Testing of the Geek Group Caps

Results of testing one of the Geek Groups MMC caps on 5-19-2001: by Terry Fritz

Cap type:

Cornell Dubilier

942C20P15K    0.15MFD (uF) +-10%    2000VDC

Measured value = 0.1504uF

Length 1.734  Diameter 1.043    Radial

Lead diameter 0.0464"    Lead length 1.67"

Weight 27.9 grams


The first test was the RMS current handling capability of the cap at 350kHz.

The equipment used was a Pearson 4100 current monitor, TEK 3012 scope, AE PDX 900, Bertan XR High voltage supply, and a few other toys.

Test Set Up

Power Supply and Scope

Scope waveform

The cap heated 2.6 degrees C over ambient after 15 minutes with 3.5 amps at 350kHz of current going through it.  Since the maximum the caps should be operated at for a short time (hours) is 15 degrees C and the heating is an I^2 function.  The maximum RMS current rating for these caps works out to SQRT(15 / 2.6) x 3.5 = 8.4 amps.  Thus you can safely put about 8 amps per string through these caps at Tesla coil frequencies.

A high voltage test found that the caps arc internally at 5700 volts.  The stored energy at that point is 1/2CV^2 = 2.4 joules and the caps make quite in internal bang :-))  However, they seem undamaged despite the very high energy.  These caps play a little trick to be able to dissipate this energy which I won't mention incase it is a secret.  Since the energy is so high, I would avoid repeated over voltage hits to these caps since they are not as likely to recover as the lower stored energy types.  However, with a breakdown at 5700 volts through their heavy 1 mil dielectric, this should never be a problem.

The internal foil was found to be of typical construction of high quality.  The end cap and lead attach is very strong.  The leads are spiraled at the ends to give high strength and good contact area.

So they look great! 8 amps RMS, 2000 Volts, at 0.15uF for only $2.25 each!!  The thermal dissipation value for these caps the MMCalc asks for should be "22".  A very low value! :-))