circuit has been disconnected.
So search for those empty pads where original heat sink was mounted and join them with a piece of wire.
It might be all you need.

Q5 is still the original RCA 40409 with the original heat sink. Seems odd he'd only replace one.

Well, if it was working fine ..........

I'm assuming I should see something like 10VAC on Q5 and Q7, but it's just the 3.6V that's also on the output.

Like Enzo said, voltage gain is in Q3 .
Any AC voltage you get on its collector you will also get on following transistors and output, because those provide only current gain, but unity voltage gain.

I don't know much about solid state amps, so I don't know what I should be looking for AC-wise. Clearly something's not making enough voltage gain, but I don't understand the schematic well enough to pick a culprit.

Voltage gain is set by NFB so itīs 220k/2k7 = 80 or thereabouts.
Real world specified 10V/0.14V=71 ... real close.

You have less output voltage than expected, but it may have 2 very different origins: bad voltage gain or plain inability to drive 15 ohms, hence the first test suggested.

Anyway, now I'm getting 4V out; Q3 collector is still at 43V, base is at 0.6V, emitter is grounded. Does have continuity to ground.

Fine.

I'm assuming the DC voltage reading high is a flag of some sort. Lifted 1 leg of R9 and it's reading fine (6k8 from base to ground).
R18 and R19 are reading okay in-circuit, 1k3 so I didn't bother to lift one leg first.
D3 and D4 are both reading okay; 0.5V forward and over-limit reverse.
I pulled C8 to measure it (27pF from Q3 collector to base) and my DMM is awful for reading capacitors but it measured around 70pF. Was not shorted. I didn't want to risk oscillation by powering up without it installed, so I kludged together three 100pF caps in series. Didn't have an effect, so the original cap was probably still OK. I'll replace it with a new 27pF.

Do the first test and restore heatsink "jumper" continuity.

Possibly unrelated, but what's the purpose of R4? It's supposed to be selected from beween 68k to open. There was a 260k resistor installed there - it looked really rough so I pulled it. No noticeable effect on output, so I'll probably replace it with a less-eroded resistor. I'll get another 260k unless there's some target I should be shooting for- I'm guessing it's biasing Q1?

Itīs a Factory adjusted mid point biasing resistor,cheaper than a preset and canī+t be messed with by Musicianīs itchy fingers.
Noitce they also avoided a bias trimmer, probably for the same reason.

[potatofarmer]

Please retest with NO load .
I want to separate the "not enough voltage gain" and "not enough current into a load" problems.

140mV in, 3.02V out with no load. I did check loaded vs unloaded early on and didn't notice a difference so I just kept the load connected as it's already hooked up to my scope and such.

40409/40410 had fixed heat sinks, not removable, which to boot had 2 mounting pins on opposite sides which were soldered to the board into their own holes and solder pads for mechanical stability.
They also electrically joined (shorted together) such pads through heatsink metal, and in fact acting like a big "hidden in plain sight" jumper between them.

On this board, the heatsink pins don't seem to be connected to anything. I've attached a picture of the top and bottom of the board. I could run a jumper to the outside of the case if you think that'd help, but I'm not sure what it should go to.



I haven't contacted the tech, but he is a reputable tech and builder, so I'm giving him the benefit of the doubt that it was working fine when it left his shop.

Q5 is still the original RCA 40409 with the original heat sink. Seems odd he'd only replace one.

Well, if it was working fine ..........

Ah, ok. I was under the impression that output transistors were typically replaced in pairs, and I've seen posts about people being concerned that they're matched or not. Scoping the output showed a symmetrical sine wave, so that seems fine at least.

I'm reasonably confident that this is just a voltage gain issue - early in my testing I increased the input signal until I got full output, clipping on both halves of the waveform at around 20VRMS or so, but it needed around 400mV of input to get there. Sorry; I was trying to keep my first post concise and apparently forgot to mention that.

[J M Fahey]

Ok, I see that in this particular board heat sink was not used as a free link/jumper but I have seen others, so just keep it in mind it happens to be needed.
The gain is set up by R14 220k , R12 2k7 , and C7 4.7uF .
Check them and absolute worst case replace them, then retest.

[potatofarmer]

Ok, I see that in this particular board heat sink was not used as a free link/jumper but I have seen others, so just keep it in mind it happens to be needed.
The gain is set up by R14 220k , R12 2k7 , and C7 4.7uF .
Check them and absolute worst case replace them, then retest.

Yep, it was C7! The resistors were fine and since I don't have a good way to test capacitors I just replaced it. Swapped back in the original Q3 (since it didn't need to be replaced) and now the DC and AC voltages are spot on. Actually got this finished a week or two ago, just forgot to update this thread. Thanks for all your help!