High Voltage

While no stranger to high current, I'm a relative newcomer to high voltage projects. Hanging around MITERS for long enough, it was hard not to become interested in the field. So far I've only done one HV project, and I wouldn't say I'm an expert at it by any means, but it's a fun complement to motor drive-style power electronics that happens to share some similar techniques.

HV Projects:

DRSSTC 1 [Design Notes] [Post Archive] [MIDI Demo]

HV Links:
[MITERS] [oneTesla] [Loneoceans Labs] [Daniel Kramnik] [I Heart Clare] [TC-Engineering]
[JAVATC] [TCFormulas]

DRSSTC 1

Status: Built Winter 2013, Software Completed Fall 2015

Specifications:

Type: Dual-Resonant Solid State Tesla Coil (DRSSTC)

Resonant Frequency: ~150kHz

Driver: Full-Bridge, 2xIRFP4668+1xAPT100S20B per leg

Gate Drive: ACNW-3190 in Modular Half Bridge (Cap Kart) Configuration

Controller: STM32F103CBT6

MIDI Streaming Software: 3-Track Custom Interface

For my first experience with coiling, I designed and built a small-medium size DRSSTC. Since I had experience with controlling switching power electronics, I went straight for the digitally-controlled Tesla coil type. For me, the challenge was learning the rest of what's required: designing the primary and secondary and actually building it. For that I read lots of work done by other MITERS members (see links) and then dove right in to FEMM, which has helped me design brushless motors in the past and serves just as well for predicting Tesla coil inductance, capacitance, and coupling parameters. I designed and built the whole thing in one sprint during the Winter of 2013.

The driver is built as if it were an H-Bridge motor driver, since that's what I know how to do. It's FETs and optically isolated gate drivers, rather than IGBTs and gate drive transformers. It certainly can't match the output power of other drivers for similarly sized coils, but it is simple and compact. And since it's designed with hard-switching in mind, it offers a degree of control over pulse generation that makes the software engineer that I probably secretly am quite happy.

Although I did some software for it when it was originally built, I set the project aside for a while when I moved out to Seattle. In 2015, I finally got around to writing the complete software interface for pulse generation, camera shutter triggering, and, most importantly, MIDI streaming.

Shutter-synced arc and primary current waveform. [Post]

MIDI streaming torture test. [Post]