Great to see the UD is lives on. Thanks, Thorben!

Kizmo said important thing. And I also should mention FDD8424H, which was good in 2009 and with kind of coils like in that past, in next two specific cases could be unreliable.
1. Some hard-gate-capacitance igbt like 1000A+ with freewheeling.
2. Some QCW like 50%+ of duty cycle with more than 2 bridges in parallel.

That is of course not a common cases, but I think that it will be good to get away from balancing on the probability that something not covering the continuous-wave-mode will fail by betting on the probability that it will not fail.

I understand, the goal of this was got packed with jlcpcb assembling, but if you have some bright thoughts about how to get this onboard, it will be nice to move this prj on this way.

My forthcoming QCW gate in attachment. TO-220 only because it has copper-plates to transport heat to radiator much more faster than fdd8424h. If there is something way to combine this case to jlcpcb assembling case it will be beautiful to add it to UD pcb prj, if you have interest to do that.

Here you go:

https://github.com/Netzpfuscher/UD3/commits/intra

*0*
Thank you!

With the VMS it was intended to make the QCW MIDI synth also block based configurable but I had not time to do it yet. So I disabled the QCW synth for safety reasons now.

Which commit is last safe for using QCW synth? This one? https://github.com/Netzpfuscher/UD3/commit/cd4d18a4d5818ee01e0f0742d89642772fda708a

There was also a lot of timing warnings when i built and programmed the board.

Not sure you can ignore this exactly warnings but I got a bunch warning with my setup and seems to it's okay. Try to skip it.

Do i need an external crystal for this board?

No.

Is the correct order to build the "UARTldr_QFN", then program the board, then build the "UD3_QFN" and program the board?

Yes. I built all then from PSoC Programmer upload "UARTldr_QFN" then upload "UD3_QFN".

As you can see in the photo im using a USB to TTL converter to connect to my laptop. I see the default baud rate set to 460800 but it looks like my computers USB can output a max of 128000. Do I change it in the PSoC creator to match my USB port?

Not sure I'm right person to answer on that, but I bet for no. What laptop model do you use?

im powering it at 17v because the on board regulators get way too hot at 24v im afraid they will die.  Could this be the problem

Yes. You should use 24v and some way cold down your board regulators or use another way to deliver regulated power to board. SMD regulators on high load without decent cooling system is in risk of overheating and damaging. For SMD there is a special boards classified as "one-sided aluminum PCB", which capable to be screwed up to heatsink it's aluminium side and pass heat to heatsink through the board's holes under component if maked, but this profdc9's SMD version did not made for that boards. So regulators and if will be high load on gate transistor then the gate transistors too should not be in SMD and should be in TO-220 or something and should be screwed up on decent size heatsinks through aluminium nitride ceramic substrate with Arctic MX-5 or another.

Do i need it to play SID music files?

I used Teslaterm to play SID files. If I remember right then you need drag and drop the SID file to Teslaterm to play it in SID mode and enable SID mode in Teslaterm

When i try running UD3-node.exe from https://github.com/malte0811/Teslaterm/releases/tag/v0.0.2  i get this error message.
I also see this version of UD3 node from https://github.com/Netzpfuscher/UD3-node but i don't know what to do with those files, so i will stick with using Teslaterm on malte0811's github.

Try this build https://drive.google.com/file/d/1pMMMhAq4u94cykIYivyjZuGO93iYG1C2/view

It‘s been silicone for the last two years; polycarbonate was before. Silicone based layout is linked in the specs section. Anyways, neither one should break down that easily at 2mm thickness so I assume it‘s the air. Even then I got 6mm of creepage between these rails and during assembly I thoroughly cleaned all surfaces. Which is why I was so surprised to see these discharges.

silicone are flexible. too much ways it can let down a task. it can become easier to squeeze when heated and it can make busses too close to each other because of force from plastic screws. not much to say without experiments.

Not quite sure what you mean by „discharge a bus cap to brick“ but I can guarantee you that the bus caps do not discharge any significant amount of energy into those small arc overs. They‘re thin, blue, sporadic. Nothing energetic. Unlike the arc over at the rectifier; that one did shorten the bus caps and you can see the results in the video.

I said that about this case https://forum.mosfetkiller.de/viewtopic.php?p=282738#p282738
Discharge a bus cap to brick means if any issue in gate signal and you will get opened an igbt which should not be opened at this time. current from electrolytic capacitor then will go through the two igbt with no significant resistance and warm up igbt core causing it to explode.

About driver which is in box. Do you have ferrite chokes between gate drive and gdt?

Uuuhm… I haven‘t heard of any 300A brick that would handle anything near 5kA reliably.

See "Wed Sep 24 2008, 03:42am" in https://4hv.org/e107_plugins/forum/forum_viewtopic.php?p=2&id=54347

I wrote inrush current limiter; making sure the 8000uF bus capacitance don‘t blow all fuses.

Sorry, my bad.

@Intra: good catch! That‘s actually something that changed already before the first crash. I forgot about that. All crashes occured with an UD+ inside a metal box.

Then it will be good if you show fresh design.

Also, for 75kHz fres, Ae=200mm^2, 240mT of N30, and 24v (if it so), it good to use 4 GDT windings but you have 13. What's on scope?

Also, pvc screw can get warm from aluminium bus and loose conduction from bus to igbt\cap. I recommend brass screw.

And rigid FR4.

(If anybody knows where to buy magnet wire in quantities greater  than 10lbs and less than 80, let me know.)

You can wind bifilar from two 26awg from this 6290ft $58+shipping spools https://www.ebay.com/itm/251040537280
For 12.75x64.11 need about 6299ft per spool which fits wery well and it also decrease secondary heating

have experimented with a larger one, but it causes overcoupling racing sparks, even after raising the secondary 1.5"

With 36x10 toroid and 22awg 12.75x48 coil you have 78.5k secondary impedance while with 50k or less you will have longer spark on same other parameters.
With 0.15 coupling your spark will be shorter than with 0.18, but racing sparks will gone. 0.11 is too low for 12.75 pipe.
Other option to improve perfomance you may try up voltage to 800v with pfc or rewire rectifier to doubler.

The toroid was also incredibly expensive, by far the most expensive part of this build at $2000USD shipped

Toroids can be self-made from tubes https://www.lowes.com/pd/Steelworks-3-4-in-dia-x-3-ft-L-Mill-Finished-Aluminum-Round-Tube/3053653
by tube roller https://www.ebay.com/itm/224567948035
My brand new TR-60 costs me $260 instead of $470 because I found it in my town.

For 62.3x12.4 toroid tubes in Lowes should cost about $400. I'm too far from LA, maybe you know better place to buy tubes or bender arround you.

You can make toroid even with aluminium foil tape. Check "Mon Jul 16 2012, 03:08pm" post from here https://4hv.org/e107_plugins/forum/forum_viewtopic.php?p=1&id=141872

The coil is also at the maximum height that can fit in my garage assembled. I could store each piece separately, but it is so convenient to be able to just wheel it in and out of the garage instead of carefully packing the toroid and secondary safely away where they will not get damaged, then needing to unpack and re-assemble it every time I want to run it.

I store my twins on 5 level of condominium with no elevator.

QCW MIDI is awesome :3

I won't change the max start cycles. There must be something wrong with your coil. You know that your coil is hard switching during startup? 

But you can do it yourself, don't set values over 254, it is only a 8 bit counter:

https://github.com/Netzpfuscher/UD3/blob/8485ae9c3ac583cc6fc78cb531a84307acef4161/common/ud3core/cli_common.c#L216

Thanks! Yes, you were right. Something was wrong with a coil. After deep analysis I found that my flashovers from previous test some way burnt sensor's comparator or +\-12v supply driver chain. After I cut out all peripheral driver part the signal became looking good.

Anyway, I choosed secondary and primary inductors for mmc I have and looks like the impedances is not matched good. Result of 20mS under 60A current limit is about 10-15cm spark.


Nevertheless I want to try to oscilate whole coil on start freq because I see the coil grab a feedback and sometimes jump between poles and I do not see another way besides to force a coil to sometimes hardswitch through the whole freq sweeping down under spark loading so please if you can change the start cycles counter to 16 or 32 bit I will be very appreciated.

The driver is connected to PC this way. It good when about 30V on bus and EMI is not so big but with 310V it crashing connection with Teslaterm even when watchdog is off and when connection is crashed then signal is stopped generating a ramp.