Bench power supply outputs and displays high voltage until a load is attached

[sycho123321]

    In the past couple months my el cheapo HY3005D bench supply will only go as low as ~8v with no load attached. The power supply does output that voltage when measured with a multimeter, but once any load is attached it drops to zero volts and must be adjusted up from there. This began occurring a few months ago, and I have no idea what caused this behavior. Any ideas?

Thanks!

[CJay]

Schematic contained in this thread I *think*

https://www.eevblog.com/forum/beginners/fixing-my-mastech-hy3005d-3/

Basic tests to do would be to check the pass transistors (2N3055), their drivers and the current sense resistors, check the control pots and make sure the resistance rises and falls smoothly as they are rotated.

[Decapitator]

It appears that there is a small amount of leakage current in the pass transistor section. If you have a sensitive enough voltmeter, you *may* be able to detect an extremely small voltage across the low value/high wattage emitter resistor associated associated with the offending '3055. Otherwise pull the pass transistors one at a time. CJay's advice is on the right track.

[sycho123321]

Thanks guys, I'll check around there.

[sycho123321]

If you have a sensitive enough voltmeter, you *may* be able to detect an extremely small voltage across the low value/high wattage emitter resistor associated associated with the offending '3055.

I measured each of the three high power resistors connected to the 3055s and each measured ~2mV, I'm not sure how small of a voltage you were talking, but each reads approximately the same value.

Edit: also checked the pots they seem fine

[Kleinstein]

There is a resistor as a internal minimal load to take into account some leakage of the output stage - this might be broken, or may be to large for the transistors used.
If something like a 2 K resistor load is enough to bring the voltage down, chances are high it is the internal resistor (2.2. K).

[sycho123321]

There is a resistor as a internal minimal load to take into account some leakage of the output stage - this might be broken, or may be to large for the transistors used.
If something like a 2 K resistor load is enough to bring the voltage down, chances are high it is the internal resistor (2.2. K).

Just tested, and attaching a 2.2k resistor actually causes the voltage to rise from 8.8v (lowest the psu will adjust) to 11v.

[Kleinstein]

If a small load make the voltage rise, this points toward some kind of AC problem. So either ripple of oscillation at low load. This would be a more difficult problem.

A first step in a repair is often a visual inspection. Sometime broken parts / lose wires are visible.

An other test without opening, similar to the external 2 K resistor would be an external electrolytic capacitor in the 100-500 µF range, just in case the internal one is broken / loose.

The circuit diagram shown in the other thread looks a little odd. There seem to be at least one error (missing GND symbol ?). The shown circuit might be prone to oscillation if the voltage is set to minimum. So it might be better if the pot is not all the way towards 0.

[sycho123321]

With a 470uf cap it goes down to zero volts just fine. I took a look at the output without a cap with my scope and it outputs noisy ramps

[tooki]

I suspect the output cap (the electrolytic mounted on the banana jack PCB, if your 3005D is like mine) has failed.

I've been working on pimping my 3005D (better displays, output load switch, etc.), and one thing I discovered is that with that cap disconnected, the thing couldn't really regulate at all. (It stayed at something like 55V!)

Your discovery that adding a cap stabilizes it would support the failed output cap theory.

[james_s]

Certainly sounds like it. Classic symptoms of an oscillating regulator.

[sycho123321]

Took it apart to get to the banana board, I checked the 470uf cap and it was a sick puppy read 0 uf. Replaced it and it's all fixed!    Thank you, guys!

[james_s]

Nice. I think I might be inclined to replace the rest of the electrolytic caps in it with reasonably quality name brand parts.

[sycho123321]

Yeah I probably should, but all I have on hand currently are wun hung lo caps. I'll have to order some decent quality ones.