Power Supply putting out to much v.

[ShackMan]

I came across an Workman PS20 12v 20amp power supply for a few bucks at a yard sale the other day. It is putting out 24.4v. It does not use the 723 regulator chip my other PS use. It has a PCB with a D313 feeding a 2n3771 that drives 4 2N3771’s. How can I check them to find out where the prob is and fix it?

Bob..

[Malc C]

it maybe rated at 20amp under load. Without load on the voltage may well be more that that stated on the device

[JohnMuchow]

I'm with Malc C, try putting the power supply under a 10A (or so) load and measure the output voltage then. 24.4V is a pretty darn high no-load voltage for a 12V supply but it's worth trying.

John

[Malc C]

Have you posted over on http://forums.qrz.com/showthread.php?t=177262 ?

Saves posting the info here - but lots of good advice

[ShackMan]

Thanks everyone. I did post it there and started checking part per part this afternoon. Still have a few small NPN s9013 and a s9014 to pull and check them. So far all is checking good.
Thanks.

[Ron W3FJW]

Have you considered that the PS may be mis-labeled?? I've had that happen in the past.

[vintage1]

OK, remove ALL the 2n3771s and test with cheap resistance meter. All should show a diode junction between base and emitter and base and collector.

Chances are one or more will be dead shorts. Replace and try again.

The voltage looks like its the unregulated unloaded output of the bridge rectifier/capacitor part of the PSU.

Someone has probably shorted the PSU and blown the output transistors, and probably an internal fuse, and then replaced the internal fuse and found it no longer regulates.

If the basic thing is working apart from that, you would find the voltage on the base of the driver 2n3771 will be below its emitter as the control stuff tries to turn off the output stage.

If all this fails to work its probably time to trace the circuit..and post it here.

Ae there any integrated circuits at all? if not that's good news.

[andrew b]

I have seen these before in various guises. They use a 15v 400mW zener as current source, thats amplified by a med power transistor (2sd313?) and then to a few power transistors, 2n3771 / 2n3055 etc. bolted onto a slab of aluminium sheet Quite a simple circuit (current follower) with no o/p current sensing or shut down but a short seems to destroy ALL the semiconductors on the board.

When something goes short (usually the o/p transistiors) then full rectified voltage appears on the o/p terminals and cooks up CB radios etc that are conected to it. :(

The mains transformer is usually about 20 volts AC out into a bridge / big cap rectifier giving about 25 DC volts. The bridge is usually 4 big chunky diodes and I would take the "20 amps" with a LARGE spoon of salt, anything over 8 amps will get it hot. :censored:


Remember 3 Taiwanese / Italian Amps = 1 imperial Amp :)

[vintage1]

I have seen these before in various guises. They use a 15v 400mW zener as current source, thats amplified by a med power transistor (2sd313?) and then to a few power transistors, 2n3771 / 2n3055 etc. bolted onto a slab of aluminium sheet Quite a simple circuit (current follower) with no o/p current sensing or shut down but a short seems to destroy ALL the semiconductors on the board.

When something goes short (usually the o/p transistiors) then full rectified voltage appears on the o/p terminals and cooks up CB radios etc that are conected to it. :(

The mains transformer is usually about 20 volts AC out into a bridge / big cap rectifier giving about 25 DC volts. The bridge is usually 4 big chunky diodes and I would take the "20 amps" with a LARGE spoon of salt, anything over 8 amps will get it hot. :censored:


Remember 3 Taiwanese / Italian Amps = 1 imperial Amp :)

If its a triple darlington follower to a crude zener, then a short will indeed take the ruddy lot out.

And at 20A it might easily be pushing 60-70W out in heat. `That is not beyond 4 TO3 output transistors on a fairly large heatsink, but you will be up around 100C on the transistors.

[ShackMan]

I pulled the four pass 2N3771's one at a time and tested with my DMM and they all tested good. I then tested the driver 2N3771 and it showed good. So I turned to the reg. board and started testing there. The d313 and two s9013's and one s9014 checked good. 4 small diodes and one 1n5404 checked good also.

I thought this just can't be right. So I put together a res and an led and a 9 volt batt. and went back and started testing again. I found one of the pass 2N3771's was indeed bad. I replaced it with a 2N3055 I had on hand and now all is well. The power supply is working great. I'll order a full set of 2N771's and few extra's to have on hand.

Thank you all for your help.

Bob..

[olmod]

Try and get GE ones there some awfull brands out there ;)

[vintage1]

I pulled the four pass 2N3771's one at a time and tested with my DMM and they all tested good. I then tested the driver 2N3771 and it showed good. So I turned to the reg. board and started testing there. The d313 and two s9013's and one s9014 checked good. 4 small diodes and one 1n5404 checked good also.

I thought this just can't be right. So I put together a res and an led and a 9 volt batt. and went back and started testing again. I found one of the pass 2N3771's was indeed bad. I replaced it with a 2N3055 I had on hand and now all is well. The power supply is working great. I'll order a full set of 2N771's and few extra's to have on hand.

Thank you all for your help.

Bob..

That is a pretty fortunate fault..often more than one go!

Since the design is pretty much open-loop there is no problem with substituting transistors with widely differing frequency performance (3055's are not exactly fast) and its a pretty reliable if rather low spec gain wise device.

But then the 2N3771 is no great shakes either. I would say a 2N3055 is a perfect swap out actually.

Do the power transistors have emitter resistors in to current balance? If not that might be a reason why one has popped..probably about 0.1 ohms is a sensible value.