A non-PL quasi-comp output transistor question

[Gepetto]

As time and patience allowed, I continue trying to find the fault with this board.

Something I recently did and probably should have done at the get go, was measure voltages at all the pins. Then identify, as Joe recommended, which components they lead to. I also noted what voltages should be present.

Then doing likewise with the functioning right channel board for comparison.

Everything on the right side more or less corresponds with the schematic. Still have a DBT in place so some actual readings were a bit below but (+) voltages were (+) voltages, (-) voltages were (-) voltages and millivolts were millivolts.

Not so with the troublesome left channel board. The (-)35 volts triggering protection via pin 5 is also present at:

~(-)35 volts @ pin 7 to collector of output transistor Q003 - diagram says 0 volts
~(-)35 volts @ pin 8 to emitter of output transistor Q001 - diagram says (+)0.06 volts
~(-)35 volts @ pin 9 to emitter of output transistor Q002 - diagram says (+)0.06 volts
~(-)35 volts @ pin 14 to base of output transistor Q001 - diagram says (+)0.65 volts
~(-)35 volts @ pin 15 to base of output transistor Q002 - diagram says (+)0.65 volts

~(-)35 volts @ pin 2 to cathode of heat sensing heatsink diode - didn't see diagram spec but right channel pin 2 reads (-)178 mv
~(-)35 volts @ pin 3 to anode of heat sensing heatsink diode - didn't see diagram spec but right channel pin 3 reads (+)1.8 volts

I am very likely trouble shooting all wrong but since contending with negative voltages in the wrong places, my thought was to identify the various transistors on the board that see -B negative volatages that are somehow linked to those pins.

That conclusion has made me focus on:

Q601 and Q602 - a pair of 2SA750's......I have read 2SA970's can replace those and I have some on hand.
Q605 and Q606 - a pair of 2SC1628's....I think KSC3503 can replace those. If so, I have those also.
Q610 - a 2SA814. This one was already replaced at some time in the past with a 2SB546A. By chance, the case split open of the corresponding 2SA814 on the right board. I replaced it with an MJE15033 after learning it was acceptible and I have more of them.

Any further insights or opinions are most certainly appreciated.

View attachment 47104

Yes, further insight. You will not have any method of getting off of the negative 35V floor UNLESS you have some path to positive.

Either the amp is trying to drive to that negative rail OR it is trying to regulate but has no path to the positive rail.

The more meaningful data may exist on the upper half of your schematic.

With no Q001 - Q004 installed, what results when you lift the collector lead of Q605. That is the class A amplifier in this schematic. Does it magically go positive then or are things still stuck negative?

 

[62vauxhall]

Thanks Joe for directing me to the collector of Q605. That transistor is installed backwards. Someone in the past worked on this amplifier before I received it.

I'll pull it, test it and report back. If it's salvageable, will reinstall the right way round.

EDIT: Emitter & collector were reversed.

 

[mlucitt]

Emitter & collector were reversed.

We see this from time to time.

 

[Gepetto]

Thanks Joe for directing me to the collector of Q605. That transistor is installed backwards. Someone in the past worked on this amplifier before I received it.

I'll pull it, test it and report back. If it's salvageable, will reinstall the right way round.

EDIT: Emitter & collector were reversed.

Recall earlier today, I described zener breakdown in the emitter-base junction if reverse biased in an adjacent post? That would be occurring here. Not sure if you have a survivor after this or not. Might be walking wounded.

 

[Gepetto]

An MPSA42 is a suitable and better device if you happen to have any.

Was this heat sunk in any way?

 

[62vauxhall]

The eureka! moment was short lived.

Pulled the transistor, did a quick check on my "device", seemed OK and reinstalled. No change. Still in protection and still (-)35 volts. Pulled again, installed without collector connected and again, in protection and (-)35 volts.

No heatsinks. Just raised off of the board a bit.

If not a survivor, I don't have MPSA42's.

KSC3503 are a no go? I have a couple and could replace Q605 & Q606 to make a pair.

 

[Gepetto]

The eureka! moment was short lived.

Pulled the transistor, did a quick check on my "device", seemed OK and reinstalled. No change. Still in protection and still (-)35 volts. Pulled again, installed without collector connected and again, in protection and (-)35 volts.

No heatsinks. Just raised off of the board a bit.

If not a survivor, I don't have MPSA42's.

KSC3503 are a no go? I have a couple and could replace Q605 & Q606 to make a pair.

OK so keep going with the debug. You removed the source of any negative drive by lifting the collector of Q605 if the schematic is accurate.

Next step, with Q605 still lifted, next lift the EMITTER of Q608 and check if it stays negative or if it pops up to a positive voltage. You have to follow the trail.

 

[62vauxhall]

Q605 collector still lifted, lifting emitter of Q608 caused a change. Still goes into protection but negative voltage dropped from (-)35 volts to (-)25 volts.

 

[Gepetto]

Q605 collector still lifted, lifting emitter of Q608 caused a change. Still goes into protection but negative voltage dropped from (-)35 volts to (-)25 volts.

And now with that in the same condition, lifting one end of R627 gets you what result?

 

[62vauxhall]

Negative voltage down further to (-)16.3 volts.

 

[Gepetto]

Negative voltage down further to (-)16.3 volts.

So at this point you have every path to negative disconnected with one possible long shot. Lift one end of R624 and see if it changes.

 

[62vauxhall]

Big change. All points with negative voltages are now (+)52 volts.

 

[Gepetto]

Big change. All points with negative voltages are now (+)52 volts.

OK that indicates to me that you may have some leakers that may have been damaged by the Q605 put in the wrong way (reversed).

I would start with putting an MJE15033 in location Q608 with all 3 leads attached and see the result. (the 2SA818 data sheet says it is a TO-3 with a B-C-E pinout order, do you agree?)

And leave all the other conditions the same as before with just this one change.

If you are still good with positive voltages, then I would reattach R624 next and measure the result.

 

[WOPL Sniffer]

 

[62vauxhall]

View attachment 47125

It's because I haven't put in any of the expensive parts yet.

 

[Gepetto]

It's because I haven't put in any of the expensive parts yet.

And you can't until this thing is back regulating again like it is supposed to.

 

[62vauxhall]

OK that indicates to me that you may have some leakers that may have been damaged by the Q605 put in the wrong way (reversed).

I would start with putting an MJE15033 in location Q608 with all 3 leads attached and see the result. (the 2SA818 data sheet says it is a TO-3 with a B-C-E pinout order, do you agree?)

And leave all the other conditions the same as before with just this one change.

If you are still good with positive voltages, then I would reattach R624 next and measure the result.

Not a TO3 (TO220) but that is the right pinout.

I'll do as you instruct and report back.

 

[Gepetto]

Not a TO3 (TO220) but that is the right pinout.

I'll do as you instruct and report back.

Yes, meant TO-220 and typed TO-3 out of habit.

 

[62vauxhall]

With the MJE15033 installed (bent legs to adjust for pinout) voltage remains (+)52 volts.

Reconnected leg of R624 and voltage still (+)52 volts.

EDIT: And now to bed. I must get up in 6 hours.

 

[Gepetto]

With the MJE15033 installed (bent legs to adjust for pinout) voltage remains (+)52 volts.

Reconnected leg of R624 and voltage still (+)52 volts.

EDIT: And now to bed. I must get up in 6 hours.

Not sure what "bent legs to adjust for pinout" means? The MJE is TO-220 with B-C-E pinout. The 2SA818 is TO-220 with B-C-E pinout according to datasheet.