SWR Workingman's 15- Voltages- DC on Speaker

[strato56]

If this amp was on my bench :
I would desolder Q4 and Q10. The drivers Q3 and Q9 have DC-paths on their own and should be able to stabilize the output (No-load) to zero volts.
If that doesn't happen, well maybe some new voltages would learn me something.

[glebert]

Noticed the schematic calls out "power ground" over by the zobel network. I don't know what is being used for reference for these measurements but it seems like that is where they should be referenced to.

[BStringThumper]

I know you said the voltages are changing as you are probing, but these numbers don't make sense. To get the the 55-ish volt drop across R30 you have to have 2.5mA of current, and Q6 is not conducting because it is reverse biased (and no voltage drop across R30-B) so the 3mA is going through Q7 but that should cause a 0.5V drop across R30-A, which is not there.

Hi Glebert, I will investigate the POWER GND reference, but I think my voltage rails are reading correctly. I'll try anything. I also changed the resistors on R30A & R30B to exact 220 ohm (matching). The numbers don't make sense to me either.

Update. I upload the new voltages. I only updated the numbers in the circle because everything is reading higher today, so please ignore the outputs and bias string.
The POWER GND is going to the speaker jack negative side. It has good continuity to ground...…….. the jacks had star washers on them.... but maybe they should be isolated? That wouldn't be much of a ground. I'll unground them next try. The Q6 & 7 numbers changed.... again. Maybe the matched 220 resistors?

Thank you everyone for helping.

Update: ungrounding speaker jack from chassis ground made no difference to the 31Vdc on output.

[glebert]

Do you still have this powered up with a light bulb limiter? I ran out of ideas a long time ago, now grasping at straws.

[DrGonz78]

I know that you replaced Q5 with a good substitute. I was looking at the picture that you had posted of the main board and was looking at the screw missing connecting the Q5 to the heat sink, at least that's what it looked liked. Very tough to tell, as I only could see there was thermal past but could not tell if there was mica insulator in place. It does look like other screws are missing and I would want to confirm there are no ground connection missing connected to the power amp board.

[log1982]

As I wrote before, and as glebert alluded to, the circuit is doing exactly what schematic says it should do. Need more pulldown current than what the schematic is showing.
Wanna test the theory? Change R23 from 2.7k to 27k.

Jcon.

[BStringThumper]

I know that you replaced Q5 with a good substitute. I was looking at the picture that you had posted of the main board and was looking at the screw missing connecting the Q5 to the heat sink, at least that's what it looked liked. Very tough to tell, as I only could see there was thermal past but could not tell if there was mica insulator in place. It does look like other screws are missing and I would want to confirm there are no ground connection missing connected to the power amp board.

Hi DrGonz78, Q5 is not bolted down. There are mica insulator there for Q4, Q5, Q10, so no grounds there. There are also no grounds underneath the circuit board hold down screws. Great thought. Please keep them coming.

[BStringThumper]

As I wrote before, and as glebert alluded to, the circuit is doing exactly what schematic says it should do. Need more pulldown current than what the schematic is showing.
Wanna test the theory? Change R23 from 2.7k to 27k.

Jcon.

Hi log1982, thanks for the suggestion. 27k seems a bit extreme, should I try something in-between 2.7k & 27k? I'll give it a try.

[BStringThumper]

Yes, sir, still on the limiter. I'm grasping at straws and pulling what's left of my hair out.

I'll look at it again tomorrow. Good night.

[glebert]

Yes, sir, still on the limiter.

Is the bulb glowing at all? If not you should be able to take it off the limiter, might make the voltages stabilize better.

[g1]

Hi log1982, thanks for the suggestion. 27k seems a bit extreme, should I try something in-between 2.7k & 27k? I'll give it a try.

What value do you find R23 in your amp to be?
If log1982 is correct, then there is a typo and someone has changed that resistor in your amp to match the typo. Obviously the amp worked at some point so it's not possible this is some design flaw.
What is the history of the amp? Something someone tried to fix or was discarded? Or something that was working one day and not the next, as seen by the owner?

edit: also please post the values found installed for the other resistors log1982 mentioned:
R23, R24, R30, R39a, or R37.

[Enzo]

I thought log was suggesting the different resistor only to prove a point that it would change the current through the VAS. I didn't get the impression he thought that was what was supposed to be there.

I could be wrong too.

One thing to remember is this amp used to work. Unless we believe this thing always put 30v on the speaker right from the factory. SO redesigning the circuit wouldn't seem to be the real way to fix it. Like if my closet door won't close right because a coat hanger is in the way, I don't get out a plane and shave the door down to "fix" it.

I hada no idea we were still on the bulb limiter. The limiter is only for saving fuse blows and not stressing an amp with serious problems. But this amp is not blowing fuses, so get rid of the bulb, it does skew all voltage readings.

All that follows could be wrong:
I think the circuit wants Q8 turned on harder. You now seem to have 10ma through R39 and 7ma through R39A. The more positive the base of Q6 the more it turns off (remember PNP and "upside down") That tends to drop current through R39, making the voltage atop it more negative. That lowers the voltage on Q8E, relative to the base. That turns it on harder. Likewise reducing current through Q6 means the drop across R30/30B drops, making it more positive. Over on Q7, the base is tied to zero volts where the signal enters. If R30 goes more positive, that means the emitter of Q7 goes more positive relative to the base. The base thus more negative. Again PNP, so that should turn on Q7 harder, when pulls Q7C more positive. it is wired to the base of Q8 so that should turn on harder, which is what we want.

Looking at some other SWR circuits I see the same 2.7k 470 470 and bias transistor in the collector circuit of the VAS.

[glebert]

Don't know if it helps much, but this thread (https://music-electronics-forum.com/...ad.php?t=34764) has some voltages called out for this model amp that is a lot closer to working than the OPs. See post #9.

[Enzo]

His numbers show that our amp here is trying to correct the DC offset, and can't.

[glebert]

Rather than saying Q8 isn't pulling enough current could it be that Q5 is not allowing it to have enough? Since Q5 is replaced by a substitute part and there is a note about changing the value of R27 based on the Q5 beta I wonder if that factors in. Original part had a large hfe range (15-150), replacement part specd at 40.