Next Gen DRSSTC

[coilerer]

First of all, Thank you, everyone, for your development of the UD3. I'm so excited to get mine integrated into my qcw coil.
On that note. I spoke to TMAX a while ago over email and confirmed that there is no buck ramp qcw mode currently available. Steve Ward mentioned he got "hacky" buck ramps to work on the early UD3. Is there any plan for developing buck ramp output as an alternative to phase shift? It would be great to get the arduino out of my build and just use the UD3.

Another question. Currently, I trigger my coil with an interrupter pin from the arduino. How would I go about telling the UD3 when the arudino is sending a ramp? This may become clearer to me as I mess with it, but I am still waiting on my programmer.

Thanks again for your hard work.

Daniel

[Hydron]

You could probably do your own "hacky" ramp by fiddling with the PSoC code. Likely would just need tweaking the appropriate logic schematic to output pwm for the ramp instead of a variable phase shift, plus possibly a bit of SW tweaking, but the majority of the code can probably be left as is.

[Netzpfuscher]

We have no DAC-Channels free. I can give you a PWM output of the ramp, should be no big change.

[coilerer]

You could probably do your own "hacky" ramp by fiddling with the PSoC code. Likely would just need tweaking the appropriate logic schematic to output pwm for the ramp instead of a variable phase shift, plus possibly a bit of SW tweaking, but the majority of the code can probably be left as is.

That's a bit out of my short-term comfort zone, but maybe I can try in the future . Thanks!

We have no DAC-Channels free. I can give you a PWM output of the ramp, should be no big change.

That would perfectly replace my Arduino. It would also mean I don't need a wifi connection (to my Arduino) to control my ramps!

[Jj]

I'll be releasing that version as the UD3.1c once I verified the changes. The next batch of UD3's in my shop will be that version.

Hi Tmax do you have an idea of when your boards may be available again? I’m very keen to get a few to upgrade my old DRSSTCs.

[Mads Barnkob]

I am gathering parts from around the lab and work shop, to get the driver running. When looking for the Ibus current sensor, I did not have a LEM 306-S in my collection, but I did have some Honeywell CSNS300-001 hall effect sensors.

So far so good. Looking at its datasheet, it also takes +/- 12V and delivers 150mA at 300A. Identical to the LEM 306-S. But checking the numbers, 150mA into 50R would give 7.5V, not 5V maximum of the Ibus input.


It is also listed to be a 1000:1 ratio CT (in schematic), but both are 1:2000


So there might be a few errors like wrongly stated turns ratio and shunt resistor should have been 33R instead of 50R?

[flyingperson23]

150 mA would indeed give 7.5 volts. My LEM LF 305-S seems to work fine. It's supplied +-5v, so can only output 100mA so is limited to measuring 200A. I put in 2000 as ct2_ratio and it works fine, although it appears to be a few hundred milliamps off. I know the current goes through a RMS filter, could that be causing the error? I need accurate current measurements to control a boost converter with the UD3.1; what would be the best way to get instantaneous/non-filtered readings?

[Netzpfuscher]

@Mads

It's no problem to change the CT2 shunt to something different. It needs no change in the code. Simply solder whatever you like, then change the "ct2_burden" (Ohms) accordingly. It's better not to rely on the internal clamping diodes of the PSOC, so make the shunt
small enough not to fry the MCU input. (There is a 470 Ohms in series, but better safe than sorry). Then change the "ct2_ratio" to 2000 and you're fine.


@flyingperson23
The UD3 software is not designed to output every current sample. It uses DMA to collect 25 samples from the ADC before interrupting the CPU to calculate everything. I implemented this to decrease interrupt load. The current is sampled with 8 kHz the CPU is interrupted at a rate of 320 Hz so that's the rate you're getting 25 sampled at a time. Is that fast enough? Otherwise you need to decrease the
https://github.com/Netzpfuscher/UD3/blob/892b8c25da2784e880c0c2617d417b14c3421ecd/common/ud3core/config.h#L40

This is the line calculating RMS:
https://github.com/Netzpfuscher/UD3/blob/892b8c25da2784e880c0c2617d417b14c3421ecd/common/ud3core/tasks/tsk_analog.c#L261

[Kevindk9]

I have a few suggestions to help make this an actual 'universal driver'

-Manual controls
Ability to operate the driver without connecting with a computer. Just a way to use an on/off foot switch and potentiometer to control power. Like it only runs when a button is pushed.

Also, It would be nice to have buttons and hotkeys in Tesla Term also, that way you don't have to go through the drop down menu to do stuff. (the tesla coil messes with my track pad)



-Induction heater mode
Currently im using a slightly modified version of the UD3 code and I run my induction heater in QCW mode. I modified the max on time and minimum repeat time so a new QCW shot happens as soon as the old one ends. The problem is it starts the whole QCW routine each time (hard switching and tracking). I believe this killed my IGBTs once because it hard switches for 10s of cycles at a very high voltage.
Ideally the IH mode would only do the routine a single time and implement a debounce feature to make sure the tank has time to ring down incase you interrupt it.


-Transformer mode
Just a simple mode to run a ferrite transformer. Ability to set a fixed frequency. You could control it using a button press or run it from Tesla Term and use it to modulate the arc and play music.

[TMaxElectronics]

Interesting ideas, some of which I have actually at least played around with in the past.
I have a little ferrite core HV source that runs from a UD3 just using pulse width modulation. Pretty much just like you said: you could use the the pw slider in teslaterm to control the output power. I've also used the UD3 as a induction heater on a few occasions (once inadvertandly by having a hammer next to my big coil during testing... that then burned my hand ).

The problem that I have with implementing that as a permanent feature is that there are some medium to large hardware changes required (especially for ferrite core mode) that aren't easy to do on the fly in the psoc. At least i don't know how Just placing different hardware next to the existing one and not enabling it until needed is unfortunately not an option as some types of the psoc hardware are already 100% used. I suppose the induction heater thing might work, but we'd really need to plan a proper mode switch in for that. Not just set the max. dutycycle to 100%

The idea of operation without a computer is nice and is actually partially implemented. The autostart option makes the coil behave exactly like a UD2.x or older coil with an interrupter input. But connecting anything to the UD3 directly (as in: without fiberoptic isolation) is not that good of an idea. We've already killed a few PSOCs just with crap coupled in through WS2812 LED strips

But in the end it is an open source project: if somebody has an idea they can get the code and play around with it themselves, and if it works out well contribute back to the main codebase

[malte0811]

Also, It would be nice to have buttons and hotkeys in Tesla Term also, that way you don't have to go through the drop down menu to do stuff. (the tesla coil messes with my track pad)

This is something I have considered previously (mostly for the "kill" button), I just never got around to implementing it. I am currently in the middle of a larger refactor in TT (connecting to multiple coils with a single TT instance), once that is done I can look into hotkeys. Can you open an issue for it at https://github.com/malte0811/Teslaterm so I do not forget about it? In general hotkeys are going to be a "fallback" option though, with a mouse being the intended method of interaction. Have you tried using a "traditional" mouse rather than a touchpad? In my experience this is much more reliable around coils.

Since I'm already posting here: Currently Teslaterm has the option to run as a "headless" server (e.g. on an RPi installed in the coil), which the actual control computer then connects to using a standard browser. Is anyone actively using this mode ("node")? It complicates the code significantly in some places, so I would like to remove it if possible.

[Mads Barnkob]

This is the first video in a series of 3, on how to get started with the UD3.1b Tesla coil driver, FiberNet and TeslaTerm.

I love you all for the great work!

[Jj]

Excellent video on the UD3!

I ordered some UD3.1C boards from JLCPCB with help from TMAX earlier this year to upgrade a DRSSTC I built way back in 2006. Quite a steep learning curve and very happy with the results so far. 1.6M from the coil to ladder. That seems to be the limit with the 150mm x 600mm toroid. I’m currently changing to a larger one and think it will make 1.8M which is the design target. I don’t seem to have much luck with higher impedance primaries and am currently running with .75uF primary cap (fres 50kHz) with the peak current around 400A. I’ll start a thread for that coil at some point and see if I can’t get some of the excellent wisdom at this forum.

[Benjamin Lockhart]

400A seems really low for a 0.75uF tank cap! My 60kHz coil with a 0.365uF tank cap gets up to 1000A with over 2m sparks at 130us on time.

[Jj]

The videos of your coil are very impressive.

Current is intentionally low. The on time for 400A is quite short at 130us. Past their current rises very rapidly if the secondaries energy doesn’t go into a streamer and returns to the primary. The setup is a half bridge of mini blocks and 500V DC. The IXYN 110N120C4H1 used are rated for 220A continuous and 760A 1ms. None have been lost yet, but I am weary of pushing them to far. K is about .17 which seems to really limit the current rise before the notch. I had planned to upgrade to a full bridge, but realised that would really only increase the current and the primary cap would need to be halved!

The secondary is 200mm x 620mm wound with .28mm wire and at least 2000 turns. Way back when it was originally made I was shooting for low fres mainly to be easier on the older IGBTs we had then. My original driver had no phase lead and I suspect the higher current switching was the main cause of IGBT failure. That was with a dual bridge of TO247s and split primary capacitor.

I have some CM600 bricks, but they would be a little wasted so the future upgrade path is looking like SKM300 or 400.

[flyingperson23]

I'm running my big and small coils both at their datasheet specified limit - 1200A for CM600 and 300A for BSM150, both for 1ms max on time with pulse skip drivers. The long on times help me get great arcs without pushing IGBTs past their rated limits. Together they've run at least a dozen hours and so far no IGBT failures, so you're probably good to go up to at least 760A assuming you have decent cooling or a thermal shutoff switch. If you're worried about your IGBTs, a thermal shutoff switch might be a good idea. It's saved my (passively cooled) small coil many times.