Coupling coefficient and some more Questions

[Admiral Aaron Ravensdale]

Hi all,
I show you my coil already here and get some help from you.
https://highvoltageforum.net/index.php?topic=2821.0
Now I run into a few more issues.

Here is the Video and the questions.

• The 2x 10K 50W Bleeder Resistor over the two Caps of the Voltage doubler gets terrible hot. Maybe I need a Headsink or can I use the of the IGBTs? When YES, is an electric insulation nessesary?
• The Outer primary Tab starts glowing also the 8cm behind of the 6mm copper tube. Should I use water cooling?
• The round flashover in the middle of the video round the complete Secondary is anoying me. The Secondary starts 1cm below Primary with a coupling of 0.134 JavaTC like recommend 0.13. When I lift it by 1cm and start at same hight with Primary I get 0.122 will this improve it? Picture and JavaTC values attached
• No real issue. Still don`t get to the 400A OCD the Variac can deliver 20A but the 35A Pulses will pop the 16A Line braker.

CM200 Coil.zip

THANKS for your support!

[Benjamin Lockhart]

The bleeders getting hot is somewhat normal you could heatsink them or swap them for higher resistance ones. The main heatsink would be ok.

If you're not even at 400A primary current you should not need water cooling. I'd say the need for water cooling starts closer to 1000A
It looks like your primary clip might be steel! That would definitely be the problem and why it's glowing red. You should use copper or brass for the clip there.

Regarding the flashover, is the inner turn of your primary coil connected to the bridge or the MMC? It should be connected to the bridge or you will have a high voltage node there.
I don't recommend having the secondary winding start below the primary no matter what the coupling is. I'd raise it to be level or even a tiny bit above.

If you want higher coupling the best way is a fatter secondary coil. I also really recommend a much larger toroid if you can make one, at least the same diameter as the height of your secondary or even a bit bigger.

On a side note 400A seems quite low considering how low your primary impedance is! I'm not quite sure why? What on-time are you running at?

Your MMC value also seems a bit big for the spark output you're getting, if you can use a higher primary impedance that still gets up to the same primary current just with longer on-time. I think things would work better and you would have less flashover because the fast primary current ramp up with the big MMC causes a high voltage gradient on the secondary coil.

In short, I'd aim for a fatter secondary and smaller MMC (with more primary turns) but still higher primary current, to get big sparks without flashover.
Take a look at my DRSSTC3 project. It is similar to yours but does 2.5m sparks.

This is by best understanding right now so other members feel free to correct me if I'm wrong.

[Admiral Aaron Ravensdale]

The bleeders getting hot is somewhat normal you could heatsink them or swap them for higher resistance ones. The main heatsink would be ok.
-->OK, they get terrible hot after 5min. When I mount them to the IGBT heatsink should I isolate them it's a 50W resistor with metal housing.

If you're not even at 400A primary current you should not need water cooling. I'd say the need for water cooling starts closer to 1000A
It looks like your primary clip might be steel! That would definitely be the problem and why it's glowing red. You should use copper or brass for the clip there.
--> doh, now when you mention it's metal... put a iron sheet between an induction coil, is maybe the most stupid idea from me!!! No comment, but sometimes you recognise not the easiest reason.

Regarding the flashover, is the inner turn of your primary coil connected to the bridge or the MMC? It should be connected to the bridge or you will have a high voltage node there.
I don't recommend having the secondary winding start below the primary no matter what the coupling is. I'd raise it to be level or even a tiny bit above.
--> yes the inner turn is behind the MMC, was not aware that this can be an issue! will change it, maybe i need to turn the FB too.
--> Today I already print a spacer, the secondary will start at the upper half of the 6mm tubing.

If you want higher coupling the best way is a fatter secondary coil. I also really recommend a much larger toroid if you can make one, at least the same diameter as the height of your secondary or even a bit bigger.
--> its already 550mm i don't want a bigger because the frequency is already at below 90kHz with brake out Pin.

On a side note 400A seems quite low considering how low your primary impedance is! I'm not quite sure why? What on-time are you running at?
--> yes in a first step I dan't want only double the CM200 limit later it was planned to go to 550-600A or maybe use CM300 when I found a good offer.
--> On time is currently limited by the output of the Variac and the one phase power line. I already build on a VAriac stack with 2 phase input and 40A output.

Your MMC value also seems a bit big for the spark output you're getting, if you can use a higher primary impedance that still gets up to the same primary current just with longer on-time. I think things would work better and you would have less flashover because the fast primary current ramp up with the big MMC causes a high voltage gradient on the secondary coil.
--> OK will keep this in mind when there are still flash overs after raising the secondary. I only get the flash over by a special music note in the song 3x flashover in 2min 50s long song. hopefully the mmc connetion switch and raising improve it enough!

In short, I'd aim for a fatter secondary and smaller MMC (with more primary turns) but still higher primary current, to get big sparks without flashover.
Take a look at my DRSSTC3 project. It is similar to yours but does 2.5m sparks.
--> Maybe it is the distance of the cam and the angle. I already got on every brakeout point up to 1.8m sparks

This is by best understanding right now so other members feel free to correct me if I'm wrong.

Thanks a lot for your Input

[Benjamin Lockhart]

I believe those bleeder resistors are already isolated to a few kV.

With a fatter secondary coil and thicker wire with less turns you could have the same resonant frequency but higher coupling. This would mean a compete rebuild of the primary as well though.
For reference, the coupling on my coil is 0.162 and I've never had any flashover and the primary is level with the bottom of the secondary.

why don't you want a lower frequency? It's easier on the IGBTs. My coil runs at 60kHz. You would need more primary turns for it though.

[Anders Mikkelsen]

What's that primary clip made of? It looks like iron. That would explain it overheating, as iron is very lossy at high frequencies due to the high permeability.

[Admiral Aaron Ravensdale]

What's that primary clip made of? It looks like iron. That would explain it overheating, as iron is very lossy at high frequencies due to the high permeability.

Yes its iron I mention it above in my answer...
FUN FACT: I order non ferrus screws and wingnuts because it's so close to the primary :-) sometimes I make very crazy mistakes...

[davekni]

The round flashover in the middle of the video round the complete Secondary is anoying me. The Secondary starts 1cm below Primary with a coupling of 0.134 JavaTC like recommend 0.13. When I lift it by 1cm and start at same hight with Primary I get 0.122 will this improve it? Picture and JavaTC values attached

Regarding the flashover, is the inner turn of your primary coil connected to the bridge or the MMC? It should be connected to the bridge or you will have a high voltage node there.
I don't recommend having the secondary winding start below the primary no matter what the coupling is. I'd raise it to be level or even a tiny bit above.
If you want higher coupling the best way is a fatter secondary coil.

JavaTC's recommended coupling is for spark-gap coils (SGTCs).  DRSSTCs are generally better with as high coupling as possible without getting racing sparks or flashovers.

High coupling is difficult with a relatively tall skinny secondary (as my DRSSTC also has).  I started with a relatively-small-diameter flat primary ~4cm above bottom of secondary.  Dropped that to 2cm above almost immediately due to racing sparks along lower part of secondary.  Used it that way for 5 years, then had some issues after multiple ground-ring strikes ionized adjacent air triggering more racing sparks.  Made a larger-diameter conical primary starting at secondary bottom and ending 14cm above secondary bottom.  New primary uses large diameter tubing to avoid too much inductance increase from larger diameter.  Increased coupling from 0.141 to 0.16 and eliminated racing sparks.

I also really recommend a much larger toroid if you can make one, at least the same diameter as the height of your secondary or even a bit bigger.

Yes, bigger is always better.  Same goes for secondary winding diameter.  Of course, construction considerations limit what's practical and affordable.

[Admiral Aaron Ravensdale]

@davekni
With existing Material I am able to make 2 changes beside raise the secondary and switch MMC cable.
I am able to use a 0,3uF MMC instead of 0,44uF
and maybe with additional work I can reduce the Frequence to 80kHz with a bigger toroid (lower and I run out for primary turns)
here are the JavaTC files. I am not sure if this is an improvement that I spend time on.

2M_0,3uf.zip
2M_0,3uF_68cm_Toro.zip

[davekni]

I am able to use a 0,3uF MMC instead of 0,44uF

This will increase primary impedance (reduce primary current).  Good for running long interrupter on times.  What enable pulse width range have you been using?  May not be much value unless existing enable times are quite short.

maybe with additional work I can reduce the Frequence to 80kHz with a bigger toroid (lower and I run out for primary turns)

The bigger toroid should improve performance a bit.  Large toroid will reduce secondary frequency change as arcs grow.  Existing low coupling factor limits frequency range over which primary can transfer energy to secondary effectively.  So reducing frequency change with arc length is helpful.
Before spending time on a nice larger toroid, perhaps experiment with quick cheap simulations.  For example, cover a large cardboard disk with foil and set on top of existing toroid.