A non-PL quasi-comp output transistor question

[62vauxhall]

I strung three 1/2 watt 3.3K ohm resistors together. Measured total was 9.91K ohms.

Voltages at both ends of R611 are very close. By positive top rail of R611, is that the end of R611 that joins R607?

So just so I am clear....tack the 10K oh resistor from there to the collector of Q603?

And can you clarify:
....to see if it moves the needle or not.

 

[Gepetto]

I strung three 1/2 watt 3.3K ohm resistors together. Measured total was 9.91K ohms.

Voltages at both ends of R611 are very close. By positive top rail of R611, is that the end of R611 that joins R607?

So just so I am clear....tack the 10K oh resistor from there to the collector of Q603?

And can you clarify:
....to see if it moves the needle or not.

Yes where R611 joins R607, correct. One end of your resistor goes here.

The other end of this 10K resistor goes to Q603 collector.

Moving the needle means does it make a difference in the outcome or not. Or is it no change?

 

[62vauxhall]

Changes, yes.
There is no relay click
It does not go into protection
There is only a few millivolts at pin 5 were there used to be about 35 volts. And I believe that was why protection kicked in.

Still no adjustment possible by the bias pot. Test point voltage remains at 0.0 volts.

I tried using a CD player into AUX and listening through headphones. I believe the headphone jack recieves a stepped down signal from the main amplifier. But there was no audio.

There are only two of four output transistors in each channel. I thought that would be sufficient to amplify signal but perhaps not?

No relay click presumably means main amplifiers not engaging. I considered that maybe the relay might be dedicated to the protection circuit. Up until now, when the relay clicked, the protection LED lit up immediately.

 

[Gepetto]

Changes, yes.
There is no relay click
It does not go into protection
There is only a few millivolts at pin 5 were there used to be about 35 volts. And I believe that was why protection kicked in.

Still no adjustment possible by the bias pot. Test point voltage remains at 0.0 volts.

I tried using a CD player into AUX and listening through headphones. I believe the headphone jack recieves a stepped down signal from the main amplifier. But there was no audio.

There are only two of four output transistors in each channel. I thought that would be sufficient to amplify signal but perhaps not?

No relay click presumably means main amplifiers not engaging. I considered that maybe the relay might be dedicated to the protection circuit. Up until now, when the relay clicked, the protection LED lit up immediately.

Seems that Q603 is at least one of your problems since it moved the needle.

You are indicating that with this pullup applied, the output went from -35V to a few millivolts of ground. Please confirm.

Do you have an MJE15033 or MJE15035 to replace it with. The pinout looks the same.

Get that Japanese junk out of there and put in a good old American transistor. That 2SA818 only has a collector current maximum of 50mA for a TO-220 case.

Get the current source working first and then we will go after the bias next.

 

[WOPL Sniffer]

Seems that Q603 is at least one of your problems since it moved the needle.

You are indicating that with this pullup applied, the output went from -35V to a few millivolts of ground. Please confirm.

Do you have an MJE15033 or MJE15035 to replace it with. The pinout looks the same.

Get that Japanese junk out of there and put in a good old American transistor. That 2SA818 only has a collector current maximum of 50mA for a TO-220 case.

Get the current source working first and then we will go after the bias next.

But they make good cars don't they Joe?

 

[laatsch55]

Jani had a 1982 Honda Accord, sold it 5 years ago....still a good car..

 

[62vauxhall]

I had an MJE15033 (one only) until earlier this week. As fate woulg have it, I broke a 2SA818 on the opposite channel board and used that MJE15033 to replace it.

Tomorrow, I can go to a repair shop that I've bought a few parts from and ask if they have either MJE15033 or 15035 on hand. If they have anything at all, there is a good possibility an NTE will be offered as a substitute. If that's the case, I'll get it anyway. At least it would be something use it for the time being.

If that doesn't pan out, I'll need to order from Digikey, Mouser, etc. That would be a pity as I wanted to get the full set of 8 output transistors for this thing along fuses, and some other bit & bobs. But I didn't want to buy eight output transitors until I knew they wouldn't be money wasted.

 

[Gepetto]

But they make good cars don't they Joe?

Seems I have heard that before Perry, and recently

I wish they would use better steel in the bodies and hardware bolts.

 

[62vauxhall]

Seems that Q603 is at least one of your problems since it moved the needle.

You are indicating that with this pullup applied, the output went from -35V to a few millivolts of ground. Please confirm.

Do you have an MJE15033 or MJE15035 to replace it with. The pinout looks the same.

Get that Japanese junk out of there and put in a good old American transistor. That 2SA818 only has a collector current maximum of 50mA for a TO-220 case.

Get the current source working first and then we will go after the bias next.

What I can say is that I measured -35volts at pin 5 several times before putting in that 10K ohm bridge. I can say too that once that bridge was in place, I got 2 to 3 millivolts with respect to ground.

However, I just now removed that 10K ohm bridge, and there was no reverting to previous conditions.

And I also just now saw that both 5A rail fuses on this board got blown in the process. I don't have any more so something else I can try finding tomorrow.

At first I thought Q603 somehow got damaged but using that same DMM series of measurments, it tested fine for a PNP transistor.

 

[Gepetto]

What I can say is that I measured -35volts at pin 5 several times before putting in that 10K ohm bridge. I can say too that once that bridge was in place, I got 2 to 3 millivolts with respect to ground.

However, I just now removed that 10K ohm bridge, and there was no reverting to previous conditions.

And I also just now saw that both 5A rail fuses on this board got blown in the process. I don't have any more so something else I can try finding tomorrow.

At first I thought Q603 somehow got damaged but using that same DMM series of measurments, it tested fine for a PNP transistor.

Have you checked the 4 diode array mounted to the heatsink to see if it is good or not?

 

[62vauxhall]

There are what I thought were single diodes - one per main amp board that connect to pins 2 & 3, and do attach to the heatsink . No, I had not checked them.

There are four diodes inside? Heat sensing I presume.

Putting meter leads to pins 2 & 3, there is .928 volts in one direction, OL in the other.

 

[Gepetto]

There are what I thought were single diodes - one per main amp board that connect to pins 2 & 3, and do attach to the heatsink . No, I had not checked them.

There are four diodes inside? Heat sensing I presume.
WHat does your good channel measure?

Putting meter leads to pins 2 & 3, there is .928 volts in one direction, OL in the other.

View attachment 46683

 

[62vauxhall]

The good channel is .999 volts one way, OL the other.

 

[Gepetto]

Clearly not 4 silicon diodes in series...but both look like possibly 2 in series.

 

[Gepetto]

Have you confirmed that Q604 is good? Your fuse blowing when applying the 10K may indicate your bias is not working properly and Q604 is instrumental in maintaining bias. If open, it will not be possible to maintain bias control. The fuse blowing is often a result of bias runaway. Do you have any 1N4148 diodes?

 

[62vauxhall]

I think Q603 met it's end. There is now voltage where the meter should say OL.

I was only able to get six 5A fuses today. No MJE15033 or 15035 on hand so can't get one until I order online. I thought there was still some life left in the 2SA818 but nope.

When I put two fresh fuses in, I disconnected the 10K reisistor. As soon as power was turned on they popped. Two more fuses and now with DBT and fuses stayed intact. DBT was quite bright and it used to dim right down.

Again, I checked Q603 both with my testing device then a DMM. The testing device display said "two diodes". With red at C and black at E the DMM displayed .732. Should say OL.

So I guess my next step is ordering MJE15033 or 35's. And the output transistors too which I think are $10 each. I was hoping not to do that until this amplifier board was functioning.

EDIT: I left the 10K resistor disconnected during the above.

 

[62vauxhall]

Have you confirmed that Q604 is good? Your fuse blowing when applying the 10K may indicate your bias is not working properly and Q604 is instrumental in maintaining bias. If open, it will not be possible to maintain bias control. The fuse blowing is often a result of bias runaway. Do you have any 1N4148 diodes?

I just now checked Q604 and all good for an NPN.

And yes, 1N4148 diodes are on hand.

I didn't see this post until after my posting the one above. Can I assume that Q603 failing is another reason fuses blow and DBT stays bright?

 

[Gepetto]

I think Q603 met it's end. There is now voltage where the meter should say OL.

I was only able to get six 5A fuses today. No MJE15033 or 15035 on hand so can't get one until I order online. I thought there was still some life left in the 2SA818 but nope.

When I put two fresh fuses in, I disconnected the 10K reisistor. As soon as power was turned on they popped. Two more fuses and now with DBT and fuses stayed intact. DBT was quite bright and it used to dim right down.

Again, I checked Q603 both with my testing device then a DMM. The testing device display said "two diodes". With red at C and black at E the DMM displayed .732. Should say OL.

So I guess my next step is ordering MJE15033 or 35's. And the output transistors too which I think are $10 each. I was hoping not to do that until this amplifier board was functioning.

EDIT: I left the 10K resistor disconnected during the above.

What are you planning on putting in for outputs?

 

[Gepetto]

I just now checked Q604 and all good for an NPN.

And yes, 1N4148 diodes are on hand.

I didn't see this post until after my posting the one above. Can I assume that Q603 failing is another reason fuses blow and DBT stays bright?

I was going to tell you to put 2 1N4148s in series and attach the anode of the pair to the collector of Q604 and the cathode at the emitter of Q604

This will make the bias low but should allow for a safe bring up.

Are the 2 outputs shorted now? Are they damaged?

 

[62vauxhall]

I was going to tell you to put 2 1N4148s in series and attach the anode of the pair to the collector of Q604 and the cathode at the emitter of Q604

This will make the bias low but should allow for a safe bring up.

Are the 2 outputs shorted now? Are they damaged?

First answer, either MJ15022 or MJ21194. Which ever I can get. Digikey had stock some weeks ago.

Good call on the outputs. The two I had in the left channel are blown.