Re: Capacitor test results
>From ed-at-alumni.caltech.eduSat Jun 8 21:42:53 1996
>Date: Sat, 8 Jun 1996 17:02:45 -0700 (PDT)
>From: "Edward V. Phillips" <ed-at-alumni.caltech.edu>
>Subject: Re: Capacitor test results
>Re: Current from "30 ma" neon transformer
> The rated current is with a RESISTIVE load!
> With a capacitive load the current will be higher, because the
>capacitive reactance cancels part of the inductive reactance
>of the transformer. The approximate reactance of the transformer
>is 12000/.03 = 400,000 ohms. Subtracting 106,000 you get
>294 k. The net current is thus 41 ma, and the voltage
>across the capacitor should reach around 4300. Surely enough for
Ed Phillips, all,
You sound a little miffed. Despite using different mathemagic, and I'm
not saying there is anything wrong with yours, I still maintain that
in the end the result is that Ed Sonderman still can't get enough
power out of his 30 mA (nameplate rated) neon to swing his capacitor
under test to anywhere's near the voltages and dielectric dissipation
that it will see in actual Tesla coil use.
We all know that Tesla service represents the most brutal and abusive
application of a capacitor yet invented by the most extraordinary of
In Tesla service with 15 KVAC RMS applied, the capacitor will see
peaks at 120 times a second of 1.414 times this figure or 21
kilovolts with continuous, recurring symmetrical voltage reversal.
RF ringing peaks, in packets occuring at the break rate superimposed on the
120 PPS wavetrain, and also subject to recurring full power reversal will
approach 2.818 times this or about 42 kilovolts, and this higher voltage is at 100
or more kilohertz. The fact that there are any dielectrics that can stand this
punishment without getting warm (I can count them on about three fingers),
is totally amazing.
So I must ask you, how is testing with 4,300 volts with a gap that won't fire
properly and therefore not developing those more stressful RF ringing voltages
going to give us a meaningful GOOD or BAD assessment of this capacitor as to
its suitability for high powered Tesla service when hooked up to Ed's pole
pig at 14,400 VRMS? You just cannot prove that a racing car engine
will perform at full output for the duration of a gruelling race by
demonstrating how nicely it will idle in the garage!
I stand by my prevous posting!, and I'll even up the ante!
If I may be so bold, I am prepared to go on the record here as saying
that IN MY OPINION THE TESTS IMPLEMENTED BY CONDENSER
PRODUCTS CO. ON THIS CAPACITOR MAY NOT BE ABUSIVE
ENOUGH TO BE REPRESENTATIVE OF TYPICAL TESLA COIL
OPERATING CONDITIONS. After all, Tesla coil capacitors probably
represent about the smallest segment of custom built to order pulse capacitors
in their overall business. They have
undoubtedly over the years developed a series of standard tests which
best mimic worst case field conditions for which their product is
normally employed. Our application is not some one shot per hour
Marx bank in a big government lab, or some similarly wimpy SDI
application, we STRESSSSS capacitors with Tesla Coils! A Tesla
coil is a capacitor destructive tester par excellence! I'm not
saying that Condenser Products Co. doesn't make the highest quality
capacitor, far from it. I have several of their commercial
polypropelene and silicon oil, and teflon and silicon oil pulse
capacitors in my larger coils and they are excellent! These are
however not their economy, Tesla coil market directed type cap. I got mine
surplus from a big government lab when they found they no longer
needed them nearly as often as once per hour. Some of these cost more than
$2K each. Nevertheless, I have no reason to doubt the overall excellence of their
economy caps. I have no experience with them, but I do recognize that even an
occasional Rolls Royce can have problems.
Let us review the statement regarding test parameters which Ed
Sonderman posted on Tuesday, May 21/96.
"Well, I finally got the test results back this morning from
Condenser Products. I feel somewhat foolish for sending them a
perfectly good capacitor. They hipot tested the unit at 35 kv with
rapid charge and discharge. They pulse tested it at 22 kv with 3
pulses per second and then went up to 8 pulses per second. They
tested the internal resistance and the capacitance value -
everything looks great...."
Does anyone else see anything wrong here? Tesla service from 15 KV RMS will
pound this cap at over 40 kV, they applied 22kV at a PRF of 3 PPS and then
increased it to 8. We coilers typically run a cap this big with a
rotary break and therefore subject it to a rate of often 400 PPS to nearly
1000 PPS! On top of this we superimpose RF ringing with wicked reversal at
100,000 PPS or more! I suspect that the capacitor (and its awesomely heavy
burden) in a large Tesla coil deserves just a little more respect than it seems to
There, I've boldly drawn my line in the sand. I thought it was time for a little
reality check. Anyone out there wish to comment?
Slightly resonated, rwstephens