Good PSU repair shop in the UK?

[Wuerstchenhund]

Hi All!

I have a switch-mode power supply (Artesyn/Computer Products XL130-3630E) that needs repair as one of the voltage rails produces a too low outout (0.7v-1.7v instead of the 3.5. expected; all other voltages are fine). All caps have been replaced already, as have shunts and most transistors. There is no obvious fault (i.e. burn damage on components), and all the minor parts like diodes and resistors haven't been checked, so the problem could easily be something simple.

The thing is that switch-mode PSUs are not my strongest point (and this PSU is quite complex) and that due to other, more pressing stuff I simply also don't have the time to search for the fault myself. In addition, my home lab is still sparsely equipped with spare components which doesn't help either.

I don't have the schematics as Emerson Embedded Power (who now owns Artesyn) won't give out schematics to this 19 year old PSU because of 'IP protectsion' (it's not as if the schematics couldn't be derived from the PSU itself, given enough time, but here we go).

So I am now looking for a shop in the UK (or if there's none I would also take one in another EU country) that can fix such a PSU for a reasonable amount of money. Google found me a few shops but they are working with businesses only and generally charge in excess of £200 for repairs.

So I was hoping that someone might know a shop in the UK who can fix a somewhat complex switch-mode PSU for a reasonable price.

[Rerouter]

the price you have been seeing sounds about right on the money, when you consider the hours labor required to troubleshoot and repair a circuit they have likely never seen before, nor one that a schematic can be sourced for, in fact in aus it may be considered cheap,

[darrylp]

If you don't feel upto fixing it yourself. Then why not add circuit to give you a new -3.5 output.
Not much more than a lm337 and a few caps and resistor will get you the voltage you need. Surely that sat on a stripboard will fit inside ?

Darryl

[Wuerstchenhund]

the price you have been seeing sounds about right on the money, when you consider the hours labor required to troubleshoot and repair a circuit they have likely never seen before, nor one that a schematic can be sourced for, in fact in aus it may be considered cheap,

Well, maybe, but then I would assume that a shop that specializes on fixing PSUs should have seen most of the ways a PSU can be designed, and the fact that certain components have been replaced should at least in theory limit the components that need to be checked to find the fault.

I would think for someone with experience in power electronics and who has access to the right tools and spare parts it probably won't take much time to find out what's wrong.

[Wuerstchenhund]

If you don't feel upto fixing it yourself. Then why not add circuit to give you a new -3.5 output.
Not much more than a lm337 and a few caps and resistor will get you the voltage you need. Surely that sat on a stripboard will fit inside ?

I thought about that but there are various reasons why I this is no alternative, and one of them is space.

[AndyC_772]

If you asked me to quote for repairing a 19 year old PSU, I'd quote you a lot more than £200. If you then told me you'd 'had a go' at repairing it, I'd take whatever figure I'd just quoted and double it.

A product which has failed in service may be straightforward enough to repair, given schematics and a parts list. With experience, a technician should be able to work out which parts are likely to have inherently finite lives, or be under stresses that may cause them to fail.

Without the schematics it's a much harder job, as you need to try and reverse engineer the design before having a clue how it's supposed to behave. Alternatively just replace any parts that look likely candidates, such as electrolytic caps or power transistors, and hope for the best.

But if someone's already 'had a go'? Then you're in a whole world of hurt. Not only do you know the original fault wasn't trivial, you now need to deal with the fact that any of the components on the board might be incorrect, fitted backwards, poorly soldered, or have unrepaired damage underneath.

A repair technician is likely to have good practical soldering skills, can probably read a schematic, and will have experience with certain popular models of whatever it is he works on, but he's unlikely to be a genius power supply expert who can simply look at a board and spot the fault straight away.

[Wuerstchenhund]

If you asked me to quote for repairing a 19 year old PSU, I'd quote you a lot more than £200. If you then told me you'd 'had a go' at repairing it, I'd take whatever figure I'd just quoted and double it.

A product which has failed in service may be straightforward enough to repair, given schematics and a parts list. With experience, a technician should be able to work out which parts are likely to have inherently finite lives, or be under stresses that may cause them to fail.

Without the schematics it's a much harder job, as you need to try and reverse engineer the design before having a clue how it's supposed to behave. Alternatively just replace any parts that look likely candidates, such as electrolytic caps or power transistors, and hope for the best.

But if someone's already 'had a go'? Then you're in a whole world of hurt. Not only do you know the original fault wasn't trivial, you now need to deal with the fact that any of the components on the board might be incorrect, fitted backwards, poorly soldered, or have unrepaired damage underneath.

You seem to think that some idiot with no clue had a stab on it, but quite frankly after two decades as RF engineer I do believe that I can solder correctly, and I usually make sure I put the stuff in the right way, thanks. And the parts that have been replaced were replaced with identical types, except the low ESR caps where the new ones are rated for higher voltages and in some cases also temperatures. The PCB has also been checked and seems fine, at least there is no damage on the outside tracks.

In addition, the problem with one of several rails producing a too low voltage did already exist before the parts had been replaced, and hasn't changed since.

A repair technician is likely to have good practical soldering skills, can probably read a schematic, and will have experience with certain popular models of whatever it is he works on, but he's unlikely to be a genius power supply expert who can simply look at a board and spot the fault straight away.

I don't expect anyone to spot the fault straight away. But I believe that someone who has more experience in power electronics than I do may well be familiar with the specific design, and if it's a professional shop may even have the schematics (as this PSU has been used in various applications, not just one, and at least many PSU repair shops in the US list that type as one they have experience with). In my field of work I occasionally get across devices where schematics are not readily available, so I know how difficult reverse engineering can be. But again, there are certain things that can be done in just a few ways, and therefore I usually recognize quickly. That should not be much different in other areas of electronics.

I'm sure that if I put enough time in it I will find the fault sooner or later. But the reason I'd rather have it repaired is that I simply don't have the time, and while at work we have advanced labs with all tools I could ever want unfortunately my home lab is still really very basic (two DSOs, a simple PSU, two DMMs and a crap Chinese solder station) and cramped.

But I guess it will have to wait until I find the time then.

[Jonny]

one of the voltage rails produces a too low outout

So the other rails are ok, which means the primary side of things (assuming there's only one switching transformer) is working as it should. So the fault will be in the secondary side in which I'd imagine one of the transformer winding's (if there's a separate one for this rail) will go into a diode, then a capacitor, then some component or combination of components for regulating down to the required voltage. If not a separate winding then a feed from a higher rail into the regulator.
You have a lab, so you have a multimeter, so you can check diodes (especially zener ones, often made of glass or tell by the symbol on pcb ) make sure they're not shorted, make sure you lift one leg or remove it completely so you get an accurate reading to.
Try and draw the circuit tracing back from the output wire, I can't imagine there will be a lot before you get to the primary line that feeds it and you'll learn so much! I did that with a faulty inverter (12V to 240V unit) after staring blankly at the circuit for a while and the answer basically jumped off the page at me. Was well worth the time.

Good luck!

[Wuerstchenhund]

Try and draw the circuit tracing back from the output wire, I can't imagine there will be a lot before you get to the primary line that feeds it and you'll learn so much! I did that with a faulty inverter (12V to 240V unit) after staring blankly at the circuit for a while and the answer basically jumped off the page at me. Was well worth the time.

Good luck!

Thanks, I guess I will leave it until I find the time to do it myself, and you're right it may be worth just for learning a bit more about power electronics. The learning curve might be a bit steeper as this thing is apparently a bit more complicated than most of the other switch-mode PSUs I have seen in commercial stuff.

[AndyC_772]

You seem to think that some idiot with no clue had a stab on it

Not at all, though I see how what I wrote could be interpreted that way. My point was simply that a technician given a faulty PSU that someone's tried and failed to fix doesn't know that the attempted repair was a competent one, and any rework is (as I'm sure you'd agree) something to be suspicious of. Unless I personally knew the person who attempted the repair, my heart would sink if anyone handed me a product that someone had already tried to fix.

[Pentium100]

Thanks, I guess I will leave it until I find the time to do it myself, and you're right it may be worth just for learning a bit more about power electronics. The learning curve might be a bit steeper as this thing is apparently a bit more complicated than most of the other switch-mode PSUs I have seen in commercial stuff.

Yes, things like the common ATX power supplies are quite simple, but the server power supplies (or whatever yours is can be complicated).

I was repairing a server power supply and it took me a while just to figure out how to make it turn on (as I did not have the backplane to which it was supposed to be connected), then it would work for a few seconds and start producing much lower voltage under load than rated (with a small load the voltage was fine). Turned out to be a leaky ceramic cap in the power on reset circuit - it leaked enough to cause the PFC controller to reset (and shut down) after a few seconds, resulting in a too low primary voltage.

[EngineeringTech2013]

Our company's been using a company in the UK called ' A1 Power Engineering (UK)'. Very good service and don't charge to evaluate a product - though they are only business to business and won't do work for private individuals I think. We have Lambda, Artesyn and Weir power supplies serviced by these guys and they've not let us down yet.

[Rufus]

I have a switch-mode power supply (Artesyn/Computer Products XL130-3630E) that needs repair

http://www.advanceproductservices.co.uk/
http://www.qesltd.co.uk/

Both list that part as something they can repair.

I presume it is for some HP equipment and has an HP part number of 0950-2369.
Someone in china is listing pulls from working equipment at US$500.

edit: should have said listing on ebay

[fluxcapacitor]

http://groups.yahoo.com/neo/groups/hp_agilent_equipment/conversations/topics/49935

[nctnico]

What if you posted some pictures?

I agree with Andy. If someone already had a go then chances are something got damaged or made worse. I doubt 200 pounds will cover the repair. I have repaired lots of switching power supplies but I'd probably have to charge a multitude of 200 pounds unless it is something simple.