I don't think you get it yet. Normally I'd let it go but you're going
to mislead a lot of people here if I do.
The original question involved capacitors in the range of 1000 pf, which
at 1.8 MHz have an impedance less than 100 ohms. That 100 ohms is in
parallel with whatever leakage the dielectric might have. It would take
a pretty defective capacitor for the RF voltage across it in any series
configuration to divide in any way other than per that 100 ohms AC
impedance.
Besides, if the leakage was that bad and it was part of a parallel
combination, Q would be horrible and the capacitor would burn up because
of the excess current. Just the opposite of what you stated.
I'm not in any way suggesting that anyone "just slap together some
capacitors without further thought". Quite the contrary ... I'm
suggesting they simply take into account basic electrical theory when
deciding on a configuration. You should do the same.
Dave AB7E
Bill Turner wrote:
> ORIGINAL MESSAGE:
>
> On Thu, 29 Nov 2007 12:15:26 -0700, David Gilbert
> <xdavid@cis-broadband.com> wrote:
>
>
>> I don't understand that comment. Murphy's Law has nothing to do with it
>> ... Ohms Law does.
>>
>
> ------------ REPLY FOLLOWS ------------
>
> Murphy will be glad to provide you with the occasional capacitor which
> has considerably more leakage than another. In a parallel circuit,
> this may be of no importance, but in a series circuit it may be fatal.
> Likewise, since this box is going to be outdoors, Murphy may cause dew
> to condense more on one cap than the other, again of little or no
> consequence in a parallel arrangement.
>
> These things can be overcome, but the designer must take them into
> account before construction begins. Just slapping a couple of caps in
> series with no further thought is asking for trouble.
>
> 73, Bill W6WRT
>
> _______________________________________________
>
>
>
> _______________________________________________
> TowerTalk mailing list
> TowerTalk@contesting.com
> http://lists.contesting.com/mailman/listinfo/towertalk
>
>
_______________________________________________
_______________________________________________
TowerTalk mailing list
TowerTalk@contesting.com
http://lists.contesting.com/mailman/listinfo/towertalk
|