Carver FTM-25

[lowell]

This amp came in with one channel sounding distorted. Found that the test points in the power amp Left channel were off. There should be +/-3.2v on the bases of the pre-drivers Q207/209. I resoldered the heatsink mounted Vbe multiplier - Q205 and the issue was fixed. Gave back to customer. He came back, same issue with those voltage test points. I resoldered the entire board around the bias and driver section of the amp, was fixed, gave back to him. He came back, same issue. I put it on my bench and the voltages are wrong again. I poked and prodded, checked voltages again, now it's working again. So, at this point I'm thinking I need to replace the smaller electrolytics in the circuit, ala C205 which is across the Vbe transistor. And any other similar electrolytics.

I don't want to do this without first really finding the issue, but I've failed at that up till now. I have tried tapping, pushing, freeze spray etc... none of which seem to change the condition. Only powering down, waiting, powering up seem to change things from working to faulty.

How would y'all go about this? Advice please. I fear replacing these caps, it then working fine, giving it back to him, and then having a very angry customer on my hands. Y'all know how it is, I just want to fix his gear for him in the end, but having to bring it back 3 times is quite frustrating.

[The Dude]

I don't have any tips. It sounds like you've tried most everything. The only thing I might try is ESR testing the caps, if you have a meter. Sure you can replace them, but as you said, then you won't know for sure if you fixed something. If the caps are drying out, an ESR test might show that. I feel your pain!

[lowell]

I have an esr meter Ill see if it reveals anything. Thanks!

[MarkusBass]

I poked and prodded, checked voltages again, now it's working again. So, at this point I'm thinking I need to replace the smaller electrolytics in the circuit, ala C205 which is across the Vbe transistor. And any other similar electrolytics.

Lowell, you know that this is the worst method of fixing amps and this was stated here many times by Enzo and other guys (including me). It's simple because of the fact that after replacing this capacitor, you wouldn't be able to tell whether the amp is now fixed, or not. Another question is: how much would you charge the customer if you don't know whether the amp is fixed.
For me, you haven't tried any basic test procedures that should be applied in such a case. If you have 0 Volts on the capacitor (you wrote that "test points were off" - I assume it means 0V), I would keep the amp in "failed" state and measure surrounding voltages. If you have power supply rails and 0 Volts on some capacitor in the middle of the circuit, then somewhere in between there is some interesting voltage that could help you to solve the problem.
For example voltages on Q201, D201, D203 and similar. Why haven't you done this? This would bring you much closer to the solution of the problem. "Poking and prodding" is not a good method in this case. You can use it but only if afterwards you identify the problem. Otherwise, you are left without any knowledge about the failure. You also leave us without any reasonable info about the problem but, on the other hand, you expect advice. My advice is: do it properly - meaning do more measurements.
Apart from this: the only reason related to C205 that there is 0 Volts across the cap is that it is shorted. And shorted capacitor can be verify without desoldering it (in circuit). Have you tried this? If the capacitor is not shorted, I would search for the problem elsewhere. This could be for example cracked track (to be verified with magnifying glass).

Mark

[lowell]

Markus thanks for the reply. I have tested voltages and attempted to track down the issue. Problem is, after testing a few points....I mean a few, the problem mysteriously goes away. Hence the poking and prodding to to get the problem to RETURN.

The first time I fixed it the voltages earlier in circuit seemed off... at the Vbe mult transistor. So I resoldered it as my meter showed no continuity from it to the board. I hope this clairifies things. I'm poking and prodding to UPSET the circuit. Not to fix it.

Those test points both read -1.3v when the fault happens.

[MarkusBass]

Those test points both read -1.3v when the fault happens.

OK, now I understand. But you only wrote that "the test points were off" (whatever this means) and now you say that both test points were -1.3 V. This is not the same. You also did some measurements but you didn't tell us about it. You realize, that this make diagnosis much more difficult, don't you? The more we know, the better we can help you.

Mark

[lowell]

I realize Mark. The trouble is, that everytime I start measuring I'll get one or two measurments, then the thing suddenly starts working right. At that point, I cannot get the FAULT condition to happen. It's very frustrating and seems that only a complete power down and hours later does it do it again. I'm convinced it's electrolytic related because nothing seems to be loose and solder joints don't seem to be the issue. I'm gonna toss it on the bench right now and see how many measurements I can get.

I do recall testing both Vbe transistors as well as the current source transistors and all surring resistors and diodes. All tested fine with the meter. Again, all this points to a leaky electrolytic. Let me see if I can get some more solid details for you and I'll post them.

[lowell]

Well now we really have something to troubleshoot. Borrowing a friends meter, it doesn't have a clear marking on the selector knob. I accidentally had it on ma instead of vdc (they're 180deg on the turn dial). Blew the fuse and now the thing is blowing fuses. I ended up shorting Q209 (pnp driver) emitter to gound, so that transistor was a dead short. Replaced it with MPSA92 after swapping the base/collector holes. So it's still drawing too much current, and I'm using the bulb limiter.

I've checked all the outputs, no shorts. Voltage are all really low unless I lift the Re ballast resistors R247/249. I've had them lifted for troubleshooting as the voltages seem good and also symmetrical with them lifted.

Now, I have +5v at BOTH Q201/203 collectors. Q203 should be a low negative voltage. I've disconnected Q205/237 and C205. Still +5v in both spots. Q201 has Vbe of 0v. I replaced it thinking it had a b/e short or internal open circuit. No change. So I then questioned how this could be. The only way it could be was if Q103 in the preamp wasn't on. Sure enough no voltage across R125.

Q103 has a Vbe of -.8v. (using bulb limiter so voltages are low)
Q103 Ve = 0v.

I've tested all the resistors in the Base biasing string. All test fine.
+/-15v are reading +/-3.3v (seems low because of bulb but still symmetrical)

IC101 has these voltages:
1 -1.5v (here's my problem, but why and how?)
2 0v
3 0v
4 -3.3v
5 0v
6 0v
7 280mv
8 +3.3v

Please help!

[MarkusBass]

I haven't looked at all the voltages you listed, but it's clear that you shorted some transistors. Keep checking.
The voltages on IC101 pins #4 and #8 should be -/+15V (instead of 3.3V). This may be due to bulb limiter but I would check it.

Mark

[g1]

Also, with your supply voltages so low with bulb, I don't think you should expect the IC to behave properly, so the DC offset at pin 1 may just be due to the low supply.

[lowell]

Thanks for your help guys. Yeah I realize I should ditch the bulb with the ballasts lifted since it's not in over-current. I'll report back.

[lowell]

Ok so with ballasts disconnected I have:

IC101
1 -14v
2 +5v

With pins 1/2 disconnected and R111 lifted I have these voltages on the solder pads:

1 +32mv
2 +5.5v

The only place I see thats DC connected to pin2 is via R113 which is the 0v point between both halves of the PA. So...there is offset. +5v of offset. This could only be from the limiter transistors as I see it. Q301 is either on more than Q303...or Q303 isn't on enough.

I also see only 88mv across R289 on Q203 NPN collector. The other working PA has 300mv across that resistor.

Q203 tests fine with a meter. There is 750mv across D205/207. There is 930mv across D201/203.

The positive side of the PA seems to be on more. Not EXACTLY sure why. Please, suggestions!

[lowell]

Please ignore my previous post as everything has changed.

Ok I lifted D309/307 and saw smoke. Turns out that was a really bad idea. I meant to lift D305/307 to take the limiters out of circuit. R243/245 are burning up and smoking. I reconnected the diodes and replaced the resistors and Q215 as it was shorted in and out of circuit. I used a KSC1845 for Q215 as it was the closest I had. I swapped the B/C pins.

R243/245 are still smoking. Using variac at 22V and 100w bulb limiter. I have to take these measurements fast cause the resistors keep smoking.

Admittedly....I seem to be making this worse the more I work on it.

Q207
Ve -11v
Q209
Ve -16.5v
Q201
Vc -10v
Vb +21.6v
R209 has 62mv across it
Q203
Vc -18v
Vb -19v
R211 has 2.4v across it

Seems Q203 is on WAY MORE than Q201

So tracing backwards the Rc of Q103/105 in the preamp.

Q103
R125 has 150mv across it
Q105
R127 has 800mv across it

I lifted IC101 pin1 AND R111.
The junction of R117/119 is at -170mv.

+/-15v rails (with variac at 22v and 100w bulb) are at +4.3 and -4.7.

I have a million questions but I think I'm just gonna leave this here and see if anyone has any insight. I'm not feeling too competent at this point.

[lowell]

In such a situation. What do you experts do? Would you replace the output transistors with resistors? Would you disconnect the preamp and use resistors to "pull down/up" the bases of the 2 PA input transistors?

I feel I'm not dividing and conquering enough and am chasing my tail.

Finally, any references or books on advanced repair techniques for direct coupled transistor amplifiers?

I understand how transistors work. And THINK I understand this amp... but it's the approach, and isolating smaller sections of the amp for troubleshooting that's throwing me for a loop.

[Jazz P Bass]

I would have to wonder why R243 & R245 are getting so hot.

Where is the current path?
Are they not they part of the limiter circuit?
If so, you can remove the limiter transistors for now.

As to inserting resistors instead of the output transistors, I have always used a 1K/1 watt resistor from Base to Emitter.
For testing, you only need to add one per rail if the transistors are in parallel.

(aside: I will not take in anything made by Bob Carver!)