[Prev][Next][Index][Thread]

Cap Confusion



Originally starting out as a letter to Richard Hull,
I thought the group might want to put their 2 cents worth in,
and or hear about the following.

Richard,

I am specing and pricing out 2 .1uf caps.

Questions for the weekend:

Condenser Products state that their caps that are rated for 20KV AC RMS
are good for a peak DC voltage of 135 KV.(I believe he said 34 mil plate
spacing).

NWL states that for my 20KV AC RMS xfmr they would give me a cap
that they rate as 60KV AC RMS good for a peak DC voltage of 85 KV.

How would you explain these differences?
Which would you go with, based only on this, so far?

NWL eng's will be getting back ahold of me Monday.

Condenser Products eng seemed very familliar with Tesla applications.
They are specing a .1uf cap good for 110 amps rms for 2 min intervals
5 min rest, 500 bps ascyronous gap. Terminals will be 1" copper fittings
due to the temp and current and skin effect he stated that 1/2" brass
would get too hot. (Why did you spec smaller 3/8" terminals for caps
rated at higher currents? or why would their eng spec so large terminals?)

NWL, for a starting point I gave them the same criteria, but told them I
wanted 400 amps rms -at- 150khz. (figureing that if the price was too high, I
could
go down to 110 amps rms -at- 130 khz)
Both NWL eng's seemed very knowledgeable, but did not know a thing about
Tesla coils. (they do recall making your caps) and I explained exactly how a
standard Tesla coil worked)
The currents I can jostle with on my own, but I wanted your input on the
voltage situation, as it is not appearing to be apples to apples.

CP's caps are rated for 125000 amps peak with 100 nH internal inductance.
NWL will spec 70,000 amps peak with an unknown internal inductance.
Is this a difference worth worrining about?

As far as using these caps later on for pulse discharge experiments, is
the value of .1 too small to bother with? CP caps are long as opposed
to NWL's square, so applying that current to a small inductance would be
harder
due to the added inductance of runs from the cap to whatever.

All my figureing is based on using these caps in 4 configureations that will
be used in the future.
1). 80 khz, 2 x .1 balanced, l= 89uhy, zsurge=42, peak amps=673,
rms amps (20%)= 134
2). 80 khz, 1 x .1 unbalanced, l= 43uhy, zsurge=20, peak amps=1415,
rms amps (20%)= 283
3).150 khz, 1 x .1 unbalanced, l=31uhy, zsurge=24, peak amps=1179,
rms amps (20%)= 235 (guess I could put 2 in series & change inductance,
then again I would like to compare configureations myself)
4). 50khz, 2 x .1 (parallel) unbalanced, l=56uhy, zsurge=16, peak amps=1768
rms amps (20%)= 353

CP caps eng figured that a 130khz circuit would draw 110 amps rms (he did
quite a bit of calculating, that at times I must admit was over my head)


What input do you guys have on all of my ramblings so far?

Kevin M. Conkey