SKM150 DRSSTC III

[Mathieu thm]

Hi

I just finished my DRSSTC III and I wanted to share it with you.
All electronics fit in a small 29x29x17cm box.

SPECS :

• secondary : 125mm x 410mm (125kHz)
• Toroid : 480mm x 125mm
• primary :  6.35mm dia. copper tube
• MMC : 5s15p of 0.039nF 2Kv (R75UN239050H3J) (forced air cooling)
• IGBTs: 2x SKM150GB128D
• Gate resistor : 5ohms for turn on and 1.1ohm for turn off
• Bus cap : 350v 6800µF
• Driver : UD2.7
• OCD : 375A 400A
• 1.3m+ arcs

Photos of the build : https://photos.app.goo.gl/C4tJ3HKW8YK31F247



I set the OCD at 375A and the results are quite good, approximately 1 meter of arc at 150Vac and 120µs :

/>
The OCD did not trigger, I still have to test it with more voltage, I think set the OCD at 400A.

I haven't seen any DRSSTC with SKM150s, has anyone ever done a DRSSTC with SKM150s ?

[flyingperson23]

I think generally when people have decided to use 150A IGBTs, BSM150 or CM150 is cheaper. Kaizer's drsstc 4 used skm145, but I'm not aware of any others.

[markus]

Wow, really nice coil and such a clean build! Everything looks well planned, I assume you used CAD to model your setup in advance?

[Mathieu thm]

Thanks ! Yes, I used SolidWorks, very useful for making a clean build without too much difficulty during assembly.

[Mads Barnkob]

That is a beautiful little coil! The large table top version, which is a size that very few build. I think this is why about the SKM150 modules, it falls in between the realms of small coils that can run on TO-247 devices and the somewhat larger floor models on wheels that goes up into 200-300A range IGBT bricks.

Your 3D printer has also been on a lot of work for this one, so does that work easily from SolidWorks and pick a part directly to 3D printer?

There is one concern, which might not be a problem at 375A OCD setting. Your MMC is NOT designed for heat dissipation. It is more designed for destruction and meltdown of the middle.

You should have space between each individual capacitor, in both directions. Best way to achieve a MMC that can be cooled is to use small cuts of copper or brass plate/band to solder them together with. Film capacitors get rid of their heat from 2/3 through the leads and 1/3 through the package. The small copper/brass plates will work as heat sinks and with space between all caps, forced air cooling can flow between them.

Tried  to visualize that suggested small plates, add distance between the caps in your head, paint.exe was not good enough to illustrate that

Red circle is where a capacitor will fail first, due to over heating. It cannot radiate heat from the side and leads are almost gone and only connects to another warm capacitor.

 

[Mathieu thm]

That is a beautiful little coil! The large table top version, which is a size that very few build. I think this is why about the SKM150 modules, it falls in between the realms of small coils that can run on TO-247 devices and the somewhat larger floor models on wheels that goes up into 200-300A range IGBT bricks.

Thanks !

Your 3D printer has also been on a lot of work for this one, so does that work easily from SolidWorks and pick a part directly to 3D printer?

You can't print directly from SolidWork, you have to export in STL format and use Cura.

There is one concern, which might not be a problem at 375A OCD setting. Your MMC is NOT designed for heat dissipation. It is more designed for destruction and meltdown of the middle.

You should have space between each individual capacitor, in both directions. Best way to achieve a MMC that can be cooled is to use small cuts of copper or brass plate/band to solder them together with. Film capacitors get rid of their heat from 2/3 through the leads and 1/3 through the package. The small copper/brass plates will work as heat sinks and with space between all caps, forced air cooling can flow between them.

Tried  to visualize that suggested small plates, add distance between the caps in your head, paint.exe was not good enough to illustrate that

Red circle is where a capacitor will fail first, due to over heating. It cannot radiate heat from the side and leads are almost gone and only connects to another warm capacitor.

Yes of course I know that my MMC is not designed in the best way !
I'm planning to make a kind of "duct" just before the air outlet grille which is just in front of the MMC, so that all the air will pass through the MMC before going out.
I like the idea of copper strips, I've got some 0.75mm copper lying around, I think it'll do the trick.

[Mathieu thm]

For now, I've done this so that there's a constant flow of air around the MMC :



I'll have to think about how I can place copper cuts on the caps connection

[Mathieu thm]

I set the OCD to 400A and did tests at 110µs ontime and 210Vac.
I lowered the primary frequency, I now have 1.2m to 1.3m arc maximum.
I think these are beautiful results and everything remains cold.
I'll probably test it at 425A OCD later.




 

/>

[Benjamin Lockhart]

Very nice! I'm also working on a similar size coil with a single CM300 half bridge, 0.19uF MMC, and smaller secondary.

[Mathieu thm]

Thanks !
Can't wait to see the finished build if you make a thread !