The Individual Philips Capacitors:

• Capacitor Make/Model: Philips 376 KP/MMKP, rated for pulse discharge applications
• Internal Construction: Metallized Polypropylene
• End connection type: Metallized foil endplating
• Rated 6.2nF, 2kVDC 650VAC

The Philips 376 KP/MMKP Capacitors:

My MMC Capacitor:

• Rated 6kVAC, 20kVDC (very underrated for this application, but still works great)
• 17 strings, 10 capacitors per string
• Capacitance of one string (calculated): .62nF
• Total measured capacitance (new): .010663uF
• Measured capacitance after ~ 2.5 hours of runtime: .01uF ± 5%
• 9 strings on the top "card", 8 strings on the lower card.
• Each string is soldered to 1" wide brass bars for low ESL/ESR (high Q)
• No bleeder resistors are used
• Used with a 15kV 60mA Neon Sign Transformer to achieve 60Hz resonant charging

Frontal View:

Left Side View:

Bottom view of one of the "cards":

Closeup of the top:

Connections between individual capacitors are made with #22 bare copper wire, wrapped twice around each lead and soldered. Capacitors are mounted on the cheapest of perfboard, and the unit has two small "L" brackets on the bottom for mounting. (seen in the frontal view)

Thoughts On 60Hz Resonant Charging, and That Theoretically "Perfect" Tank Capacitance Value

My MMC capacitor measured out to be .010663uF when it was brand new. Calculated values come out to .0106uF, so it's definately there! I chose this value in order to take advantage of the resonant charging effect that comes into play when the charging transformer and capacitor impedance values are equal at 60Hz. These values are calculated using the following formulae:

Transformer Impedance:

Z = E / I

Capacitor Impedance:

Z = 1 / ( 2 * pi * F * C )

Where:

Z is the impedance in Ohms

E is the transformer output voltage in Volts

I is the transformer output current in Amperes

F is the mains frequency in Hz (60Hz in the United States)

C is the capacitance in Farads

When the impedance values are equal, the two components transfer power very efficiently because they are resonant. Theoretically, power transfer is maximum at resonance. When capacitance values other than the resonant value are used, the capacitor is deemed to be "out of sync" with the incoming AC waveform from the transformer. This doesn't mean that power is lost in the transfer, it only means that the power is not being utilized in the most efficient way.

Unfortunately, a theoretically "perfect" capacitance value cannot be chosen based merely on this principle. Many other factors come into play when the tank capacitance value is changed. The Lp/Ls (primary inductance/secondary inductance) ratio is changed due to tuning corrections. I^2R losses in the primary coil change. Spark gap firing rates change drastically.

It's difficult to tell what value to use when choosing a capacitor, but going with the resonant value is neat for several reasons. For one, having the spark gap fire at 10% of full input voltage has a unique feel to it. Instead of slowly cranking up the variac and having the gap roar into life when the variac reaches 60 or more percent, the resonant capacitor value causes the gap to fire at a very low rate starting at 10% of the full input voltage! That's around TEN or so volts when you're putting 120V in. Also, with capacitor values that are much larger than the resonant value, you don't have much control over things. You can be halfway up on the variac and things still not happen, but then you give it another 1/8th turn and your coil roars into life at 60%. I prefer to have things start up slowly at 10%.

Other's have reported excellent coil performance with values approaching twice that of the resonant value. Since coil performance is what we're after, I'd go with something in that ballpark. You lose some input power control over the coil, but performance is gained. Sounds good to me.

Back to the main page