Print

Please login or register.
1 Hour 1 Day 1 Week 1 Month Forever
Login with username, password and session length

[sewage666]

First post after finding this forum...

I'm looking for some help. I've done just a tiny bit of tube amp repair for me and friends, but now I've got an old humdinger of a fix-it project back on the bench.

My old Ampeg SVT-200T was fried some time ago at a Bay Area club with bad power. Two other bass rigs were burned up that night (I guess the power wiring there is crap, or three people in a row don't know how to hook up their amps). It was fixed by a tech who just installed an entirely new power amp board from SLM. As I recall, it fried again in the same club and I just put the thing in storage until I could suss it out. Installing a new power amp board is no longer an option, Ampeg having been bought by LOUD.

On opening, the T-90 relay was burnt and fused and R28 and R29 were toast. I replaced all of these. Not one fuse was burnt, though, so I guess those didn't really work to save anything.

Now the amp will power on, but no sound. The pre-amp works when I put a lead from the pre-out to another amp. The contact from the pre to the power amp in the loop section seems good. I've removed and tested the power transistors and they all seemed fine.

I tested all the points listed on the schematic... being a bit of a n00b inside an solid state amp, I'm not sure entirely what's going on, but TP10 is definitely way off, reading .2mVDC when it should have 16 volts. I'm going to replace the diode and electrolytic there.

I've attached the schematic here with my test notes on it (hopefully it's big enough to view). Is there anything else I should be looking for? And what the eff could've happened?

I've searched for anyone else working on this amp, but didn't find much online. It seems like when a '90s Ampeg solid state breaks, it gets tossed. These PCBs are a pain, but I'm looking to this as an educational experience and maybe get a back-up amp for my tube monster. Any help is appreciated.

[Enzo]

The fuses are not there to protect the amp.  They are there to protect your house FROM the amp.   They are there to help prevent the thing catching fire in event of failure.

TP10 is your +16v rail, it needs to be there.  Measure the diode, is it shorted?  If not, I suspect an open R24 rather than the zener and cap.   Is there +60v on the other end of R24?

SImple test.  Leave the amp on a couple minutes.  I suspect you have +60, but start with that, is there a +60 supply present?  (Can't make +16 from +60 if there ain;t no +60)  So assuming +60 on one end of R24, and essentially zero on the other end, either the resistor is open, which would leave it cold to the touch.  Or the zener/cap is shorted, which would make R24 REAL hot.

[sewage666]

Enzo... you are king.

I did as you advised. The resistor got hot, and then I realized I'd hit the wrong part of the test point I'd marked on the PCB while tracing the circuit. Oops.

It is indeed getting 16V there.

Now I'm really stumped. I'm attaching the revised schematic with notes. Any idea of where to look next would be appreciated. The only real idea I have now is to do an audio probe.

[sewage666]

I was able to follow the audio up to R1 that leads back into IC1 on pin 6. Then it disappears. R1 tests fine, as does the cap and resistor between pins 6 and 7 on IC1.

My limited knowledge would assume I could have a problem with the IC and / or with the optocoupler OC1. I don't even know what an optocoupler does here, though, much less how to test it.

[sewage666]

I found this to test an octocoupler...
http://www.antimath.info/electro/how-to-test-an-optocoupler/
And by my reckoning, I guess my octocoupler is working.

[J M Fahey]

That Led/LDR optocoupler is being used as part of a compressor/limiter.
I distrust more IC1.
Check that it's fed +/-15V to pins 8/4 respectively.
Also that there's no DC (beyond a couple millivolts) on pins 1/7 .
You will *not* have audio on pin 6 but yes on pin 7 .
I mean, *if* you inject some signal on the Line in jack.
Worst case, carefully replace IC1.
You may temporarily unsolder and lift the LDR end of the Opto, just to put it out of the way.
In that case, you must have 2X what you inject.

[Roly]

First up, thanks for being your own best friend and posting the circuit/schem complete with your measurements - that's a huge flying start.

I'm curious about how the club power situation could cause an amp to fry, but we'll leave that for the moment.

IC1 - you won't see any signal on pin 6 because pin 5 is connected to ground and that makes pin 6 a "virtual earth" point, but you should see signal on the output pin 7.

The readings at TP4 and TP5 say the +/-16V supplies are okay, despite the apparent reading at TP10, and I wouldn't expect any AC at TP4 or 5 (see table).

Like JM, if I had to guess between the opto and the op-amp I'd go for the op-amp every time.  These LDR optos hardly ever fail, and it would be very unlikely for it to fail shorted across the LDR without some external help in the form of a loose solder dag or such.  Op-amps however are a bit more fragile and not unknown to simply give up.

I feel quite strongly that you have two different, and quite possibly unrelated, situations here; the left hand side around the low level stages and op-amps, and the right hand side around the main power output stages, and I don't think they are actually related.

I think the reason the amp has stopped is around IC1, and that the damage around the output stage is an accumulation over time that had not yet got so bad as to do any serious damage - that TP13 in the speaker protector goes low and pulls in the relay T90 strongly suggests that there is nothing wrong with the power amp portion.

Again the voltages at TP8 and TP9 tell us that the main supplies are okay.

I find R28, R29 and the relay T90 being toasted particularly interesting.

TP13 should only be low and relay T90 pulled in if the output of the main amp ("half rail") is very close to ground, which it should be (the relay etc being there to protect the speakers from being burned up by DC).  So it's possible that R28 & 29 had only got very hot, and not actually failed, otherwise I would have expected to see some other damage (dead transistors?) or off voltages around the output stage.

That is, the primary problem in the output stage may have been the protection relay itself, or at least its contacts.

If you could provide a voltage reading between point (E) - the main output - and ground, and across each of R31 through R38, the 3.3 and 0.33 ohm emitter resistors (~1V, and ~10-20mV), that would help a lot in getting a picture of the condition of the output stage.

I would nominate IC1 as your most likely suspect simply because it is fairly exposed to the outside world via the 'Line In' connector, and dodgy club power could potentially have put a high voltage in via this connector if it happened to be being used at the time.

HTH

[sewage666]

Thanks so much all for the help! I'm gonna try to address this all logically and in order...

J M Fahey:
Check that it's fed +/-15V to pins 8/4 respectively.
Also that there's no DC (beyond a couple millivolts) on pins 1/7 .
You will *not* have audio on pin 6 but yes on pin 7 .

Pin 8 is getting 16.5V and pin 4 is getting -16.5
I could hear but the faintest of audio with the gain and volume cranked all the way on pin 7. I mean REALLY faint.

J M Fahey:
Worst case, carefully replace IC1.

The local Fry's Electronics store has the NTE equivalent in stock, so I'm going to do just that, after I install an IC socket... doing myself a favor for the future.

Roly:
First up, thanks for being your own best friend and posting the circuit/schem complete with your measurements - that's a huge flying start.

I've done enough lurking around forums to know better.  :)

Roly:
The readings at TP4 and TP5 say the +/-16V supplies are okay, despite the apparent reading at TP10, and I wouldn't expect any AC at TP4 or 5 (see table).

I'd misread TP10, it's good at 16V. I don't know what's going on at TP3, however... my new reading is .47VDC and -1.2mVAC. That's all sorts of not matching the table.

Roly:
TP13 should only be low and relay T90 pulled in if the output of the main amp ("half rail") is very close to ground, which it should be (the relay etc being there to protect the speakers from being burned up by DC).  So it's possible that R28 & 29 had only got very hot, and not actually failed, otherwise I would have expected to see some other damage (dead transistors?) or off voltages around the output stage.

It's been so long since I reopened this can of worms... I should've checked if those resistors were toasted or just toasty.

Roly:
If you could provide a voltage reading between point (E) - the main output - and ground, and across each of R31 through R38, the 3.3 and 0.33 ohm emitter resistors (~1V, and ~10-20mV), that would help a lot in getting a picture of the condition of the output stage.

I'm not getting any voltage at E... and I'm wondering about R39. I checked the Ohms in circuit, and it reads 0... I know that's not the proper way to check, but R40 is the same type and it reads 10 Ohm. I'll unsolder R39 tomorrow and double check. I'm not sure what you mean by read across the emitter resistors though... read each side of each resistor? Or from emitter to emitter?

Tomorrow, I'll celebrate Independence Day by shopping for some parts. Yes!

[Roly]

I'm not getting any voltage at E... and I'm wondering about R39. I checked the Ohms in circuit, and it reads 0... I know that's not the proper way to check, but R40 is the same type and it reads 10 Ohm. I'll unsolder R39 tomorrow and double check. I'm not sure what you mean by read across the emitter resistors though... read each side of each resistor? Or from emitter to emitter?

Zero volts to ground on point (E) is good, it tells us that the whole DC-coupled output stage seems to be balancing itself correctly, which in turn means it is very unlikely to have any dead or dodgy devices.

I mean measure the voltage across each of R31 to R38, probe at each end of each resistor in turn to measure the voltage between the ends of each resistor.  R31 and 32 should have about a volt across each, while the rest should have about 10 to 20mV across each.  Basically this is to check that the two drivers, Q11 and 12 are passing about the right current through R31 and 32, and that each of the output transistors Q13 to Q18 are all passing about the expected current, and therefore can be presumed healthy.

The reason this is important is that the toasty resistor R28 and R29, and the relay contacts on T90 all suggest excessive current in the output stage at some time.  This may have been due to a fault since repaired, or simply that the resistors and relay contacts were a bit under-specified and are showing some age.  I strongly suspect the latter, that there is no serious problem around the output stage (which is a damn good thing 'coz these stages can often develop terrible cases of galloping silicon cancer).

R39 is generally only used as a former for the coil L1 but may also be a damper; shorted by the coil it will only read zero ohms.

Until you can see about a volts AC of signal at TP2 there isn't much else you can do, and I suspect that when IC1 is replaced you will restore the signal level and overall operation.

Just be certain you get the new op-amp in the right way around.

[J M Fahey]

Agree and add: please do not use an NTE equivalent if at all possible.
NE5532 are very common, order a couple from a known supplier , keep tem as spares.
That said, they are not even needed in IC1, *very* light duty there, you can use a common TL072 .
Of course, as Roly said, make sure you replace it pointing the same way.
Good luck.

[sewage666]

Agree and add: please do not use an NTE equivalent if at all possible.
NE5532 are very common, order a couple from a known supplier , keep tem as spares.
That said, they are not even needed in IC1, *very* light duty there, you can use a common TL072

I replaced IC1 with a TL072. No result.

I mean measure the voltage across each of R31 to R38, probe at each end of each resistor in turn to measure the voltage between the ends of each resistor.  R31 and 32 should have about a volt across each, while the rest should have about 10 to 20mV across each.  Basically this is to check that the two drivers, Q11 and 12 are passing about the right current through R31 and 32, and that each of the output transistors Q13 to Q18 are all passing about the expected current, and therefore can be presumed healthy.

I've measured 0V passing through R31 or R32. That can't be good. Same for the R33-R38. I assume since these are ceramic resistors, they should also be running at least a little hot, which they aren't. I just pulled out Q7, Q9, Q11 and tested them according to first test here:
http://www.kpsec.freeuk.com/components/tran.htm#testing
I don't know if there's a better, more assured way to see if a transistor is failing?

I'm baffled what to try next. I'm beginning to understand why, years ago, one tech I asked about working on this amp said it was on a short list of amps he refuses. Oi vey.

[J M Fahey]

There's something that does not add up, so please do the following:
1) inject, say, a 200 mV sinewave into the line in jack.
Use an actual plug there, not simply soldering some wire to its pads.
2) measure that you actually have those 200mV there.
Any half decent multimeter will have a 200mV AC scale.
If you do not have an audio generator, download a 400Hz to 1 KHz MP3 wave and play it with a portable MP3 player into the line in.
Most of them can provide around 200mV at the earphone jack.
3) follow those 200mV along their path, you may have a cracked track or bad solder anywhere.
So check: Line in jack "hot" terminal> C1 (both ends)> R1 (first end)> IC1 pin 7 (you should have 300mV AC there).
4) if AC reaches R1 and can't be found after that, turn the amp off, wait for all PSU voltages to disappear and measure the resistance between IC1 pins 6 and 7 (I expect around 33K) ; and between 6 and 5 (I expect "open" or at least over 100/200K).
Use the 2K and higher resistance scale.

[sewage666]

Big update... I was trying JM's test, and I started getting the faintest hint of a sound coming through my speaker. A little 500Hz sinewave I was playing through my iPod into the line in. Weird, I thought. I switched ICs into the sockets I'd installed, and it didn't make a difference. So I think the ICs were fine.

While I was probing to do the voltage test, I accidentally made contact on the tip end of the line-in jack/switch. Noise. Big noise. I switched my iPod onto The Fog soundtrack (yay) and there it was, nice and loud, coming through the power amp section. Then I accidentally bumped my probe into the thermal switch. POP! I blew one of the NE5532s. So I swapped a TL072 in.

I'd had this same kinda problem with the effects loop of a VH-140C (it was cutting the signal in and out), so I had some of the weird L leg Cliff jacks that Ampeg used in the 90s lying around. This is all despite the original jack continuity testing fine, keep in mind... same thing on the VH-140C. I can't explain it.

So, I swapped the jacks for new ones, and behold! My problem was mechanical all along (and I feel really dumb)! Many thanks in helping me trudge along, or I wouldn't have kept testing. And I hate Cliff jacks, now. BUT...

Now that the amp is amplifying, it's buzzing and distorting REALLY bad, and the signal cuts abruptly while the instrument's strings are still decaying. It also only makes noise if the volume and gain are at least half way up. Could this be because of the TL072? I guess I can't really answer that until I order some new NE5532s and get this thing back to spec. Could it be the octocoupler? Could be from the original damage inflicted the first time it fried, or damage I just did shorting the thermal switch? Time to stop for now, order some ICs, and then watch fireworks. Happy fourth!

[Roly]

I'm beginning to understand why, years ago, one tech I asked about working on this amp said it was on a short list of amps he refuses. Oi vey.

If they can build it and get it going, you can repair it.  The tech's attitude may have had more to do with the commercial aspect, being time consuming and unprofitable; but as it's your gear, and your time is your own, the economics are quite different, so don't get discouraged - okay?  You have already made very serious progress.

I accidentally made contact on the tip end of the line-in jack/switch. Noise. Big noise.

(AAARRRGGHH!  )  YAY!  I'm sure JM and I would both have mentioned the jack switch contact as a very common cause of this sort of problem, were it not for the fact that it was marked on the circuit "ckd continuity"!

So, I swapped the jacks for new ones, and behold! My problem was mechanical all along (and I feel really dumb)!

Most of the time it is, and some of the time you do, but with time you learn to double check the stupid stuff first.

Now;

Now that the amp is amplifying, it's buzzing and distorting REALLY bad, and the signal cuts abruptly while the instrument's strings are still decaying.

Okay, now you are WAY ahead, but earlier you said...

I've measured 0V passing through R31 or R32.

...and what you are now observing is perfectly consistent with what you previously observed, the output stage is operating without any bias; there is no voltage across them because there is no current flowing through them (and stony cold) - and this most likely goes back to the burnt resistors R29 and R29.  We are very hot on the trail here, but now we (you) must be very careful because one wrong move around the output stage could produce quite a lot of damage.

I think that this is almost certainly a pre-existing problem, not something you have just caused.

The table says you should have +40 volts on TP6 and -40V on TP7.  My guess is that you won't have, that these voltages will be quite a bit lower.

At this point I normally revert to cold checking, and in this case I'd be pulling first Q6, the bias transistor, and Q7 and Q8 the pre-drivers to make sure they are good.

It could well be that somebody has simply fiddled with the bias setting pot (looks like "AR1" on my cct, in series with R23), but I'd want to be sure before I went changing it.  I'd also want to know the two protection transistors Q2 and Q3 are healthy.

In all these cases you must be extra careful to restore the right transistor to the right position, the right way around.

I doubt the op-amp has anything to do with this problem, but by all means restore IC1 to a 5532 if you feel happier about it.

While testing around the output stage it is best to disconnect your loudspeakers and check for near-zero volts between (K) and ground before you reconnect them each time. (Yeah, I know it has a speaker protector, but the amp also has faults and an unknown tech work history; better to be cautious...).  I would also be using a limiting lamp in the mains feed while I was working on the output stage.

You have made considerable progress and I think you are not far off finding and clearing the remaining fault(s)

[J M Fahey]

Agree and add: BUILD A LAMP BULB CURRENT LIMITER AND USE ITSearch this Forum for it.
And   disconnect speakers for now, we'll suggest some DC voltage measurements, no signal applied.