The MMC (Multiple Miniature Capacitor) is a concept invented on the Pupman Teslacoil Research List and brought from idea, to prototype, to working concept through the experimentation and research of the dedicated amateur scientists worldwide that make up the Pupman membership. Many of the members of the Pupman list are also Group members and we have much of our early beginnings on the Pupman list.

The idea is simple. Tesla Coil duty is just about the hardest imaginable application for a capacitor. Cap lifetime is a function of rating, voltage, current and number of charges. In the typical Teslacoil application the capacitor can easily be pushed to it’s maximum voltage thousands of times a minute. Overvoltages are commonplace and typically result in the destruction of the capacitor in short order. Because of this, most cap manufacturers will not warranty a cap that is put into coiling service, and the lifetime of a typical coiling cap is measures in hours, often minutes.

This technical problem led many coilers to actually construct their own caps from scratch. Using aluminum foil, roof flashing, large sheets of plastic or even baggies in some cases. This was combined with PVC pipe for a housing, Castor or mineral oil for a dielectric, and lots of caulk and hope. The result was typically a leaking mess that was very heavy, very inefficient, and often died in the first 30 seconds after taking days or weeks to build. Other types of caps have all been tried, glass plates from windowpanes stacked neatly and held gingerly with bolts, used industrial surplus caps with questionable lifetimes and made from dangerous components, some even experimented with hanging large sheets of metal in hopes of making air-cored caps.

Chris Boden has been kicking around the idea of a density capacitor for over a decade now and despite several prototypes, has yet to get a working model above 20kV.

Capacitors are a very difficult thing to make for high voltages and high pulse duty cycle applications, there’s just no easy answer.

The closest thing is the MMC.

The idea is that while it’s very difficult to make large caps in the values needed for a coil, very small caps are cheap and work very well. The simple act of taking the proper capacitors and stringing them together in series to get the voltage you need, then paralleling the strings to get the amperage you need works wonderfully. The load is evenly spread across all of the caps and they are then easily able to withstand the operating conditions that the coiling environment puts them through. The proper caps also have the miraculous ability to withstand overvoltage in a special way.

With most caps, when you hit them with 3x their normal voltage you have just created either a very expensive firecracker or a very expensive doorstop. With the self-healing caps the plates are made from a very thin film. This film reacts differently than a typical plate. When it experiences a dielectric failure (punch-through and carbon tracking) the foil ablates well away from the initial arc site on *both* plates. This opens the circuit around the carbon track and the cap keeps working just fine. The MMC caps we use can tolerate over 250,000 overvoltage events and not only keep working just fine, they stay within 10% of tolerances while doing so.

This simple concept created a revolution in the coiling community.

The Geek Group has a long history with MMC capacitor arrays. We were one of the major pioneers in bringing the MMC caps to the worldwide coiling community and are massive supporters of the concept and technology to this day. All of the Group’s lab and demo Tesla coils use MMC technology and have for several years.

Because we use so many of them we wanted to have a portable MMC array that we can move around the lab and use on a variety of different coils in all of our common sizes. We wanted the array to be robust enough to stand up to daily abuse and moving, small enough to be stored reasonably, and all the while be easy to maintain and adjust from one coil to the next in just a few minutes.

Getting all of these properties in the same device is not unlike trying to herd 500 cats through a tuna cannery.

Of course, since it’s a Geek Group project we had to add another level of complexity. Let’s not only build a laboratory MMC array, let’s build the largest MMC array in the world.

So that’s just what we did.