MMC Capacitor

My latest MMC uses Cornell-Dubilier's 942C20P15K-F 
arranged in 5 trays (one not shown) each tray giving
25nF capacity, rated to 36kV. This gives a total of 54
caps per tray for a grand total of 270 caps for the 5 trays.

High voltage capacitors that can survive the grueling
conditions of a tesla coil are hard to find, and therefore
inevitably expensive if you do manage to locate some.
Making your own rolled poly capacitor is another option,
but this option rarely tends to be successful for very long.

Nikola Tesla (1856 - 1943) used salt-water capacitors which consisted of wine bottles with a brine solution in them. This option will work and it can also be very cheap if your short of funds. However they are not very efficient. Glass has very high losses when it comes to RF frequencies, so your coil's power output will be down. It is also difficult to estimate the value of capacitance that you have made, important when it comes to tuning, unless you have a suitable meter capable of measuring the capacitance.

The option that a lot of people use is to create a MMC or Multi-Mini Capacitors. These consist of quite a few lower voltage rated capacitors joined in a series/parallel combination. This means that 10 individual capacitors each rated at 2000v, could now handle 20,000v being applied across their series string of 10. The drawback is that capacitors behave the opposite to resistors in respect of their overall value when connected in series.
This means that once you have enough capacitors in a string to meet the voltage requirement, you then need to keep adding identical strings in parallel to bring the capacitance value back up to what you desire. This is illustrated below:

Size	0.015uF  15nF  15000pF
Voltage DC rating	2000V
Voltage AC rating	500V
Part Number	942C20p15K
Length Inches	1.161
Diameter Inches	1.811
Lead Diameter Inches	0.047
dv/dt	2879
Ipeak	432
Irms 70 deg C & 100kHz	13.5 A

CD942 capacitors
The individual capacitors that I use are the very robust and popular 942C20P15K-F made by Cornell-Dubilier. These are rated at 2000v each and have a value of 0.15uF, and are of the so called metal-foil capacitors which are preferred over the metal-film type.
The difference is mainly to do with the connection onto the end of the plates, and it has been found the latter type can suffer from failure, unless the peak current is kept within the capacitor's specifications. Currently I use 125nF and power from a Pole Distribution Transformer or 'PIG' as they are often called, but these caps are just as useful for a single NST set up, although with a much reduced overall value.
These can be difficult to source outside of the USA however, and you may need to import them. In the China Huasing Electronics provided the majority of my own.






A clumsy accident occurred in which a piece of scrap wire dropped across the MMC,
resulting in this one single cap (a Cornell- Dubilier 942C20P15K) ending up being
connected across a 12Kv 'Pig'.
The centre immediately blew out and was found to be completely carbonised.
The lesson learnt is not to put 12Kv across a 2Kv capacitor, and more importantly,
also make sure no pieces of scrap wire are left unaccounted for!



blown Tesla coil capacitor

Tesla coil capacitor 10 foot strike

   
Circled above Another way of mounting many capacitors in a small space!

Before I was using the five trays shown above, I had a 72nF MMC that needed another 30nF, but with no room to do so, hence the circular MMC.
The reason for the increase to 100nF, was Richie Burnett's: Resonant Charging advice, see also More Detail in a TCML Post
Richie Burnett's resonant charging ideas relate to the fact that with a synchronous RSG, rather than the resonant frequency of the charging system being solely dependent on the mains frequency and ballast value, it is also affected by the SRSG's bps rate.
I have two pages on this site that deal with this subject:
Resonant Charging (1) & Resonant Charging (2)

Based on Richie's advice I therefore needed another four strings of capacitors bringing the total number being used to 216 (12 strings of 18 in series. The circular arrangement saved space and avoided having to rebuild the whole coil assembly where the existing capacitors are presently mounted.
There are 76 capacitors in this MMC occupying a floor space of only eight inches diameter. The easiest way to assemble the strings when using this method is to construct them externally, and then put them in once assembled. They were then held in place by cable ties passing through holes drilled in the exterior. The finished auxiliary MMC measured 33.8nF (expected value was 33.3nF).
The replacement of a single capacitor merely involves cutting all the cable ties that secure that column, and desoldering the string top and bottom. You can then easily work on the removed string to replace the faulty item
The four strings each consist of eighteen in series. Each vertical column of capacitors consists of nine units with a link across the bottom to its corresponding partner on the other side.
The coil to MMC connections are the two nuts that can be seen at the top at the 4'o'clock and 8'o'clock positions.