Recommended development boards for QCWDRSSTC

[dru]

Hello, I finally decided to join this forum after visiting it frequently for the past few years. I am 20 years old from east coast USA. Please call me out if I am posting something that's already been addressed or if I'm posting in the wrong topic area, I know there are a lot of posts about QCWDRSSTCs but I haven't been able to find a solid thread for this specific question. For the last couple of months I have been working on a QCW DRSSTC with a half bridge topology that utilizes the voltage ramps of my US mains 120vac 60hz to achieve the QCW effect. When I first started working on it, I was designing my circuit based on one I found online from Labcoatz, but I was having a lot of trouble with dying IGBTs and faulty control circuitry. In general I tend to have a lot more success when I design the circuits from scratch, as I find it very challenging to trust the calculations of others without making them myself and running simulations so I can optimize it on my own time. I think Labcoatz design was great, but it was missing a few things that would require me to completely redesign the PCB and control circuitry. Anyways that's a lot of text sorry, my question is if anyone has any good recommendations for development boards from reliable manufacturers that are good for precision PWM control, durable in intense EM fields (or at least easy to put in a faraday cage), and preferably use C++? I'd be flexible on the programming language, I really only know python, c++ and then branches like Sagemath and Matlab. I have a lot of free time to work on circuits right now because I'm taking a year off from school. I have been using random boards that I got from someone on Ebay along with a lot of scrap electronics, so they are a little outdated but are useable. My current focus is on programming the voltage ramps, modulating audio through off-time (though I'm curious about the possibility of carrier waves but not sure that's feasible here), and designing the external control circuit. I'm also open to any advice on general design, tuning, feedback methods, or links to papers so I can better understand the governing dynamics of QCW coils. I included a 10 second exposure photo from the last time the coil was operational. Thank you and apologies for the length of this!

[Rafft]

hi
are you good at coding?
I would start off at generating ramps. 0~5vdc. forum member Finn Hammer has created a ramp+interrupter generator from arduino.

then the buck converter. Gao has made a simple one.

the rest is your basic UniversalDriver with parts for operating on 350~450KHz. then the resonator.

audio can be generated by swapping the 'ramp-up' part w audio tone.

meanwhile  here is my 6 month qcw journey from scratch.
https://highvoltageforum.net/index.php?topic=1914.0

[davekni]

In general I tend to have a lot more success when I design the circuits from scratch, as I find it very challenging to trust the calculations of others without making them myself and running simulations so I can optimize it on my own time.

Wonderful outlook, especially the use of simulations!

or at least easy to put in a faraday cage

I have no specific processor boards to recommend.  There's been some recent mention here about Raspberry Pi Pico, but I haven't seen the ECB layout.  A solid ground plane is key to a robust board.  Yes, a Faraday cage can help greatly, especially with less-than-optimal boards.  Even foil covering a cardboard box can work.  Wires/cables between inside and outside of cage are the most problematic.  Common-mode chokes are a huge help.  Signals to the power electronics (half-bridge) are often cause the most issues, so need the larger chokes.  Sometimes I use two stages, several turns around a low-frequency core followed by one or two turns around a higher-frequency (ie. NiZn ferrite) core.  Connecting through optical fiber is ideal, but adds complexity.

I'm also open to any advice on general design, tuning, feedback methods, or links to papers so I can better understand the governing dynamics of QCW coils.

Don't know of any concise papers.  I don't think QCW arc behavior is all that well understood by anyone.  There are some interesting discussions on this forum on modeling arc loading for both QCW and normal DRSSTC coils, including some simulation models.
In general, it seems that QCW coil designs are best with as high coupling factor as is possible.  (Limited by flash-over from secondary-to-primary due to geometry required to get high coupling.)  So far my only personal experience is with my low-frequency (100kHz) QCW experiment.

[dru]

hi
are you good at coding?

Hi Rafft, thanks for responding! I am alright at coding, I've taken a few computer science classes at uni level and took linear algebra a while ago so I have a good hang on most languages. The buck converter is one thing I haven't implemented in my design yet because I'm worried I will have more trouble with the IGBTs. I run my coil from a variac right now that can supply 0-230 volts 60hz at 3000VA. What advantage would the buck converter offer? The igbts I usually use are the FGA60N65SMD, so they can handle 650 volts and 180 amps pulsed. They still frequently die, which I think is from the faulty controller I was using.

[dru]

I have no specific processor boards to recommend.  There's been some recent mention here about Raspberry Pi Pico, but I haven't seen the ECB layout.  A solid ground plane is key to a robust board.  Yes, a Faraday cage can help greatly, especially with less-than-optimal boards.  Even foil covering a cardboard box can work.  Wires/cables between inside and outside of cage are the most problematic.  Common-mode chokes are a huge help.  Signals to the power electronics (half-bridge) are often cause the most issues, so need the larger chokes.  Sometimes I use two stages, several turns around a low-frequency core followed by one or two turns around a higher-frequency (ie. NiZn ferrite) core.  Connecting through optical fiber is ideal, but adds complexity.
Don't know of any concise papers.  I don't think QCW arc behavior is all that well understood by anyone.  There are some interesting discussions on this forum on modeling arc loading for both QCW and normal DRSSTC coils, including some simulation models.
In general, it seems that QCW coil designs are best with as high coupling factor as is possible.  (Limited by flash-over from secondary-to-primary due to geometry required to get high coupling.)  So far my only personal experience is with my low-frequency (100kHz) QCW experiment.

That is good advice about the chokes I hadn't thought of that, thank you! The fiber optic cables are what I'm planning on using for the connection between controller and gate drivers enable pin


Yes my most recent coil had a coupling factor of .577 which definitely seemed to help. I'm wondering though, I've seen a lot of QCW coils that have a low aspect ratio, while mine is at around 3:1. Should I try a different coil size/what advantages would that bring? Thank you!

[davekni]

Yes my most recent coil had a coupling factor of .577 which definitely seemed to help. I'm wondering though, I've seen a lot of QCW coils that have a low aspect ratio, while mine is at around 3:1. Should I try a different coil size/what advantages would that bring? Thank you!

0.577 coupling sounds great.
I've heard of only two reasons for low aspect ratio.  One is to get high coupling.  You are already high enough.  Other reason is that many people rate QCW coils based on the ratio of arc length to secondary length.  A short secondary makes that ratio higher for a given arc length.  Matters only if you personally care about that arc/secondary ratio metric.