Shorting the diodes ensures no bias voltage can develop between the driver base's when it is all coupled up. As long as the outputs and drivers are functioning normally (which what we are doing now is to prove) then no DC fault can cause excessive current to flow in the outputs.
You should see the same 0.88v also on TP15 (meaning 0 volts between TP15 and TP16) and that shows no current is flowing in the output stage.
If that is OK then connect R51/52 back in place. Keep using the bulb tester and check that there is still no voltage between TP15 and TP16 (it doesn't matter if there is a high DC offset, just make sure there is still no voltage between those test points).
Lets confirm we can get it to that state.
You should see the same 0.88v also on TP15 (meaning 0 volts between TP15 and TP16) and that shows no current is flowing in the output stage.
If that is OK then connect R51/52 back in place. Keep using the bulb tester and check that there is still no voltage between TP15 and TP16 (it doesn't matter if there is a high DC offset, just make sure there is still no voltage between those test points).
Lets confirm we can get it to that state.
with R51/52 lifted and earthed as before TP15 was 0.88v as was TP16.
With R51/52 put back in place TP15 and TP 16 are at -ve 34.5v and as such there is 0v between TP15 and TP16
With R51/52 put back in place TP15 and TP 16 are at -ve 34.5v and as such there is 0v between TP15 and TP16
Those results are fine as far as it goes... the output stage appears to be operating correctly and not drawing excess current.
I would begin by checking that the - and + rails are present on R43 and R46 (the emitters of Q18 and Q23. I'm guess they are low because of the bulb tester but are they equal and opposite in value?
It is worth checking all the base/emitter volt drops around that area Q12 through to Q23. You have to be extremely careful as you are measuring the voltage across the junctions as that is the only reliable way to see. They should all be around 600mv give or take with the NPN's having the emitter as the most negative and the PNP's with the emitter as most positive.
I would begin by checking that the - and + rails are present on R43 and R46 (the emitters of Q18 and Q23. I'm guess they are low because of the bulb tester but are they equal and opposite in value?
It is worth checking all the base/emitter volt drops around that area Q12 through to Q23. You have to be extremely careful as you are measuring the voltage across the junctions as that is the only reliable way to see. They should all be around 600mv give or take with the NPN's having the emitter as the most negative and the PNP's with the emitter as most positive.
R43 and R46 have +ve and -ve rails at reduced voltage (about +/- 44v) and are equal.
I will now very carefully check B-E Q12...Q17 and report back
I will now very carefully check B-E Q12...Q17 and report back
wipes sweat from brow
all the measurements are between B-E
Q17 -ve 0.6v
Q15 0v
Q13 46v
Q16 0v
Q14 0v
Q12 +ve 0.6v
all the measurements are between B-E
Q17 -ve 0.6v
Q15 0v
Q13 46v
Q16 0v
Q14 0v
Q12 +ve 0.6v
Belay that
Those measurements just seemed wrong, and they are, I measured again pushing the probe harder into the solder joints and repetaed a few times.
all Q12..Q17 have 0.6v across B-E
Those measurements just seemed wrong, and they are, I measured again pushing the probe harder into the solder joints and repetaed a few times.
all Q12..Q17 have 0.6v across B-E
So to recap (for my own sanity)
All testing with DBT inline
D8/D9 shorted to give zero bias
Q19/Q20 and Q21/Q22 replaced
B-E checked and all at 0.6v
Q29/Q30 replaced
B-E Q30 is 0.6v on @29 it starts out at -0.6v the slowly drops to around -0.4v
Q25 replaced with PNP (Ksa992)
Q12..Q17 tested and all have 0.6v across B-E- which suggests they are working
TP15 and TP 16 equal- both have -34.5v
does this suggest that either Q23 or Q18 are faulty?
However B-E on both of them give 0.6v
The original fault was the three resistors, particularly R52 and R65 which were open, which could imply the problem is with the negative rail.
@Mooly suggested replacing the drivers Q31/Q32 which are 2SC2592 ans 2SA1112, though I have not found a near equivalent from the perspective of Ft, as the 2SA1112 is 200Mhz and most possible substitutes (MJE15034/15035) are 30Mhz. Both 2SC2592/2SA1112 have Vceo of 180v.
Now I am really confused
All testing with DBT inline
D8/D9 shorted to give zero bias
Q19/Q20 and Q21/Q22 replaced
B-E checked and all at 0.6v
Q29/Q30 replaced
B-E Q30 is 0.6v on @29 it starts out at -0.6v the slowly drops to around -0.4v
Q25 replaced with PNP (Ksa992)
Q12..Q17 tested and all have 0.6v across B-E- which suggests they are working
TP15 and TP 16 equal- both have -34.5v
does this suggest that either Q23 or Q18 are faulty?
However B-E on both of them give 0.6v
The original fault was the three resistors, particularly R52 and R65 which were open, which could imply the problem is with the negative rail.
@Mooly suggested replacing the drivers Q31/Q32 which are 2SC2592 ans 2SA1112, though I have not found a near equivalent from the perspective of Ft, as the 2SA1112 is 200Mhz and most possible substitutes (MJE15034/15035) are 30Mhz. Both 2SC2592/2SA1112 have Vceo of 180v.
Now I am really confused
That all reads OK. I suggested replacing drivers really as a matter of good practice when this kind of fault occurs but the later testing of the output stage shows they seem to be OK (the 0.88 volts at the output with the driver and output stage isolated).
Having said that the resistors that burnt were all in the driver stages.
One quick check. Do you have voltage across R51 and R52? There should be virtually none as the base current should be very low indeed.
Having said that the resistors that burnt were all in the driver stages.
One quick check. Do you have voltage across R51 and R52? There should be virtually none as the base current should be very low indeed.
Oooh, that's very unexpected. Right then...
Two choices, either the driver stage really is faulty in some way or there is another path. Q29, Q31 and the NPN outputs form a super high gain triple Darlington meaning that the base current should be microscopic when idling. So where is that current in R51 going?
One possibility is the protection diodes D111 to 115 but they would all have to leaky to do that surely. Its all so unlikely... you could lift any one of the three to break the chain and see if it makes a difference.
Have you got Q25 and Q26 fitted? If you have them remove them and retest.
Two choices, either the driver stage really is faulty in some way or there is another path. Q29, Q31 and the NPN outputs form a super high gain triple Darlington meaning that the base current should be microscopic when idling. So where is that current in R51 going?
One possibility is the protection diodes D111 to 115 but they would all have to leaky to do that surely. Its all so unlikely... you could lift any one of the three to break the chain and see if it makes a difference.
Have you got Q25 and Q26 fitted? If you have them remove them and retest.
With Q25 and Q26 removed, R51 and R52 are 0v
I did test D111 to 115 in circuit and they test OK, I can remove one of each if needed, however I ma trying to keep the connection/disconnection of the outputs to a minimum for fear of breaking one of the legs
Thanks again
I did test D111 to 115 in circuit and they test OK, I can remove one of each if needed, however I ma trying to keep the connection/disconnection of the outputs to a minimum for fear of breaking one of the legs
Thanks again
Just to clarify: the input sides of both R51 and R52 are grounded? R51 is PLUS 2.1V?
I'm going to project what Mooly would likely ask: would you report all emitter and collector voltages of Q29 through Q38?
Thanks.
Thanks.
If already installed, might as well leave them in place. We'd have to eventually test their presence anyway.
I'll have a look later but the amp should be fully built up at this point apart from Q25 and Q26 which should still be removed.
That is the voltage across them.
So just to be sure 🙂 There is zero volts across each resistor now whereas before when we had Q25 and Q26 fitted there was 2.1 volts across R51
That is the voltage across them.
So just to be sure 🙂 There is zero volts across each resistor now whereas before when we had Q25 and Q26 fitted there was 2.1 volts across R51