GW Instek PSP-405 powers off when turned on its side

[donturner]

I've picked up a GW Instek PSP-405 bench power supply. It works great, except that when I turn it on its side (my bench is very small so sometimes I need to save space) I hear a "click" and the display goes blank and it ceases to have any output.

I've taken the lid off and had a good look inside but can't see anything obviously wrong (no loose wires or ICs coming out of their sockets), however, I don't really know what I'm looking for.

Is this even a fault? If so, any hints on fixing it?

Photos attached.

[donturner]

I've just spotted a whole load of toffee coloured brown gunk under one of the caps - could this be the issue? What is this stuff?

[wraper]

could this be the issue? What is this stuff?

No, adhesive.

[duak]

Try wiggling some of the heavier components and see what happens.  I see a toroidal coil - these are famous for breaking their soldered connections if they were not secured properly during assembly.  Big capacitors also tend to escape.

[cvanc]

Yeah, that's just glue they use to hold larger/heavier parts in place.

Does it go totally dead, no fan, no nothing?  Just like you pulled the power cord?

You need to get the boards out of the case and inspect the back sides.  There's probably a loose connection/cold solder joint, I would guess somewhere around the AC line input circuit.  Hopefully a visual inspection will reveal it.

Good luck!

[donturner]

"Try wiggling some of the heavier components and see what happens."

Thanks I'll try that - any tips for avoiding electrocution whilst doing this?

[johnkenyon]

"Try wiggling some of the heavier components and see what happens."

Thanks I'll try that - any tips for avoiding electrocution whilst doing this?

Prod with something insulating - e.g. a plastic ruler, a 12" length of wooden dowel.

[drknob]

Use a pointed stick..... 
Seriously, 50% of the audio gear I repair suffers from connection/solder failures.

[Gyro]

"Try wiggling some of the heavier components and see what happens."

Thanks I'll try that - any tips for avoiding electrocution whilst doing this?

Lucky you asked!

That's an SMPS and a number of heatsinks and other parts are at mains potential. It is bad advice to suggest this without sensible safety precautions! 

It is almost certainly due to PCB joints flexing with gravity though. Gently push heavy components (wiring looms too) in the direction that gravity will be acting with a plastic rod or similar (as johnkenyon suggests), making sure your hand can't slip. Also check for loose screws.


P.S. This exercise may well involve taking out the power PCB at some point. You will need to ensure that the primary side capacitors are safely discharged before working on it.

[JFJ]

"Try wiggling some of the heavier components and see what happens."

Thanks I'll try that ...

Before getting too involved in looking for a PCB fault, try something simple - swap the mains lead.

Friction between the plastic mating surfaces of IEC mains connectors can sometimes prevent a good connection. Spraying IEC connectors with contact cleaner/lubricant may help, but fitting a different make of mains lead that you can feel goes all the way in, is a better solution.

[ebastler]

Before getting too involved in looking for a PCB fault, try something simple - swap the mains lead.

Friction between the plastic mating surfaces of IEC mains connectors can sometimes prevent a good connection. Spraying IEC connectors with contact cleaner/lubricant may help, but fitting a different make of mains lead that you can feel goes all the way in, is a better solution.

+1. The tolerances on those IEC connectors (mostly the outside dimensions of the plastic body, I think) can be unpleasantly large. And the cheaper cables are often quite stiff and can put drag on the plug when you move the device around.

[Nusa]

Replacing the mains cord is a good idea.

If that's not it, does it matter which side you turn it on, or does it fail on both sides?
If it's only one side, I'd suspect a loose/cracked solder joint as suggested above.

Does it fail with the lid off as well? It should be easier to probe voltages in the failed state to narrow down the fault area. Remember to stay safe while doing so.

[donturner]

Thanks for all the responses. I've spent the last hour with the lid off poking around with a plastic ruler. I was able to reproduce the issue by poking at the large yellow transformers in the middle of the PCB. I took the PCB out and reflowed them along with a few other suspect joints.

I thought that would fix it but sadly the issue still remained when I turned the unit on its side, although I couldn't reproduce it by poking at the transformers any more.

I then put the whole thing back together and tried to reproduce the issue again. Bizarrely that seems to have fixed it! No amount of shaking or turning upside down causes it to click and power off 

I tried different power cables but they all seem fine now. I'm kind of pleased that its working but also a bit disappointed that I didn't find out the root cause. If it happens again I'll start probing voltages whilst the fault is occurring.

[coromonadalix]

Maybe it's the x-formers pins between the assembly and the pcb soldering,   the wires are normally wire wrapped around the xformer base assembly pins, i did have in the past a badly soldered one on the metal pins, took me a long time to find the culprit.

[jeka1981]

Does anyone know how to convert it from 115V to 230V ?

[fzabkar]

Please show us a photo of each side of the PCB in the relevant area.

Are there two 200V bulk electrolytic capacitors on the primary side? If so, then the PSU is probably wired as a voltage doubler. To convert it for 230VAC operation, one would normally just remove a wire or link that connects AC Neutral to the midpoint of the two caps.

See the 230V/115V switch at the top left of this PSU schematic:

http://www.pavouk.org/hw/en_atxps.html

[wraper]

Does anyone know how to convert it from 115V to 230V ?

The same way as old AT/ATX psu with manually switchable input voltage.