SimpleDriver v2.3, my phase-shifting QCW DRSSTC controller

[NEYi]

Greeting and salutations!

I've been into coiling for quite some time, but my work is mostly known in a local region of Ukraine/Belarus/Russia and is very obscure overseas.
I think I'll try making articles in english and start demonstrating my stuff here on HV forum – I hope you'll like my work.

Let's start with the driver board, it's called a SimpleDriver and allows users to build QCW coils without bulky buck-converter – relying on phase-shifting technology.
Power output of the full-bridge is controlled solely by an SD board, which allows making QCWs simple and very compact.

Actually I made a huge super-detailed article about it, which also explains the thing's operation principles. Check it out!
http://tqfp.org/simple-tesla/simpledriver-v23-in-english.html#!prettyPhoto

And just to catch your attention – a couple photos!





P.S. Let me know if you liked the article, I got much more things, like wireless 6-note poly. interrupter that can be controlled from a phone.
http://simpletesla.ru/wp-content/uploads/2017/08/6006b6.png

[nabzim]

This seems to me to be very, very impressive... but then again I don't know that much!
I am really surprised that nobody else has replied yet!

I really really like how all of the control i/o pins are all along the same edge at the back end, and the fiber input and indicator LEDs are on the same side, facing you. Looks like it was really well thought-out.
Great job!

[Mads Barnkob]

Hi Eugene and thank you for sharing your driver and its incredible detailed documentation, that is high class example of how good documentation can be on a product without giving it away when you still have the software secret.

95$ is quite steep compared to some of the Chinese manufacturers that comes out everywhere by now and which are another reason why I stopped trying to sell a simple cheap driver myself, as I got no chance to sell anything with a price on a simple driver being higher than a more advanced Chinese one. However you have documentation with yours, that is seldom the case with the Chinese, at-least not in English.

I would love to hear more about your interrupters aswell.

[flyglas]

Hi Eugen,

I like your compact design and also the great documentation. I found your driver design a few months ago in a russian forum.
At the moment I am designing my own DRSSTC/QCW driver based around an Altera/Intel MAX10 FPGA (MAX1000 eval board).
My design is inspired by the UD2.7, UD3, UD+ and of course your SimpleDriver.
My goal is to implement freewheeling, phase-shift PWM with alternating hard-switching between the half-bridges and the ramp generation completely without analog circuits. At the moment the whole gate drive logic is implemented with a state machine. The ramp will be generated either directly in VHDL or within a NIOS soft core.
Unfortunately my progress is slow as I build a matching QCW tesla coil in parallel.

Is your SimpleDriver used during this performance?

[NEYi]

Oh, some people had actually replied under this topic.
I checked it maybe 4 days after posting and there was nothing – thought people were like meh.

Anyway, I really appreciate that there's some interest to my work – you'll really like my other boards, I'll write about them a bit later.

So, the question-answering time!

I really really like how all of the control i/o pins are all along the same edge at the back end, and the fiber input and indicator LEDs are on the same side, facing you. Looks like it was really well thought-out. Great job!

Actually, I had plenty of time perfecting this design – first ever version v0.1 of SimpleDriver emerged into existence long back in 2014 and looked like this. https://i.imgur.com/E3GcNpr.jpg
Over the next years there were also v1.0, v2.0, v2.1, v2.2, and now it's v2.3.
http://tqfp.org/forum/download/file.php?id=1347&mode=view (v2.0 back from 2015)
http://tqfp.org/forum/download/file.php?id=1943&mode=view (v2.2 from 2017)
Most likely there'll be new versions in the future.

Hi Eugene and thank you for sharing your driver and its incredible detailed documentation, that is high class example of how good documentation can be on a product without giving it away when you still have the software secret.

95$ is quite steep compared to some of the Chinese manufacturers that comes out everywhere by now and which are another reason why I stopped trying to sell a simple cheap driver myself, as I got no chance to sell anything with a price on a simple driver being higher than a more advanced Chinese one. However you have documentation with yours, that is seldom the case with the Chinese, at-least not in English.

There's been quite a huge battle over proprietary VS open-source hardware design philosophies.
It seems like OSH had won the hearts of common people, but my personal observations/experience tell a different story – in almost all cases OSH devs have very little motivation to work on their products.
I mean, these are rarely being developed past the point of early alpha, there are barely any documentation, and practically no user support...
Which isn't surprising, most of these projects are made not as something useful for other people, but usually to get a little fame on resources like Hackaday. After that the project usually gets completely abandoned.

In contrast, designing real commercial products can keep you motivated for years even if it's not some bestseller – you'll be happy to provide long-term design-support in new features and reliability improvements, write articles, and support every user who's currently using your design. All you need is knowing that there are actual people using your stuff, and that it pays for further work on imroving it.
Not sure how you guys, but I have a special R&D bank account that pays for all the prototype boards, parts, deliberately exploded units, and of course pizza.

On chinese manufacturers – I'm pretty sure these aren't a big threat for us westerners, at least yet. Of course, casual buyer will purchase a driver for $45 just because "woot! It's so cheap!", but I haven't seen these even having a basic phase-lead.
Those who try building Tesla coils for professional use will probably play with these for little, burn up a few IGBT bricks and then get a proper driver like UD2.7/UD3/UD+, or SD.

On my price being steep – earlier SDs had the price of only $75, but it wasn't easy providing further support to every new user. I made it $95 on v2.3 with new features like alternating shift and pulse-skip. At that moment I hanged with dudes from France and Switzerland and they said that my driver was dirt cheap cause cheapest UD v2.7 clones costed there $135.

On the software not being open-source – that's basically to avoid easy production of clones. Believe it or not, but I personally know a dude who makes whole batches of UD v2.7 clones for direct purpose of retail.
Those with at least basic knowledge of HDL programming will have no problems recreating the driver by relying on operation principles provided in the article.

My goal is to implement freewheeling, phase-shift PWM with alternating hard-switching between the half-bridges and the ramp generation completely without analog circuits. At the moment the whole gate drive logic is implemented with a state machine. The ramp will be generated either directly in VHDL or within a NIOS soft core.
Unfortunately my progress is slow as I build a matching QCW tesla coil in parallel.

Is your SimpleDriver used during this performance?

Actually, my earlier driver was completely digital – FPGA based. Required around 450 MHz for digital phase-lead. http://tqfp.org/uploads/images/00/00/01/2013/02/21/ec0a74.jpg
It had USB interface and had to be configured from PC which also required writing and supporting software for Windows.
While developing a new version... Well, I ran into the same problem – the design became too complex and progress was super-slow.
Eventually, I just designed a new one that relied on analog-digital technology – it worked.

Should be enough for now, there'a already a whole essay here. Will be happy to answer more questions though – feel free to ask. Also, bluetooth poly-interrupter topic is coming.

[Netzpfuscher]

The FPGA based version of your driver is the most interesting design for me ^^ I would like to take a look at the HDL how you implemented the whole thing in contrast to the UD3. Is there a softcore running?

[Hydron]

Thank for sharing your design here - looks like another good option (I especially like the phase shift adjustment - I tried to do a similar thing on a UD2.x design but didn't quite get it right) and the price seems very reasonable for complete tested board.

I agree that info on the FPGA implementation would be very interesting - if you are able to share anything it would be appreciated.

As for open vs closed source, it is a tricky one. If you do a bunch of hard work (both with initial design and subsequent user support) then having someone rip it off and sell to people with no support doesn't seem fair.
At the same time, open source allows me to take something and improve/modify it to do exactly what I need it to, rather than stick to a fixed design. The UD3 work being done is a good example of this - I will be taking Netzpfuscher's work and using it both to run on my coil (after I modify both the hardware and software design), and also as a way to learn about using an RTOS on ARM (I am more experienced with hardware design than coding).
Given this my personal opinion would lean toward making things open-source when practical, but I certainly understand both views!

[flyglas]

Actually, my earlier driver was completely digital – FPGA based. Required around 450 MHz for digital phase-lead.

Hello Eugen,
in my approach the phase-lead is realized as known by UD2.7. With the signal from the zero current detector the state-machine is triggered to enter the next state. During QCW operation a next state (the hard switched one) is triggered by a delay corresponding to the generated ramp.
Freewheeling in case of overcurrent and alternating half-bridges in phase-shifted QCW mode is part of the state-machine.
I hope 100 MHz is enough for this. At 420 kHz coil frequency and 100 MHz FPGA clock I will have 119 clock cycles per half-wave.

What is your experience? I am also very interested in you FPGA implementation.

[Nijin]

Great Job, I really liked the article !

[dannybeckett]

Wow, this project is incredible. Your documentation is top notch!! I'm surprised this hasn't had more people commenting - I'm really interested in purchasing your driver and interrupter for my summer Tesla project Huge congrats on a solid and complete product.

A quick question on output driving power. I'm looking to drive a apir of chunky half bridge IGBT modules:

https://www.infineon.com/cms/en/product/power/igbt/igbt-modules/ff600r12ie4/

I notice at the start of your docs you say "for use with up to CM400 IGBT bricks". If I want to drive a 1200V 600A full bridge, what modifications do I need to make to your driver? Or should I design an external driver?

Thank you!