Re: Commercial cap specs for tesla cap

Hi Jeff!

NWL (formerly Nothelfer Winding Laboratories) is a very respected name
among pulse capacitor vendors. And, NWL was the vendor of choice by
Richard Hull when he was actively coiling. They also make transformers,
reactors, and HV power supplies. If you spec your cap properly, they
WILL make you a reliable Tesla Coil cap. However, you may suffer a bit
of sticker shock... :^)

Since you'll be running off a 14.4 kv distribution transformer, and
assuming you're using inductive ballast, the peak voltage applied to
your cap can be significantly higher than expected using just an
RMS-to-peak calculation (14.4*1.414 = 20 kV). You'll also need to make
further allowances (17% more) if you plan to "overdrive" your
transformer (say to 280 volts) from your variac. Running the numbers,
Vpred = 14.4*1.414*1.17 = 23.8 kv predicted peak voltage.

However, interactions between your tank cap and LV-side ballast
inductance can further increase the peak voltage on your cap, especially
if your rotary gap misfires. For planning purposes I'd apply a minimum
1.5x - 2x "safety factor" to the above voltage, leading to a minimum
actual peak voltage rating of about 36 to 48 kv - I'd spec a 50 kv pulse

It gets worse(!) Once the gap fires, the cap then sees a voltage
reversal of as much as 80%. This means that the cap's dielectric system
actually "sees" a worst-case voltage stress equivalent to about 1.8X the
peak applied voltage. The dielectric system must be designed as though
it was handling about 90 kV. Fortunately, your capacitor vendor takes
this into account when designing the cap for oscillatory discharge use
when you tell him the expected % voltage reversal.

The peak current is really a function of your planned peak voltage, tank
capacitance and primary inductance. However, assuming you use a moderate
amount of primary inductance (good design practice, and easier on your
rotary gap's electrodes), you should be able to keep Ipeak in the range
of 500-750 amps. I'd rate it for a conservative 1000 amps. The duty
cycle will be a fuction of your breakrate, coupling coefficient,
operating frequency, and how well your rotary is quenching. Assume a
worst case of about 2%. If you use an RMS current of 20 amps, this
should keep you out of trouble. Let's also assume that a system of your
size will be tuned to operate at no more than 100 kHz - we'll use this
as the worst case ringing frequency value. 

Let them know that this cap is intended for disruptive, damped wave
discharge application - as such, pulse repetition frequency is also an
important parameter. Provide them with the typical and max values you'll
intend to hit. You might want to assume 120 BPS (synchrounous rotary)
and 420 BPS for worst-case asynchronous operation. You can spec a higher
PPS rate if you really want to live dangerously (and expensively)...
:^). The design life for pulse caps is typically rated in "shots", with
typical being about 100 million. This translates to about 116 hours at
240 BPS - not too bad! Ask them how much more a capacitor rated for for
10^9 shots would be if you're concerned about expected life, or if you
plan to run at higher break-rates. Many of the other requirements are
less unimportant - ask for the lowest-cost option. 

Putting this all together now:

  rated voltage     = 50 kV
  peak current      = 1000 A
  rms current       = 20 A 
  duty cycle        = 2%
  voltage reversal  = 80%
  ringing frequency = 100 kHz
  op. temperature   = room temperature (25 degrees C)
  design life       = 10^8 shots
  case style        = polypropylene case (usually the lowest-cost)  
  terminations      = Studs preferred (1/4" or 5/`16" typical)
  repetition rate   = 120 PPS min, 480 PPS max, 240 PPS typical

Try to see if you can come close to a standard design that they already
make - custom designs cost big money! Also, ask them if they are
currently manufacturing anything that comes close to your requirements -
excimer laser and flash-lamp pulse caps are often usable as Tesla Coil
caps. They are often made in significant quantities and capacitor
vendors sometimes have over-runs that can be picked up at a discount.
Who knows, you may luck out... :^)

Hope this helps!

-- Bert --
Bert Hickman
Stoneridge Engineering

Tesla List wrote:
> Original Poster: "Jeffery Pederson" <kd4lyh-at-webtv-dot-net>
> Hello fellow coilers,
> I contacted NWL capacitors for a cap for a tesla coil application. They
> sent me an email that I need to fill out ther cap form. If anyone out
> ther could help me give them the specs they need it would be
> appreciated. this is what they want to know: tolerance
>           rated voltage
>           peak current
>           rms current
>           duty cycle
>           op. temperature
>           voltage reversal
>           ringing frequency
>           maximum Ind.
>           design life
>           case style
>           terminations
> Im using a 14.4KV/5kva piggy running up to 6-7KVA. They said they
> couldnt build me a cap without all these specs. Im not an engineer so
> Ill need some help.
> Best Regards Jeff KD4LYH