Traditionally, you want to set the on time so that the coil turns off right around the OCD trip point. For a coil with higher primary impedance, this will mean higher on time, as current rises slower. Usually, lower current/voltage coils will use higher on time to get the same spark length as a higher current coil. Anything over 400-500us is typically done with pulse skipping. This can get you crazy long arcs (see:

but also makes music sound a lot fuller, especially bass (see:

vs my coil https://youtube.com/shorts/Ns2FfKgEC-c at 1ms with half the onetesla's bus voltage). To do this, you need a pulse skipping driver (or really high impedance primary), good igbt/primary cooling, and beefy igbts for your coil size. Id say set your ocd to at most twice the max continuous current, as opposed to around 4x which is common for lower on time coils. You probably wouldn't want to run high on times with a single to-247 bridge either, due to its low thermal mass. Single board coils like the onetesla often are restricted to <100us as they don't have an OCD, so any longer will allow the current to ramp up high enough to kill the igbts.

Hopefully this is in the correct section, I guess this could be spanned across a few sections.

Sometime ago I bought a "dawncap" from AliExpress, its finally arrived! The capacitor was sold as being 0.44uf, 6kv and a rating of *presumably* 60a rms current.

Currently (I haven't checked in a few months) there's no real information about these capacitors making them an unknown, hopefully this thread can give some information on them!

My plan is to replace the MMC in my drsstc 3 and put this single capacitor, and slowly ramp up bus voltage/on-time to see if/at what point the capacitor starts to heat. For me this is a simplistic approach, would there be a better way of testing?


They have a complete specification sheet on the net, If I find mine I will post.


Cheers.

Quote from: FonziDaytona on November 17, 2023, 11:40:01 PM

I see the on board MMC has bleeder resistors and my upgraded MMC doesn’t. Maybe startup IS the issue after all?

-Matt

This is what is tried to be fixed with a 10K resistor across the inverter output, to get rid of any residual charge that could confuse the logic at start of a pulse, not just start-up, but it would sound sputtering and unstable.

Did you take a look at all gate driver resistors, TVS and diodes from GDT to IGBTs? Many components can fail "half" at IGBT failures, like gate resistors going from 10R to 100R-10K and you get all these weird behaviours from differently damaged components in the drive chain.

Quote

Hello again, Dave, just getting down to doing a final wind of my gdt and looked a yours and madds and like the wrap method. One question, on one core only had 4 windings and the other 11 wraps, how should one go about picking a number to wrap.  I was just going to go for about half, but is their any good numbers. I usually make a gdt and test it, by checking the input and output at the frequency of my coil looking for the best response per wrap method and core.

Most people wind more turns than needed just for margin.  Down side of more turns is higher leakage inductance.  Requires higher gate resistance to avoid excessive overshoot, slowing down Vge rise and fall times.  Down side of too few turns is core saturation, which causes more serious problems.  That's why people error on the high side.  However, if calculated or measured accurately, there is no need for extra turns.

Given your prompting, I just added a GDT turn calculation post to my GDT tutorial thread:
    https://highvoltageforum.net/index.php?topic=1854.msg19355#msg19355
Please let me know if it needs any further clarification.

Quote

This project has went slow for me, found out I have four different cancers and been getting them removed. Little stressed out at times. Thank every one on this site, I always worry that some of you all will disappear from the site before I get mine running.

Sorry about your cancer recurrence.  Hope that doesn't repeat.  I'll intend to be around for a while yet.  My thyroid cancer is not too likely to show up again.

How about a split-bolt connector such as:
https://www.homedepot.com/p/Commercial-Electric-4-1-0-AWG-Copper-Split-Bolt-GOEC-25/310741810