[Prev][Next][Index][Thread]
Re: Cross-linked caps
On Thu, 25 Feb 1999 11:40:47 -0700, you wrote:
>Original Poster: Gary Lau 25-Feb-1999 0822 <lau-at-hdecad.ENET.dec-dot-com>
>
>> Original Poster: "G.Wilson / B.Petersen" <wilson-at-vossnet.de>
>snip
>> The Tank Cap will be chains of 70nF MKP - 15 in each chain and up to 6
>> chains giving up to 28nF in steps of 4,7nF. (As used with good results
>> by Reinhard Walter Buchner) Should there be any cross linking between
>> the chains ?
>
>I see two sides to this.
>
>1) If they are cross-linked, I believe the voltage across each individual
>cap will be more evenly distributed, as the capacitance variation across
>the individual units will averaged out. If one of the caps did in fact
>have an especially low value and was not cross-linked, it would see a
>disproportionally high voltage across it and be most likely to blow.
I would think it would be a good idea to use a resistor across each
cap to ensure the voltage is evenly distributed over all the caps - if
you're running close to the limit, one cap failing would probably
'unzip' all the rest. Obviously a cross-linked array requires fewer
resistors. Cross-linking should also reduce the effects of
part-to-part capacitance/esr variation, again reducing the risk of a
serial failure due to uneven voltage distribution.
>2) If the strings are not cross linked, this gives one the option to
>experiment with differing total capacitance values, by connecting or
>disconnecting individual strings at their end points. If you do plan on
>doing this, the strings should be placed far enough away from each other
>that they can't arc (since a disconnected string will look like one long
>wire).
>
>My personal choice (I've also just ordered a pile of polypro caps) is to
>cross link.
>
>Gary Lau
>Waltham, MA USA
>
>