[Test-Equipment] Question about series/parallel capacitors

[Mike Manes]

Putting caps in series across a DC voltage greater than the rated cap voltage
risks unequal shunt leakage causing an unequal distribution of the DC,
possibly exceeding the lowest-leakage cap's rated voltage and blowing it to
a short.  The the remaining caps see even higher voltage than before, so by 'n
by, the whole string shorts.

One can work around that by shunting the caps with resistors that leak 10x
as much as the normal caps do at their rated voltages, but then, you end up
with a much leakier equivalent cap.  That works fine where the excessive
leakage current is no big deal, e.g., in rectifier filter application.

Otherwise, you're better off just investing in a properly-rated cap:

http://www.jameco.com/webapp/wcs/stores/servlet/ProductDisplay?langId=-1&productId=1983424&catalogId=10001&freeText=1983424&app.products.maxperpage=15&storeId=10001&search_type=all&ddkey=http:StoreCatalogDrillDownView

That's a 0.1 uF 630V mylar cap, 29 cents ea.  Put 5 in parallel.

But since the min order qty is 100, you'll have plenty of spares :=(.

73 de Mike W5VSI

On 5/13/2011 16:59, Barry wrote:
> I need a 0.5uF capacitor that's capable of handling 600VDC.  I have several 0.5uF that are rated at 250V so I plan to connect them in series/parallel to get 0.5uF at 750V.
>
> The question I have is does it matter how these are connected?
>
> I have the choice of:
>
> A. Connecting three sets of three capacitors each in parallel to produce 1.5uF and then connecting the three sets in series yielding 0.5uF.
>
> or
>
> B. Connecting three sets of capacitors in series and then connecting the resulting 1/6uF capacitors in parallel to again make 0.5uF.
>
>
> Both methods will give me 0.5uF but it I'm wondering if one way has any advantages.
>
> Any thoughts?
>
> Thanks,
> Barry - N4BUQ
>
>
>
>
> ______________________________________________________________
> Test-Equipment mailing list
> Home: http://mailman.qth.net/mailman/listinfo/test-equipment
> Help: http://mailman.qth.net/mmfaq.htm
> Post: mailto:Test-Equipment at mailman.qth.net
>
> This list hosted by: http://www.qsl.net
> Please help support this email list: http://www.qsl.net/donate.html
>

-- 
Mike Manes    mrmanes at gmail.com     Tel: 303-979-4899
"Things should be made as simple as possible, but not more so."
A. Einstein