Cooler Master Real Power 450W badcap...

[Brian C]

Last night, when i switched on my pc power supply, i heard a "bang" sound, followed by thick white smoke.
I opened up the power supply. I found 2 bad capacitors. They are panasonic 200v/820uF in the rectification circuit. This is first time I saw panasonic cap burst. I carefully checked the PCB and found that it was the shorted MOV that cause the capacitor burst.
The 2 big 820uF are connected in series to the bridge diode +, -. Each 820uF cap connected to a MOV in parallel. The rectified DC volatge for mains AC230V is around 325V. When 1 of the MOV failed (short circuit), 325Vdc appears across the other 820uF capacitor, causing overload and burst. It is then followed by the other one, as the first burst cap become short circuit.

[starfury1]

Re: Cooler Master Real Power 450W badcap...

interesting...not a nice thing to happen and expensive repair

its not a dead short across the Active Neutral since the caps are in series in a 230V setup, if it follows the norm.
(115V voltage doubler/ 230V fullwave bridge rectifier, caps series)

So i'll have to think on this cause it sounds like the MOV's are as useful as tits on a bull if A Mov goes short when its in bridge mode of operation

I thought it was suppose to take out the mains fuse, seems not enough current was pulled when only one MOV goes short?

Humm thats is a bit of a worry

anyone else got a theory or practical on this?
I am wondering now

[Per Hansson]

Re: Cooler Master Real Power 450W badcap...

No idea, I've seen a few failed MOV's on the primary side of PSU's and each time they have saved the PSU (that is replace MOV and PSU is fully operational again)

[Kaine]

Re: Cooler Master Real Power 450W badcap...

this may sound naive but what do you mean by MOV? That is a spectacular cap failure :O

[Per Hansson]

Re: Cooler Master Real Power 450W badcap...

http://en.wikipedia.org/wiki/Metal_oxide_varistor

The MOV's sole purpose is to protect against such a spectacular failure as you saw there
When you get a overvoltage the MOV sends the excess voltage to ground/neutral

[starfury1]

Re: Cooler Master Real Power 450W badcap...

your not wrong on "spectacular Cap failure" Kaine

Yeah as Per said above on MOV's

The thing is that under the older style of supply were a selector switch is used for 120V/240V operation.

If you place it in the 120V position the circuit is a Voltage Doubler , if in the 240V position its a Fullwave Bridge Rectifier with the caps in series.

(see PeteS post "Rectifier & Filter" Here )

A schematic Here

Snippet from the schematic section of interest.




At this point the whole stage somewhat a conventional linear PSU supplying 340 Volts odd DC as a FW Bridge rectifier setup (@ 240 VAC Rms mains)

BUT note if you look at the full schematic a feed is being taken from the center of the 2 caps ,resistors and MOV's

So you just cant treat in as a single supply rail to the rest of the circuit.

The MOV's are suppose to blow (go short first) if you put a supply on 240V mains while its in 120V position.
This protects the caps by taking out the main fuse.
(this does work as stevo has found out)

In the case above the supply was in the 240V or bridge mode of operation when one MOV when short.(from the senerio outlined)

So it seems that the protection is none existent
when its in this mode of operation
My understanding is mov's will fail from repetitively being hit with over voltage over time (mains spikes and the Mov suppresses it, for example although I would think the EMI filter would cut a lot of this possibility?)

So now the question is in my mind, why didn't it take out the mains fuse?
From what I can gather not enough current is pulled when one fails
(did the inrush current limiter limit too much?)

Is this normal? or not mode of failure
(I dont recall this being a common failure but I don't fix them day in day out either, So those here that do would be best to comment on it)

If its normal then the mov is a "'self destruct" waiting to happen for the psu and depending on how often its forced into action will at some point cause this failure.
(if the above is the case then it would be recommended to replace mov's when serving a psu, is this the case?)

I find this a bit of a strange one, maybe just bad luck

[PCBONEZ]

Re: Cooler Master Real Power 450W badcap...

Just like anything else,,,,
once in a great while a fuse doesn't do what it's supposed to.

Perhaps you suffered a power surge and the fuse didn't blow.

I had a PSU blow (fail) so hard once it took out the main breaker on my house.
-
I was relatively new in that house. It was far from any major city. I eventually figured out the power to the house was so bad (unstable) because of where I was on 'the grid' that even regular light bulbs wouldn't live for more than 2 or 3 months if they were left on.
-
Once I knew that EVERYTHING I cared about got a UPS.
(Including the Printer, Scanner, TV, VCR, and Stereo.)

.