Soundcraftsmen troubles

[still4given]

Hi Guys,

I bought a PSR800 not working. While checking it out I found it had rail voltage on the right output. I opened it up and found the 2N3440'S and the MM4003'S were litererally falling off of the board. I have now replaced all of them as well as all of the 1N4150 diodes. I'm still reading rail voltage on the right side. I'm attaching the schematic in hope that you guys might give me some direction as to where to look for the problem.

Thanks, Terry

P.S. I reading negative volts on the output if that helps.

[still4given]

Maybe this will be clearer.

Blessings, Terry

[jaycee]

Maybe they have failed, overheated which has melted the solder. I would install new devices.

In fact you would probably be best checking all the semi's for open/hard shorted condition. You also have a good left channel to take readings from and compare.

[still4given]

Hi JC,

Thanks for replying. When you say replace the devices are you talking all of them?

I was thinking the same thing about comparing the readings from the two channels, as a matter of fact I was trying to do that last night, but they are layed out totally differently on the board so it is difficult to trace each of them down to compare. there is no silk-screening on the board so I have to try to follow the foil traces and track them down. That's very time consuming.
I was hoping that there were a couple of key places that the rail voltage would be likely to transfer through and I could start looking there. There are a few symbols on the schematic that I don't understand, like the
---> >--- symbols. Are those just direction arrows or do they represent a component? I know I'm probably too green at this to tackle this project but I was hoping that with your guys' kind help I might get it fixed and learn something in the process.

Thanks again, Terry

[jaycee]

No, I meant the ones you found loose, they are the VAS and the current source to drive the VAS.. if the VAS is bad it will most likely pull the output to the rail. Looking at the circuit, you may also wish to check the VBE multiplier (the transistor near the potentiometer labelled "BIAS")

It's also definately worth testing the MOSFETS, one of them could well be short due to whatever caused the VAS/driver stage transistors to unsolder themselves.

edit: I may have misunderstood when you said " I have now replaced all of them" to mean "I resoldered them into place correctly" and not "Replaced with brand new parts".

edit 2: Yeah, those funny arrows just look like arrows showing the direction of current flow through the output stage. Bit bizarre if you ask me... but...

[still4given]

Thanks Jaycee,

Yes, when I said I replaced them I meant I bought new parts and replace all of the 2N3440's and MM4003's, although I replaced the MM4003's with 2N5416's since they don't seem to make the MM4003's anymore. They weren't just unsoldered, the leads were actually eaten away by corrosion. The same with a couple of the 1N4150's so I replaced all of those.

As to the VBE, are you refering to Q5?

What's the best way to test the MOSFETs for a short? Diode test? Ohm?

Same question for the MPSL01's How best to test them in circuit?

I notice on this amp that the cases or the T-03 MOSFETS are all connected to their perspective heatsinks. Then the heatsink is connected to the PCB.

Weird thing about those little arrow looking symbols is that they are all pointing away from the MOSFET that they are associated to. If that is representing current flow, wouldn't there need to be at least one lead flowing toward the MOSFET? Maybe it represents a plugin connection. I see the same symbol at the red clip LED. That part has plugin leads. I don't think the MOSFETs plugin on this amp but maybe they did when this schematic was drawn.

Also, any idea what the diamond symbols are representing? The ones with the .400 and 0.7 inside?

Thanks, Terry

[jaycee]

As to the VBE, are you refering to Q5?

Yep, that's the one.

What's the best way to test the MOSFETs for a short? Diode test? Ohm?

Ohm test as far as I know - I'm pretty sure there was a thread a while back on here about how to test MOSFETS.

Same question for the MPSL01's How best to test them in circuit?

In circuit is probably not a good idea. Test out of circuit with a diode test. Main thing to check is that there is conduction only one way between collector/emitter.

I notice on this amp that the cases or the T-03 MOSFETS are all connected to their perspective heatsinks. Then the heatsink is connected to the PCB.

Hmm, that sounds rather unreliable to me, but a good way to do without the heatsink insulators.

Weird thing about those little arrow looking symbols is that they are all pointing away from the MOSFET that they are associated to. If that is representing current flow, wouldn't there need to be at least one lead flowing toward the MOSFET? Maybe it represents a plugin connection. I see the same symbol at the red clip LED. That part has plugin leads. I don't think the MOSFETs plugin on this amp but maybe they did when this schematic was drawn.

Aha, yes, you could be right there. Whatever they are they're not mega important though.

Also, any idea what the diamond symbols are representing? The ones with the .400 and 0.7 inside?

I would say these are indications of what the voltage should be across the part in question.

I did think of a quick test for the outputs being short. If you pull Q9 and Q12 out of circuit, there should be no voltage on the output at powerup. If there is, you have a shorted 2SJ50.

[still4given]

I would say these are indications of what the voltage should be across the part in question.

Cool, I'll check that. On second thought, that would be at full power. I better wait until I resolve the other problem before run this thing back up to full power.

I did think of a quick test for the outputs being short. If you pull Q9 and Q12 out of circuit, there should be no voltage on the output at power up. If there is, you have a shorted 2SJ50.

OK I see that. Although, if that were the case then my problem would have to be the MOSFETs wouldn't it?

Is there nowhere else that the rail power could be getting to the output?

Let me make a couple of observations and then ask a couple of questions.

When I measured the output voltage at the speaker terminals, I hooked up the red lead to the red speaker terminal and the black to black. I read -65V. This made me believe that it was the negative rail that was bleeding through.

Is this a fair assumption?

Also when I follow the negative rail on the schematic it appears that the only places that the negative rail connects to the output are at the MOSFETs Q13 & Q14 and through Q3 and Q2.

I just thought of one other thing that may or may not be important. As soon as power is applied, the red clip indicator lights up on the bad channel.

Blessings, Terry

[jaycee]

Cool, I'll check that. On second thought, that would be at full power. I better wait until I resolve the other problem before run this thing back up to full power.

I see you have a variac from earlier posts. The good old lightbulb trick might work better for this, as it's limiting current.

OK I see that. Although, if that were the case then my problem would have to be the MOSFETs wouldn't it?

If you remove both driver transistors and you're still getting rail voltage then yes, one or more MOSFETS are bad.

Is there nowhere else that the rail power could be getting to the output?

From the schematic, no. Check the PCB for shorted tracks just in case though.

When I measured the output voltage at the speaker terminals, I hooked up the red lead to the red speaker terminal and the black to black. I read -65V. This made me believe that it was the negative rail that was bleeding through.

Is this a fair assumption?

Correct. The black speaker terminal will be connected to GND of the power supply, so if youre reading -65V then that's the negative rail voltage on your output.

Also when I follow the negative rail on the schematic it appears that the only places that the negative rail connects to the output are at the MOSFETs Q13 & Q14 and through Q3 and Q2.

Not through Q3 and Q2 - the only connection to the output is via Q2's base... it's unlikely thats the cause, Q2 would probably burn up if it was short enough to cause rail voltage to flow through it to the output.

I just thought of one other thing that may or may not be important. As soon as power is applied, the red clip indicator lights up on the bad channel.

That's just the clip detector seeing the stuck rail voltage as a clip.. So at least the clipping detection works

[still4given]

OK, I pulled the Q 12 and Q9. no change still voltage on the output. So I disconnected everything that connected up to either the + or the - rail and still voltage on the output. I thne checked and the voltage is on the heatsink. After looking more closely to the schematic it appears that the heatsink is what is connected to the output. Now I don't understand MOSFETs too well so this may be a dumb question. Is it possible that a MOSFET that is not switched on could allow the rail voltage to pass. When I pulled Q9 & Q12 on the good channel, I read rail voltage there too. After I soldered them back in it dropped back down.

Is this possible?

If not then I guess my MOSFETs are toast.

Blessings, Terry