SWR Super Redhead with power amp problem

[TASBrendan]

Hi all,

I have a SWR super readhead on the bench and it has a problem with R9 smoking. I just want to see if I'm on the right track with this one. Solid state is new to me and I feel like I've been throwing stuff at the wall and nothing is sticking....
There has been some previous work not by me where the fan thermal switch was bypassed with a solder blob across the pads. When I undid that, R9 started smoking. Another part of that previous work is the ground trace on the board was cut to isolate the fan ground and a wire was run back to the transformer ground lead. The audio ground still runs to the ground tab of the speaker jack.
So where I'm at now is that R9 smokes, even having been replaced with a 1W resistor and the fan mod being put back to how it was when I got it.
I also replaced R7 with a 2W cement resistor as the carbon resistor that was there looked to be a 1W and had discolored bands, but no smoke.
I suspect Q14 is dead- hooked it up to my transistor tester and got no Beta reading. This tester is also new to me, so I don't have a lot of experience with it.
Attached is a schematic with some notes. With the fan off, R9 still smokes. Fan could be totally unrelated and just happened at the same time. Also, this happens even when I have it plugged in to a current limiter that shows no dead shorts and no obvious current increase as R9 starts to go.

Any and all help is appreciated.
Thanks
Brendan

[g1]

Do you have a load or speaker connected? You should not, until the power amp is working properly.
If you have a load connected that could explain R9 smoking if Q7 is bad.

[The Dude]

What was the original problem/symptom with the amp aside from the modification? Was there something not working correctly? Did R9 smoke prior to you undoing the modification? Aside from beta testing Q14, did you check to see if the transistor is shorted? Is there any DC voltage at the amp's output?

[TASBrendan]

No load connected, but I'll check out Q7- hadn't thought to look. I tested for shorts on Q14 with the continuity on my meter, but no shorts. The original problem was a "smell." So the first test I did was to leave it on, but idle for an hour and watch for heat. No smell, not much heat. Then I got more info that the smell happened when playing a chord at high volume or using a bass fuzz pedal. I put a sine wave in and ran it a little hot to see what warmed up, but didn't get any smoke. I noticed Q14s metal can was hot and saw that R7 was discolored. That's when I noticed the solder blob on the fan circuit. Should have just left it, didn't realize it was just bypassing the thermal switch. Have not tested for DCV on the output. Can you tell me what that would show? I can have the amp on for about a minute before R9 smokes.
Thanks

[TASBrendan]

Hi Duder,
So, yes I do get DCV at the output (J5). I brought it up on the variac and it gets to about 20VDC at J5 by the time the input is around 100VAC, which is when I can smell what I assume is R9 starting to go. I kept it low at 40VAC and probed around Q7. With the +V set to 25VDC at the collector of Q7, the base has ~ 21VDC and the emitter has 20.5VDC. The emitter voltage drops a little on the other side or R19 and a little more at J5.

[The Dude]

The fan circuit has little to do with this problem. If it wasn't running at all, it might cause the amp to overheat and fail, but otherwise you have a problem in the amplifier section. I'd ignore the fan thing for now and concentrate on the amp. The fact that the amp is outputting DC indicates an amp circuit problem, which is why I asked about output DC. The burning resistor likely indicates a short, so it will continue to run hot until the short is found. Do not hook up a speaker until you verify that there is no longer DC on the output. I would do a quick check of all transistors in the path of R9 and see if any are shorted for starters.

[TASBrendan]

Cool, thanks so much for the help! I'll go through all the transistors again and make sure I'm not missing a short somewhere. The only one that's tested no Beta on the tester so far has been Q14 and I have a replacement showing up hopefully today. The fan was modded to run all the time with the thermal switch jumpered.

[TASBrendan]

I think I'm seeing a short at Q6. I'm getting a meter reading of 1Ohm between the emitter and collector. And when I use the diode mode, I get near 0v between E and C in both directions and a really low .0024 V from base to both emitter and collector. Am I right, this is faulty? And would that cause heavy current through R9? Seems like it's on the wrong half of the amp.

[TASBrendan]

Thank you so much for your time and insight. I want to make sure I'm testing the transistors in circuit correctly. Also R20 only reads 37Ohms- should be 120Ohm and D3 and D5 both read ~ .54 V in one direction, but ~ 1.5V in the other. Diode check on Q5 gets b to e = .073v in both directions. (-) b to (+) c gets OL, and reversed ~ .58. (-) e to (+) c gets OL and reversed gets ~ .59. D4 reads ~ .54 in one direction and ~1.5 in the other, while D2 reads ~.54 in one direction on OL in the other.

[g1]

Sounds like Q5 and Q6 are both bad. Remove them and check them again out of circuit.

[J M Fahey]

Sorry but pulling all parts to be measured outside in the hope of finding a "bad part" by sheer chance is the long unfruitful path.

You need to plug the amp into a dim bulb limiter, 40 to 60W bulb recommended, remove any load, set all controls to 0, turn it ON and start measuring voltages.
Start with rails, speaker out, voltage *across* each transistor CE (collector to emitter) and BE.

In principle, any transistor must have voltage of the proper polarity (positive for NPN, negative for PNP) and same for BE, only here we expect around 0.65V

Anyone not meeting that, is suspect.

That is the practical way.

[The Dude]

......The fan was modded to run all the time with the thermal switch jumpered.

Nothing wrong with that, and probably keeps the amp much cooler. FYI, early versions of that model didn't have that 300 ohm 5w resistor in parallel with the thermal switch, so it was common for people to bypass the switch to keep the fan running and the amp cool. Later versions added the resistor so that the fan runs slower all the time and faster when the switch gets hot enough to close. But again, that is not your immediate problem.

[TASBrendan]

Thanks g1, I pulled Q5 and Q6. Q5 tests good but Q6 has a short from e to c. I don't understand if that would pull current through R9- what should I know about that? Would it be safe to run at low voltage without Q6 to verify that's also the problem with R9?

[g1]

Check that the other odd readings from post #9 are ok now with Q6 removed.
Q5 and Q6 are current limiters/protection, so you can run without them to test at lower voltages. I assume you have a variac.
R9 supplies Q4. If C6 is ok, I would think the R9 issue is Q4 related.
When you have it running at lower voltages, post DC readings for E,B,C of Q7 - Q12.

[TASBrendan]

Thanks g1, I'll give that a shot next. I do have a variac and it's been helpful for this one. I did check R20 with Q6 removed and it now reads correct 120 Ohm.
Thanks J M Fahey, that's some transistor trouble shooting info I badly needed. I do have a light bulb current limiter, but it's a 150W since I mostly deal with tube amps. I had it plugged in, but it wasn't showing me much- I've seen it light up more for tube amps idling normal. And R9 still smoked.