Possible degree of damage to PSU

Question

The following PSU accidentally got 400 V instead of 230 V at J6. There is a fuse built into the power switch, which is connected to J6. The fuse blew on the original incident. Trying to replace the fuse in the hopes that it was just a transient problem resulted in yet another blown fuse. I have tried to check a few components in-circuit, but of course this is of limited value. Before I continue by wildly desoldering components, I wanted to ask the experts (I am doing electronics only occasionally) about which components are likely to be damaged, what else I can check with the least effort, or if there is any hopes of repairing the PSU at all.

Unfortunately the PSU does a little more than just provide bipolar DC supply voltage (like charge and switch a battery), and it is from a chinese device, so just replacing it (e.g. by the same or a similar PSU) is most probably not an option.

My observations so far:

1. no apparently visible burn marks on any of the components, nor any bulged capacitors
2. C1, C2 seem to be intact (they show twice the specified capacity of 1 nF because of the shorted path over D1, see below)
3. bridge rectifier D1 seems to be blown, because it is conducting at 0 Ω on all four diodes in both directions, no threshold voltage. Btw., as opposed to the schematic, the rectifier is implemented by 4 discrete diodes of type 1N4007 (specified for 1000 V), which I had thought to be strong enough to withstand 400 V...
4. the inductor L1 seems to be okay, in as much as it shows around 0.5 Ω through both coils
5. C3 cannot be measured in-circuit because it is already shorted by the (damaged) rectifier D1; naively, I would think that this is the most interesting, because it is specified for 275 V only, but how likely is it that it has blown before the rectifier D1?
6. the resistors in the upper left part of the schematic (up to the rectifier/optocoupler) seem to be okay

So is it likely that there is also damage behind the rectifier D1? Because if yes, there is too much work to do/too high a risk that the repaired state is only transient.

Answer 1

I would worry about semiconductors exposed to the mains voltage:

• D1 rectifier bridge
• Q1 primary switch
• U2 controller chip
• D9 boost diode
• D2, D3 diodes
• The not shown feedback optocoupler
• The not shown TL431 secondary voltage sensor (a part very likely to be used)

You can unsolder all those semiconductors, and then measure the resistors and capacitors in-circuit to ensure that they are OK.

Then replace the semiconductors with new ones. You’ve not shown the full control loop of this supply though: the stuff on the secondary side is just as important! I suggest posting the full PSU schematic, perhaps split into two parts for readability.

Answer 2

Okay, this was try and error, so I don't really know if that qualifies as an answer, but I fixed it.

It turned out after desoldering the 1N4007 diodes, that all of them had gone to semiconductor heaven and are now possibly living a peaceful non-existence as (involuntary) NTC's. I replaced them by an integrated bridge rectifier of proper dimension, which required some drilling on the PCB that luckily had left some free space on it.

The input capacitors turned out to have remained unbroken (I am not sure whether I already had replaced the 33 uF/400V electrolytic C22 some time ago). Of course there were worries that downstream components had gone as well. But I prepared to run for the fire extinguisher and just switched the power supply on: proper low DC voltage at the output terminals. Next plugged the scope to the PSU, and on it went.

Sometimes one needs a piece of luck. And hopefully it stays that way.

It took about an hour, so still well worth the effort (for a 500€ scope, as far as I remember), plus it saved some natural resources.