Thanks again
a quick recap as to the state of the amp
D8/D9 shorted
Q25/Q26 removed
C17 shorted
Input shorted
D11/D12 removed
R59/R60 lifted and connected to earth
Q31/Q32 installed
when power is applied both NF and speaker relay click on, dim bulb is off, current draw is approx 0.2A
which isolates the output section from the "input" section
The voltage across R65/R62 in this configuration is 0v and TP16 is 0v
a quick recap as to the state of the amp
D8/D9 shorted
Q25/Q26 removed
C17 shorted
Input shorted
D11/D12 removed
R59/R60 lifted and connected to earth
Q31/Q32 installed
when power is applied both NF and speaker relay click on, dim bulb is off, current draw is approx 0.2A
which isolates the output section from the "input" section
The voltage across R65/R62 in this configuration is 0v and TP16 is 0v
With configuration as you describe above, would you report DC and AC on the C17 node, and emitters Q29 and Q30.
Thanks.
Thanks.
That all sounds good.
Can you measure the voltage across R56 and R57 (270 ohm) while in this state. It should be zero volts across each.
You take a step back and think and recheck 😉
No voltage across both the 270 ohms means that when the 22 ohm resistors are reconnected there should still be no voltage across the 33 ohms and nothing drawn by the final output transistors.
You have to get it to this state. First image is with R51 and R52 grounded. Second and third image show what happens if there is an offset. No current flows in any of transistors (provided there is no load attached). The voltages just follow the offset driving R51 and R52 but we still have zero current in the 270 ohm, the 33 ohm and the 0.47 ohm.
At DC and with no DC load it can't be any other way 🙂
@Mooly as soon as as I reconnect the 22 Ohm, I get 0.8v across the 33 ohm
TP16= 0v
Q29/Q30 emitter are -1/1v/1.1v
Q29/30 base is 0v
Q31/Q32 emitter are -0.8v/0/8v
and there is -1.1/1.1v across R57/R58- the 270 ohm.
In normal operation the base of Q29/Q30 is 1.8v (ignoring +/-) and emitter is 1.2v and as such they are switched on.
As Q29/Q30 have 1.1v on the emitter- does this mean they are being turned on?
TP16= 0v
Q29/Q30 emitter are -1/1v/1.1v
Q29/30 base is 0v
Q31/Q32 emitter are -0.8v/0/8v
and there is -1.1/1.1v across R57/R58- the 270 ohm.
In normal operation the base of Q29/Q30 is 1.8v (ignoring +/-) and emitter is 1.2v and as such they are switched on.
As Q29/Q30 have 1.1v on the emitter- does this mean they are being turned on?
So- with the 22ohm installed and R51/R52 grounded as in the first image, there is 0v across the 33 Ohm (R65/62).
I then installed one each of the output (2SA/2SC) and still zero volts across 33 Ohm, the relays click, there is 0 volts on TP 16 and the dim bulb is off.
This implies that some how the "front end" is causing the problem- though I am not sure how
I then installed one each of the output (2SA/2SC) and still zero volts across 33 Ohm, the relays click, there is 0 volts on TP 16 and the dim bulb is off.
This implies that some how the "front end" is causing the problem- though I am not sure how
In earlier posts, there was some evidence of oscillation: AC Voltages on bases Q29=6.4v/Q30= 6.4v
Did you intend mV rather than V?
In any case, look for AC when connected to front end. Try to determine frequency. Single sweep may help.
@BSST what would cause the oscillation - would it be a transistor or a combination of components?
there is approx 4v AC at the base of Q29/Q30 measured to TP23- ground
@Mooly
I replaced c19/c20 with 2pf (rather than1 pf) and c13 as I looked for ways the front end could impact the output.
With R51/R52 reinstalled I now have 0.3v across R62/R65, which implies some sort of progress, though as you say it should be 0 volts
there is approx 4v AC at the base of Q29/Q30 measured to TP23- ground
@Mooly
I replaced c19/c20 with 2pf (rather than1 pf) and c13 as I looked for ways the front end could impact the output.
With R51/R52 reinstalled I now have 0.3v across R62/R65, which implies some sort of progress, though as you say it should be 0 volts
The oscillation is almost certainly responsible for all the weird symptoms. I wouldn't want to even guess the culprit, but I wouldn't blame a transistor alone, but rather circuity it's embedded within. Oscillation can be "global" because of problems in the feedback path or associated compensation components or it can be a "local" oscillation provoked by nearby components, but it's seen everywhere--- very vexing to locate. I know that's not helpful. The problem could be as simple as an open cap.
I still urge trying to get a sense of oscillation frequency, as that can be helpful. The measurement doesn't have to be accurate or beautifully stable.
Did you mean C18 and C19, both originally 1pF, rather than c19/c20? That's as a good a guess as any. A similar guess would solder tack 10x bigger caps in parallel at the C15 and C16 sites; the idea is to slow the amp way down in an attempt to make it stable. That's not a fix, but a diagnostic experiment.
If an idea doesn't look promising, restore the values to original so that you don't migrate into a hodgepodge of changes. Presumably the amp worked properly at one time. Any relevant repair attempts in the past?
I still urge trying to get a sense of oscillation frequency, as that can be helpful. The measurement doesn't have to be accurate or beautifully stable.
Did you mean C18 and C19, both originally 1pF, rather than c19/c20? That's as a good a guess as any. A similar guess would solder tack 10x bigger caps in parallel at the C15 and C16 sites; the idea is to slow the amp way down in an attempt to make it stable. That's not a fix, but a diagnostic experiment.
If an idea doesn't look promising, restore the values to original so that you don't migrate into a hodgepodge of changes. Presumably the amp worked properly at one time. Any relevant repair attempts in the past?
@Mooly -
So I replaced C24 and C25, meaning all the caps except the main power supply caps have been replaced.
I also replaced the two zeners D1/D2
and got 0.27v across the 33ohm
so I got hopeful and put a pair of outputs in (2SA/2SC) and sadly very bright dim bulb.
I then took out the 2SA- relays clicked, dim bulb off
I then took out the 2SC and put in the 2SA- it looked as if the bulb was about to go bright, but it settled, went off and the relays clicked.
To me this means there is some form of "feedback" when both the outputs (2SA/2SC) are installed. TP 16 with both installed rose to at least 0.45v before I turn off the power.
Could this be the NF relay?
I am also wondering that even if the 0.47 ohm emitter resistors and the 4.7ohm R73 etc test OK out of circuit, perhaps one or more is failing under load.
My other challenge is the wires are now starting to break from the PCB being moved, even though I am being very careful.
So I replaced C24 and C25, meaning all the caps except the main power supply caps have been replaced.
I also replaced the two zeners D1/D2
and got 0.27v across the 33ohm
so I got hopeful and put a pair of outputs in (2SA/2SC) and sadly very bright dim bulb.
I then took out the 2SA- relays clicked, dim bulb off
I then took out the 2SC and put in the 2SA- it looked as if the bulb was about to go bright, but it settled, went off and the relays clicked.
To me this means there is some form of "feedback" when both the outputs (2SA/2SC) are installed. TP 16 with both installed rose to at least 0.45v before I turn off the power.
Could this be the NF relay?
I am also wondering that even if the 0.47 ohm emitter resistors and the 4.7ohm R73 etc test OK out of circuit, perhaps one or more is failing under load.
My other challenge is the wires are now starting to break from the PCB being moved, even though I am being very careful.
The only thing that makes sense is that it is not a normal DC issue.
What is bandwidth of your scope? We could be looking at something in the tens of Mhz range here. You are going to have to look (in the faulty state and using the scope) at the voltage on the output line and the voltage at the emitter ends of the 33 ohms and look for a voltage difference across these resistors. You can't normally connect the scope directly across these resistors because of grounding issues. So use DC coupling and assuming a wideband scope lets see what is there.
Unknown history is often a problem unfortunately.
I can't understand what causes oscillation when input of the output stage is grounded, like Mooly showed in his schematics. First you have to determine what causes current flow through the R57 and R58 or your kind of expression with voltage on emitters of the Q29/Q30.
Is your ground point solid, or you have some problems with your reference point, your measurement could be wrong.
I doubt there is oscillation with circuit connected like Mooly's shematics.
Is your ground point solid, or you have some problems with your reference point, your measurement could be wrong.
I doubt there is oscillation with circuit connected like Mooly's shematics.
@pitbul
If I'm not mistaken, the oscillation doesn't occur until Peter reconnects the output power section back to the amp front end. I believe the output stages are stable when R51/R52 are grounded. Peter, please correct me if I'm wrong.
I believe the amp front is grounded at Q4's input jack as a precaution, since there was some was some earlier suspicion there.
Thanks for your eyes on.
If I'm not mistaken, the oscillation doesn't occur until Peter reconnects the output power section back to the amp front end. I believe the output stages are stable when R51/R52 are grounded. Peter, please correct me if I'm wrong.
I believe the amp front is grounded at Q4's input jack as a precaution, since there was some was some earlier suspicion there.
Thanks for your eyes on.
Again, I would like to point to the S2 switch which determine is sum sensor active or it is connected via switch to the output point (on both speaker poles).
If you like to use sum sensor on the speaker side of the connection, you have to use extra wires to connect emf generated on speaker terminals, otherwise you have to switch of the sensor in the OFF position.
Maybe bad switch do not connect well on both poles, as you can see on diagram, and input stage oscillate because of missing signal from the output.
If you like to use sum sensor on the speaker side of the connection, you have to use extra wires to connect emf generated on speaker terminals, otherwise you have to switch of the sensor in the OFF position.
Maybe bad switch do not connect well on both poles, as you can see on diagram, and input stage oscillate because of missing signal from the output.
Thanks all for you input, I very much appreciate the assistance with what is turning out to be a minor nightmare.
@BSST - yes input is shorted and problem only occurs with two of the outputs (2sa/2sc) installed- whereas with Q31/Q32 installed the bulb stays off and relays click.
@pitbul - With the input shorted Q4 has 0v at the output. The earth I am using is TP 23. I will check again for a clean connection. I will also remove S2 and make sure all the switch connections are working, though I have deoxed a couple of times
@Mooly - scope is 100Mhz so will look across 33 ohm resistors- to further reduce possibilities I replaced R56/R55 and R53/R54,
If we apply occams razor, then S2 or a ground lift would seem the most obvious. I am also conscious that removing and installing the PCB and desoldering/resoldering certainly increases the chances for broken wires ( a few have broken and been reinstalled) and breaks in the track- though I test for continuity every time I change things.
I am away for a couple of days (doing fun audio things!) so will try and have a look today and then with fresh eyes when I ma back on thursday
Peter
@BSST - yes input is shorted and problem only occurs with two of the outputs (2sa/2sc) installed- whereas with Q31/Q32 installed the bulb stays off and relays click.
@pitbul - With the input shorted Q4 has 0v at the output. The earth I am using is TP 23. I will check again for a clean connection. I will also remove S2 and make sure all the switch connections are working, though I have deoxed a couple of times
@Mooly - scope is 100Mhz so will look across 33 ohm resistors- to further reduce possibilities I replaced R56/R55 and R53/R54,
If we apply occams razor, then S2 or a ground lift would seem the most obvious. I am also conscious that removing and installing the PCB and desoldering/resoldering certainly increases the chances for broken wires ( a few have broken and been reinstalled) and breaks in the track- though I test for continuity every time I change things.
I am away for a couple of days (doing fun audio things!) so will try and have a look today and then with fresh eyes when I ma back on thursday
Peter