Strange problem on PSU Korad 3005

[chipeater]

Hello,
I encountered a strange problem on a Korad KA3005 power supply.
Without any load, I can adjust the voltage from 0 to 31V without any problem (controlled with a DMM).
But as soon as I exceeded a 50mA load at low voltage ( 5V ), the power supply went into OCP even if the setting was 100mA or 1A.

I wanted to take apart the power supply and I measured the output transistors  ( D1047 ) and the driver ( D880 ). Everything was fine, idem for the 1n4007 diode.
I reassembled everything and I still have the same problem.

As a test I use a 10mm 100mA LED, set the power supply to 5V 100mA but the power supply regulates well at 100mA for less than a second then only between 50 and 60mA.

And now comes the strange part :
I connected an electronic load to my power supply and I had exactly the same problem as with the LED.
Just by curiosity I tried with a higher voltage and current ( 15V 1A ) and....and everything worked perfectly strangely 

So I retested at 5V 100mA and now everything is working again 
I touched all the cables a bit to reproduce the problem but without success.

Has anyone ever had this problem?
If not, do you know where it can come from ?
As everything is working now, it's complicated to find the origin of the problem.

I am afraid that this could happen to me again so I would like to know where it could come from and change some components to prevent it.
Do you think it could come from the TL082 for the current measurement ? from the power transistors ?

Something else?

Thank you for your help 

[Geoff-AU]

intermittent faults are the worst.

I have seen some funny business from mine, I think it was something to do with the constant current section being a bit funky when charging a battery.  but it was transient and so is my memory.  I don't tend to use OCP.

[Psi]

LEDs can be problematic. They can become partially damaged inside and then become somewhat random about the current they draw.

You often see the brightness go up/down for no reason.
It's easy to notice visually on a normal 5mm 20mA led but can be less noticeable with higher power leds.


You can also damage an LED by connecting it to a lab psu with a safe current limit set if the voltage is set to high.
The psu may limit the current but the capacitance on the output of the psu can still damage the LED before current regulation takes control.

[aargee]

Check the power supply lines on the control PCB.
Had a weird but different fault and it was a failed electrolytic in the negative supply rail. It turned the comparators biasing to crap. Turned on OK but quickly failed.

Good luck.

[chipeater]

Thank you for your answers 

The problem reappeared last night and it seems that it comes from one relay.
If I apply some pressure on it with my finger or tap it a little with a pen, I hear the contact open and close again several times and the current display swinging.

I ordered two relays and a ULN2003 just in case and will let you know 

[BMK]

I have the 'same' PSU in the 2 channel version, branded tenma. One channel has this exact fault. Normally I resolve it by the trusty kinetic reset procedure. Would be really interested to see your exact fault, it might motivate me to open mine up and actually fix it!

[chipeater]

Was yours purchased recently ?
Mine is not really a Korad either, it's a Velleman I bought a long time ago ( maybe 10 years ).

The reset procedure does not work every time for me but by setting the current limit high enough and drawing like 1 or 2A , I can make it work normally again.
but most of the time, my power supply is set to 5V and I draw less than 100mA and the problem comes back quickly.

I bought those relays : https://fr.farnell.com/finder/36-11-9-012-4011/relais-spdt-250vac-10a/dp/2311439

and I'm going to get them on Monday so I'll know soon if it's really coming from there 

[BMK]

Just looking back at PMs. I bought it 'as-is' on here in 2015 at a really good price from the famous 'toploser'. I think he recycles bulk customer returns, and mine was honestly described as having a bent chassis, buckled around the transformer mounts -likey from pulling too many g's in transit. Developed the fault last year. I put it down to a latent defect caused by its rough travels.
No Idea how long it existed before falling into my hands, but similar mfg date to yours is plausible.

[chipeater]

It was definitely the relays. I don't know which one exactly, I changed both.
If your problem also comes from there, you can choose the same relay as me, the footprint is the same.

Goog luck