Replacing 250v caps with 200v ones in an APFC PSU

[yanz]

The PSU i'll talking about is a Tt 480W TWV PSU (cousin to purepower/butterfly series). Yesterday I took this psu from the machine and did some dusting. Before i put this psu in the machine, i tested it with dummy load (bunch of resistors)..

Due to my carelessness not putting back the plastick/film insulator to the bottom of the board, it ended up with "BZZZZZZTT" loud sounds and thick magic white smoke emition... (smells like grandma make cookies).

On autopsy, 2 primary caps (panasonic TS-HA 470uF 250v) exploded and they where very very hot.. It need almost 10 minutes until they (the caps) can be touchable by a bare hand. The leads of those 250V caps was touching the metal case at the bottom. That was what made the short circuit.

I recapped them temporarily with a teapos 680uF 250v caps and a 680uF 200v caps to test if it was a total death. Fortunately it worked. Running a bare system fine.

FYI, 250V caps are hard to find here and oftenly not in proper size (diameter and height).. So i'll be helped if i can use 200v caps.

(And to put thing worst, i lose my two DMM due to my brother stupidity: he measured the voltage of his motorbike coil/inductor at startup and apparently it was too much for the DMM.. So i dont have tools right now to know exactly what voltage is running across those primary caps.)

If i remember correctly, the APFC chip is a UCC3818N..

Now in my hand are 2 NCC KMH 680uF 200V. Will it be ok if i use them for final replacement of those primary caps?

[gonzo0815]

Re: Replacing 250v caps with 200v ones in an APFC PSU

Most APFC circuits need a little higher voltage then the highest line voltage.
this is to make it possibly to draw at any moment the exact same amount of current from the mains as a resistive load would.

This is in most cases near 380v to 400v for 240v rated devices.
The UC 3xxx chip`s data sheet`s app note supports this assumption in that way, that it probably would at least be 385v.

Two 200v caps are a little close, so that there is not much or no safety margin.
I know, that a well branded capacitor probably sustain a higher voltage then it is rated for, but in general it is good design practice to use a significant safety margin.

The best solution would be, to replace bot caps with a single 330uF 450v one.

To make a long story short, you can use your 200v caps, but you are leaving comfort zone fore sure and your result may vary.

[starfury1]

Re: Replacing 250v caps with 200v ones in an APFC PSU

dont know about APFC side of this
but yeah at 240VC and assuming the caps are series it would work but sailing close to the wind which is not really A good idea.
Seems from what Gonzo said even more so in this case with the use of APFC

(no afpc old style supply) 240 AC rms x 1.414(full wave rectifying) = 340 VDC / by 2 = 170 Vdc across each cap (with 2 resistors to force equal voltage sharing across caps)
this of course will depend on how much your Mains supply is liable to vary.

Some 240AC systems here (Well WA) I believe were liable to sit closer to 260V
Australia has now reduced to 230VAC nominal but the plus minus on that still means it could be at 240 plus and be within tolerance

So thats a factor you got to consider...what the local mains is liable to be like

Iam still with Gonzo on that (get 250V or a 450V Single) for safeties sake.

I think with Mains better over kill the possible under kill

Cheers

[gonzo0815]

Re: Replacing 250v caps with 200v ones in an APFC PSU

A APFC circuit is a preregulator. Thus i would say, there is not much overshoot.
The 385 -400v would be a fairly stable value, regardless of input voltage.

Personally i have used 350v caps on a 320v device and some 400v one in a APFC stage of a Medion CRT.
Both units are still alive, but sure i always used Rubycon or similar good brands and there was only one cap, not two in series which does have it`s own constraints too.

[PeteS in CA]

Re: Replacing 250v caps with 200v ones in an APFC PSU

A APFC circuit is a preregulator. Thus i would say, there is not much overshoot.

Active PFC tends to have slow loop response, so overshoot and undershoot in the face of a sudden load change will be greater than for the main loop. So the extra margin of safety of 250V caps would be preferred. Or a single 450V cap.

[severach]

Re: Replacing 250v caps with 200v ones in an APFC PSU

>Due to my carelessness not putting back the plastick/film insulator to the bottom of the board
>The leads of those 250V caps was touching the metal case at the bottom. That was what made the short circuit.

The leads should have been shorter. The plastic film isn't there to keep the leads away when the unit is new. It's to keep the leads away when the unit gets old and the board warps.

>he measured the voltage of his motorbike coil/inductor at startup and apparently it was too much for the DMM

The meter should handle the 12v primary wire just fine. If he tried to measure the spark wire, the meter isn't ready for that. Google "high voltage probe" to see what is required to measure that sort of thing.

[yanz]

Re: Replacing 250v caps with 200v ones in an APFC PSU

The leads should have been shorter. The plastic film isn't there to keep the leads away when the unit is new. It's to keep the leads away when the unit gets old and the board warps.

That's also one thing i forgot to do (clipping thea leads of those 250v primary caps)..

The meter should handle the 12v primary wire just fine. If he tried to measure the spark wire, the meter isn't ready for that. Google "high voltage probe" to see what is required to measure that sort of thing.

now that i think about it again, i believe that was also what my little bro trying to do: measured the spark wire.. no wonder the dmm died...

Active PFC tends to have slow loop response, so overshoot and undershoot in the face of a sudden load change will be greater than for the main loop. So the extra margin of safety of 250V caps would be preferred. Or a single 450V cap.

Two 200v caps are a little close, so that there is not much or no safety margin.
I know, that a well branded capacitor probably sustain a higher voltage then it is rated for, but in general it is good design practice to use a significant safety margin

dont know about APFC side of this
but yeah at 240VC and assuming the caps are series it would work but sailing close to the wind which is not really A good idea.
Seems from what Gonzo said even more so in this case with the use of APFC

thanks for the input, guys.. i take the caution..
I have go with the dual 200v caps and it's been running fine for these past 2 weeks.. nothing bad happen (yet)..

the idea to replace those dual hv caps with a single 450v caps is tickling me, perhaps i'll do it whenever i get another apfc psu in my hands for maintenance... it will be a bit chalenging considering the tight space in that area but it'll be fun to do

[yanz]

Re: Replacing 250v caps with 200v ones in an APFC PSU

This is in most cases near 380v to 400v for 240v rated devices.
The UC 3xxx chip`s data sheet`s app note supports this assumption in that way, that it probably would at least be 385v

i should be more patient and thoroughful in reading the pdf , like what gonzo said, the output is approximately at ~385v. using 400v caps is too close to the limit. but i used a high quality jap brand (ncc) kmh series caps so perhaps they are more tolerance to a rough condition..

and starfury, last time i check here in indonesia (bandung) it was always run at 200-220v but anything can happen so i agree it's better to be safe than sorry..