SiC Tesla Coil

loneoceans, Tue Jun 07 2016, 05:40AM

Hello Coilers,

I've built several coils of different sizes ranging from small systems to some rather large ones. Large coils are typically impressive when running at high power, but they pose several logistical challenges in terms of building them, storing them, running them and setting them up (and not to mention, they're pretty expensive!). Because of this this, I've been working on building smaller coils, and it has been quite a lot of fun! Turns out, building physically small coils has its own set of challenges associated with them as well.

As some of you might have seen earlier, one of my recent small builds was my tinyQCW coil (see: This was my first real attempt to make a really small yet high performance Tesla Coil. This ended up to be quite successful, with the coil producing sparks ~13x its secondary length (of 2.4 inches), and it convinced me that small coils can also be pretty cool .

There are two main problems with small coils: (1) they typically make small sparks and (2) due to their physically small geometry, and while keeping the coil still reasonably easy to construct by hand, their resonant frequencies have to be higher. This makes the power electronics more challenging to construct! This project aims to take a look at these two problems.

Though I have been quite busy with other things, I've put some thought in this project for a while so I thought I'll create a thread to share the progress as it goes along.

There has been talk about what will be the next new thing in coiling, and one of them is the advent of some new transistor technology - in this case, Silicon Carbide FETs. These have been in the market for a while now but they have pretty expensive... until recently. So despite them already laying the foundation for next-gen power inverters in the industry, as far as I am aware, SiC hasn't made its way yet into a SSTC yet.

SiC provides some very appealing characteristics for high frequency coils, specifically their extremely fast operation, higher operating temperature, very low Rds, and extremely low (but more complex) gate drive, among other nuanced things such as the rather poor intrinsic diode performance in the SiC transistor (this will require some more complex drive schemes like synchronous rectification but more on that to come later). So I decided to make a SiC based Tesla Coil.


Note: for size comparison, the fan is a 60mm fan.

Above shows my attempt at this new SiC Tesla Coil. The board is fully integrated allowing mains operation 240VAC in. It is possible to make an even smaller design by making it more modular and removing the bus caps (and driving say via a LiPo battery pack which has been executed brilliantly by Steve and forms much of the basis of inspiration of this project), simplifying the gate drive (I probably have an over-kill gate drive!) and also going to a 4 layer board, but the result is still quite small and about the size of a scientific calculator. The board features fully isolated discrete +15 -3V gate drive for all FETs with programmable dead time, integrated power supplies, on-board CT and a full bridge of 900V 35A (90Apk) SiC FETs each with its own live heat-sink.



The main board contains the power section, gate drive and supplies, but it needs a driver. Control is provided via a very tiny driver daughter-board running a PSoC, and provides a bunch of different IO, gate drive signals, and CT feedback. The modular design allows for future more advanced controllers to be simply plugged in. Despite it's small size (1/3 the size of a UD2! or so!), it is in fact a fully-featured driver and even has 4 feedback LEDs and RS422 / Fiber communication as well as temperature sensing, UVLO, and fan power control plus a few other GPIO for future expansion. This was adopted from the very fine work done by Steve and Philip so big credit to them.



Putting it all together, it looks like this (the header pins on the driver board are flipped, they should be pointing downwards). Despite its small size, it should be capable of similar power operation to my QCW 2 coil, yet being able to completely fit inside a small box, including the toroid, with a designed operating frequency of around 800kHz or so. Hopefully, making a high frequency QCW will also yield valuable insights!

More to come soon once I get the coil constructed and the boards soldered up. Hopefully, this will achieve the goal of a really small table-top coil, yet producing pretty big sparks, hopefully something like 2 feet or more. We'll see what happens!