Stupid PSU design FAIL

[dan8139]

Got this lovely example of PSU quality in a PC after there was a bang and a smell... It's a CASECOM ATX 400W. When I saw the leaking primary cap I was very surprised...

Until I realised they are only 200V caps! We're in the UK on 240V.

Annoyingly this thing also took out the PC's motherboard :-(

Wonder how many more of these things there are out there!

ETA: I just wondered whether it was me being dumb and there's some sort of step-down transformer or something in there but I don't think so.

[mariushm]

Re: Stupid PSU design FAIL

There's TWO capacitors, the voltage is split between them.

The AC voltage is rectified, so it becomes 1.414 x 240 = 340v so each capacitor will get about 160-180v

It's not those capacitors that killed the board (though they really didn't help), it's the small capacitors on the secondary side and the lack of filtering (pi filters aka inductors and capacitors) that killed the board.

The psu is a 250-300w at best, so no wonder it can't handle modern systems that need a lot of power on 12v.

[dan8139]

Re: Stupid PSU design FAIL

Should be thread FAIL then ;-)

You learn something everyday... wasn't aware caps were used in series.

Will look elsewhere for the problem - cheers!

[Adr3nl]

Re: Stupid PSU design FAIL

Why did u multiply the AC Voltage input by 1.414? I understand it has something to do with the rectification process, but im curious =p

[mariushm]

Re: Stupid PSU design FAIL

The voltage at your mains is a RMS voltage.

So 240v actually means there's a sine wave that goes up to about 340v above 0 and 340v below 0

The peak voltage is sqrt(2) x Vrms or in other words Vrms = 0.707 Vpeak.

See : http://www.electronics-tutorials.ws/...s-voltage.html

sidenote: things change a bit with ac power that's not quite sinosoidal, for example when using some UPS that uses simulated sine wave.. the peak voltage will be about the same, but the power supply will use a different amount of current compared with pure sine wave input.

[cheapie]

Re: Stupid PSU design FAIL

Why did u multiply the AC Voltage input by 1.414? I understand it has something to do with the rectification process, but im curious =p

AC voltage is measured with RMS, which is some sort of complicated math-y thing. The peak-to-peak voltage is higher (bottom of the cycle to the top), and this is what you get when rectifying.

See here.

[ben7]

Re: Stupid PSU design FAIL

The balancing resistor across that capacitor probably went open circuit, causing an over-voltage on the capacitor.

Anyways, I doubt they are 820uF!