Posted July 28, 2025

APELC’s basis for a Marx Generator Design – Part 1 Using the doorknob-style capacitor.

Too often, we find that our potential customers look at our website, then approach us with an inquiry of a specific Marx generator model. And just as often, we find that some of the desired specs came “close” but ultimately are a bad choice. This makes me think that we’ve done a poor job with how our website communicates what we do and what we have to offer. We are working on this and hope to soon unveil a new and improved site that better serves our customer base. Until then, I’m hoping that some blogging will help, or at least be interesting to some. This is part 1.

Customers bring specs, which can be inconsistent, vague, and sometimes, crazy. But we plow through it, as best as possible. Ultimately, we must find a suitable geometry that’s not too crazy.

APELC designs Marx generators around three types of capacitors, including the doorknob, mica film, and the more traditional dielectric/foil capacitors similar to those from General Atomics.

This blog focuses on the doorknob-style capacitor.

The doorknob capacitors have been APELC’s tried and true, go-to capacitor since our start. Many of the generators listed in our portfolio use these. A few pros and cons:

(Pro) They are easy to work with and they’re robust. We often reverse these capacitors to 100% of their rating; and somehow, they keep working, and sometimes they don’t.
(Pro) Their relatively small geometry is easy to design around, meaning that we can often pack-in several capacitors into a small volume. This point caters well to APELC’s parallel-switched geometries, where we incorporate several parallel spark gap switches on each stage to reduce the generator’s series inductance. An example generator includes the MG20-22C-2000PF.
(Pro) They can be modified, if you dare. In early years, we used to shave down their length, removing excess epoxy and electrode, and sometime reducing the lengths by as much as 30%. It is a horrible way to treat a capacitor; but it can work.
(Pro) The capacitors are simply well engineered and manufactured. When I was in grad school, I did some interesting testing on ceramic doorknob capacitors, finding trends between the physical compression and failure. Maybe we will someday revisit that testing
(Con) The doorknob capacitors suffer from a horrible voltage coefficient. As the charge voltage approaches the capacitor’s voltage rating, its capacitance begins to decrease. At its voltage rating, the capacitance will have dropped by as much as 8%. So, if you are designing to a hard specification, such as pulse width, this trait should be considered.
(Con) Unfortunately, their energy density is great. As a result, they are best suited for applications requiring a few Joules to hundreds of Joules. Example APELC generators include our MG15-3C-940PF and MG16-3C-2100PF.
(Con) One problem we do run into with these capacitors, is when we want to stuff in a large number of capacitors, resulting in a series connection of several capacitors. The problem? The geometry starts appearing as a Pulse Forming Network (or PFN). The problem with the PFN affect is that it can surprise you with square-like output waveforms. Moreover, the impedances of those stacked PFN’s add, and too often to a higher than expected impedance. That’s when the “oops” comes into the conversation.
(Pro) Countering the last point, designing the multi-capacitor geometry can lead to some nice pulse-shaping, as we did with our MG24-20C-2700PF and MG18-9C-2600PF Marx generators.

We really like using the doorknob-style capacitor for its ruggedness and design flexibility. In the next post, we will explore the use of the mica film capacitor.

————————-

About the Author
Jon Mayes is the Founder and President of APELC. With over 25 years of experience in pulsed power system design and a Ph.D. in Electrical Engineering, Jon has led the development of industry-leading Marx generators, EMP simulators, and high-voltage test systems. His work has supported the Department of Defense, Department of Energy, and major research institutions across the U.S.