GW GPC-3030, constant voltage not working

[magnehh]

Hi!

I have a GW GPC-3030 lab bench power supply that has a faulty channel. The Master and 5V channels work perfectly, but the Slave output just doesn't want to limit the voltage. Constant current for this output works fine.

The Slave kind of sort of works when in parallel or series modes, but still not 100%.

I suspect it might have something to do with the ancient OP AMPs on the board, but it might just as well be something else. I noticed the +15V and -15V rails give me a differential 27V. Not sure if that's the problem, though.

Here's a largely identical schematic: https://www.dropbox.com/s/4hi27dikav54pkh/GW%20GPC-3030.pdf?dl=0

Does anybody on the forum have experience with this power supply or have you seen similar behavior in similar power supplies?

Best regards from new forum user! :-)

[macboy]

Looks like U305 is current control and U304 is voltage control for the slave side. The base of Q303 is constantly fed current via R321 and zener VR315, and this ultimately drives the output transistor(s). At any given time, either the current or voltage regulation should be active, pulling some of that base current away in order to reduce the output current and/or voltage to where it should be.

When in constant current mode, you will expect U305 output (pin 6) to be somewhere around ground* (i.e. + output terminal of slave) with D307 forward biased, and U304 output (pin 6) to be pegged at/near the positive supply (pin 7, +15V) with D305 reverse biased.

When in constant voltage mode you will expect the opposite: U305 output near +V (+15V) with D307 reverse biased, and U304 somewhere ~= gnd (+ out) with D305 forward biased.

*somewhere around ground means low enough to forward bias the diode (D305/307) and start drawing current through it, likely a few volts above gnd, but definitely not pegged near the rail.

I would look at the output of those two opamps, but also look at those diodes; in particular if D307 is shorted, or if D305 is open, it could cause your issue.

[magnehh]

Thank you very much for the tips!

I've done some more testing. It seems D307 is always reverse biased. The output of U305 swings between +15V and  -10V. Sometimes it would also settle around 0V. So the OP AMP might be fried. It does seem to get slightly warm, but not more so than the corresponding OP AMP for the Master output.

Anyway, this leads to Q303 being off constantly, when measured "a long time" after the CV/CC lights toggle. However, the base voltage does vary smoothly when current limit is set high and I vary the voltage knob. It's always negative, though, between -1.7V to -8.5V.

It just hit me that some power supplies have a sense input. This power supply does not have such a thing externally, but perhaps there is internally? That could explain the sudden jump to max voltage.

[mij59]

There is probably a short in the output stage of the channel, check Q304 and D3071, also check all auxiliary supply voltages e.g. +15V, -15V, +24V etc.

[magnehh]

Those Q304 and Q3071 seems to be fine. I tried switching transformer windings and the power transistor from the working Master channel, but the behaviour is still the same.

I removed the main board today. To my disappointment I discovered a bunch of horrible wave soldered joints. Could seem like the board was not laying flush in the wave solder machine. Nevertheless, all components seem to have sufficient solder on them to get contact with the board.

When in parallel tracking mode, the voltage on the Slave channel is controllable, but it says CC no matter what I do.

[mij59]

Since D307 is reverse biased, the base of Q303 can only be pulled down by D305, assuming R321 and D316 are OK.
Maybe the tracking switches do not work properly.
Are you sure power supply complies to schematic ?

[magnehh]

Yes, that is true. I'll have to make some more measurements, when I get home tonight.

The switches themselves operate smoothly. Perhaps the board they are mounted on is the problem. That would be board 512-3.

I feel fairly confident the schematic matches quite well, except for the fact that I have analog meters instead of digital readouts. I've not yet seen another discrepancy between my unit and that schematic. But I'm not 100% sure of course.

I'll have to study up on what those 555 timers are doing on 512-3.

[magnehh]

Allright, so I discovered a bodge. A resistor soldered between two points on 512-3. In addition to that, one of the front panel pots has started to grind. And on top of that, the soldering on the main board looks like it was done by 5 y/o kids. I'm starting to think this crappy power supply isn't worth the effort.

Clearly, Goodwill (GW) didn't care about quality.

Maybe I'll retrofit another main board and adjustment board into the chassi some day.

I'm very grateful for all the help you guys've given me! Thanks!