RE: Capacitor in series or parallel?

["Tesla list" <tesla@xxxxxxxxxx>]

Original poster: "Lau, Gary" <Gary.Lau@xxxxxx>

I think both you guys are right.  With the xfmr across the cap, the xfmr
will see a lot of RF at the tank frequency, and with the xfmr across the
gap, the xfmr will see HF-rich step functions.  Also, during the
zero-current crossings when the gap briefly turns off, there will be VHF
transients at the self-resonant frequency of the primary and other
paracitics, peaking at TWICE the tank capacitor voltage.

But the important thing is that while R-C filters don't do a great job
at filtering the main tank frequency (i.e. they won't help much if the
xfmr is across the cap), an R-C filter will do an excellent job
attenuating the VHF transients and step functions seen across the gap.
When using an R-C filter, running with the xfmr across the gap is
better, as both hazards are eliminated.  If you're not using an R-C
filter, choose your poison.  I couldn't say which is worse.

Regards, Gary Lau
MA, USA


>>> The usual answer to this is that you connect the power
>>> supply across the spark gap, in order to minimize the
>>> amount of RF fed back into the transformer. This is a
>>> moot point though: spark gaps generate RF too.
>>>
>>> Steve Conner

>> Hi Steve,
>>
>> I must disagree; it is not a moot point. the RF voltage seen by the
>> power transformer is several orders of magnitude less with the gap
>> across the transformer. With transformer across the gap, the
>> transformer sees only the RF voltage drop across the gap WHILE the
>> gap is firing, essentially a few hundred volts instead of ~20+ KV.
>> IMHO, a mouse and a moose are both mammals but not equivalent if
>> they're stepping on you, and it is a disservice to a newbie to
>> intimate that they are quantitatively similar..
>>
>> Matt D.


>It also sees a "20+kV" step function when the gap
>fires. I can imagine scenarios where that would excite
>ringing in the transformer windings and be more
>harmful than the relatively pure sine wave you get
>across the capacitor. Tesla tried it both ways and
>never seemed to come to a conclusion.
>
>Steve Conner