Is this laminated bus good enough?

[Benjamin Lockhart]

Hi all, I'm working on my large DRSSTC project and I've got this large Powerex inverter that I'd like to use but the aluminum bus bars are only laminated over the caps and the overlap stops just before the IGBTs.



The 6 bus caps here are connected as 2 series, 3 parallel.

Will this still have low enough inductance, or is that small loop at the IGBTs a problem?

I can maybe remake the aluminum plates to extend right over the IGBT's although I'll need to find a suitable thin insulating sheet for the extra bit.
I will post some info about my coil once I have more of it done.

Thanks, Benjamin

[Mads Barnkob]

The stray inductance of this inverter is better than 99% of all other amateur (and properly also professional) DRSSTCs. Do not worry about inductance being a problem here, it is afterall a proven industrial design, that you are re-using.

Does the cutouts near IGBT connection, correspond with the flaps for the DC connection near the caps? It just looks like it was space optimized to get as many busbars from a single large aluminium plate in the laser/CNC cutter.

[Benjamin Lockhart]

Yes the cutouts are the same width as the DC connection tabs, so you're probably right about the space optimizing. The cutouts are also there to allow access to the IGBT gates though.

I'm also planning to have a second identical cap bank bolted upside down above the main one by the side tabs, to get up to 9900uF. Will that work fine?

Thanks, Benjamin

[Mads Barnkob]

I'm also planning to have a second identical cap bank bolted upside down above the main one by the side tabs, to get up to 9900uF. Will that work fine?

You could also bend the DC connections down to be flat and lay the other bank directly on top of the other. Then use standoffs to connect the IGBT terminal furthest away from capacitors.

Make sure to leave room for snubber capacitors as well, directly at the IGBT terminals.

[Benjamin Lockhart]

I need to leave room for bleeder resistors on the caps and a wire going to the center tap of the caps. Also I was planning to just cut off the IGBT connection part from the second assembly. The second bank is already missing some of that part anyway.
Would there be much benefit from moving the 2 banks closer together?

[davekni]

Will this still have low enough inductance, or is that small loop at the IGBTs a problem?

Presuming 1200V IGBT bricks, I'd expect no issues.  If running 600V IGBTs at 500V, then bus connections would be critical.

Would there be much benefit from moving the 2 banks closer together?

I don't think so.  Only possible advantage that comes to mind is allowing the two banks to share ripple current at 2x TC operating frequency.  One bank is likely plenty capable of that already.  Line frequency ripple current should be well shared even with separation.  Line frequency is likely the larger source of bulk-cap heating anyway.

[Benjamin Lockhart]

Thanks. Yes they are 1200V IGBTs, CM300DU-24NFH, and I'll be running a 680V bus from doubled 240V.

Here's a picture of my coil so far.
I'm using a UD+ driver and a water cooled primary.