This is a diagram of my speculative (because I haven't built it yet) choke output DC power supply based on microwave oven transformers (MOTs). I was inspired to draw this diagram after studying Richie Burnett's web info regardingchoke output DC power supplies. Compared to my resistor output, twin MOT, pulsed DC power supply, it seems to offer certain advantages:
* Reduced Ohmic losses thanks to low-loss choke output (no hot resistors!)
* Higher voltage; up to 24kvdc compared with only 12kvdc from my old twin MOT power supply. Higher voltage is a good thing, because it allows the use of a smaller tank capacitor to get the same bang size. This implies more turns on the primary, for reduced gap losses.
* Convenient power control via a variable speed ARSG. No need for heavy, bulky, and potentially complex ballasting hardware. This feature may make unballasted 120vac operation feasible for North American coilers.
Before this love fest gets out of hand, I should point out the drawbacks of this as-yet untested power supply:
* More complex design requires four MOTs instead of two, plus more MOT capacitors and more HV rectifiers. The choke output supply will be bigger, heavier, more expensive, and more difficult to build. My original twin MOT supply has a "footprint" no bigger than a single 15/60 NST. The choke output supply will be more than twice the size and weight.
* The choke output supply requires a more complex variable speed ARSG (for power control), as opposed to the simple static gap employed in most Tesla coil systems. This may be a significant discouragement to novice coilers.
* The spare MOTs employed as output chokes are an unknown. The output choke/MOTs could fail due to the high output potential between the secondary and the core. My original twin MOT supply never subjects the MOTs to anything more than their designed voltage stress of 2kvac (3kvdc).
This project is one of those rare ones that is interesting enough to get me out of my swivel chair and into the garage to actually build something. I've got almost all the materials together. I've started fabricating the HV rectifiers from strings of 1N4007s. I'm snooping around for more MOTs. I've taken my cheap, junky ARSG out of mothballs. I hope to have a prototype running by Spring 2002. If anybody beats me to it, please share your notes.
I almost forgot about theory of operation! I shouldn't assume that everybody "gets it" on first glance. Here's how it works: viewing the diagram from left to right, the two 120vac MOTs are each rated at 2100vac output. Wired together in reverse-phase, they develop 4200vac. This is applied to a full-wave voltage doubler / filter circuit consisting of two 12kvdc, 1 amp silicon rectifiers and eight .97uF, 2200wvac microwave oven capacitors. The total charge developed across the doubler / filter capacitors is about 12kvdc. This is applied to the Tesla coil tank circuit and charges tank capacitor C1 to 12kvdc through two output chokes (actually just MOT secondaries). The 12kvdc charging current through the chokes causes them to become strongly magnetized. As C1 approaches full charge, the current drops rapidly, causing the magnetic fields around the chokes to collapse back into the windings. This induces additional current in the chokes. This extra current is prevented from returning to the doubler / filter capacitors due to the one-way action of the output rectifiers. The additional curent must go into C1, resulting in increased charge up to as much as 24kvdc (twice the original voltage). When the rotating ARSG electrodes move into firing position, C1 rapidly discharges through the Tesla coil primary, completing one firing cycle. The discharge pulse is so fast, and the inductive reactance of the output chokes is so high, that for an instant, the power supply and the Tesla coil are electrically isolated from each other--the chokes feel like a nearly infinite impedance. The charge on C1 is spent, and drops to a very low voltage. The rotating ARSG electrodes move out of firing position, and the charging cycle is repeated. Cool huh?