[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

RE: Terry's DRSSTC -"different" H-drive functions...

Original poster: Terry Fritz <teslalist@xxxxxxxxxxxxxxxxxxxxxxx>

Hi Steve,

At 03:04 AM 1/12/2005, you wrote:
>Like hitting the primary
>every tenth cycle after breakout or something to maintain say 100 amps in
>the primary.  All kinds of wild possibilities! but they require passing
>high speed digital to the coil back to computers and such.

Coincidentally I've been working on a similar idea. My plan is to chop out
cycles of gate drive as needed to keep the current below a user defined
limit. .....

But I think we need a "new" H-bridge design... More on that latter since I just thought of it while waiting in line at the fast food place ;-) But we might be able to use the same ones too... Have to think more on that...

Here is what I was thinking...

"Normally", when the gate drives are off, the IGBTs are pretty much slamming on the breaks for the primary loop current. If all the IGBTs are not conducting, then the primary Q is very low and the energy stored in the primary is very quickly being dissipated in the free-wheeling diodes in the IGBTs.

Since "quenching" is not needed in DRSSTCs, the 'having all the IGBTs off' situation really just adds to device heating for no real good reason. If the primary loop has say 2 Joules in it when we turn the IGBTs off and all that energy goes into the reverse diodes at 200BPS, we might burn off 400 IGBT watts!!!

But suppose we instead just turn off the top two IGBTs and turn on the bottom two... No energy is being added now and the H-bridge is still just basically a dead short to the primary so the primary loop current can continue ringing at a relatively low loss. By keeping the bottom two IGBTS on, we basically just let the primary loop "freewheel" and not waist the stored energy as well as reducing the thermal load on the IGBTs!! Of course, running that loop at say 300 amps for a long time will heat the IGBTs too, so that may be a wash...

Since my DRSSTC has independent gate drives for each IGBT, it would be easy for me to do this ( >:o))) Those with gate drive transformer systems might have to go to 4 drive transformers and keeping the IGBTs on for a long time might be a problem due to the transformers "letting go" after being in the DC state for too long... I have to be careful too since my IGBT drive power supplies have a time limit (bigger caps..., have to recheck that design to handle this situation)...

I wonder if there is a big IGBT motor drive system out there that already does all this? Unfortunately, the late Marc Metlicka was our IGBT motor drive expert... But I don't think any AC motor drive could touch ~200kHz operation... However, there are little motor drive H-bridge (1/2 bridge) ICs out there that can run very fast that might really simply things and they have good current drive to run gate transformers ;-)) They have things like forward reverse, free wheel and break (opposite meaning for our purpose). They also take care of things like cross conduction and all that.

Perhaps one could make a tiny DRSSTC with standard DC H-bridge motor drive ICs right off the shelf!!! They do often have reverse diodes and all that... They have good isolation and protection and could do the above function as a standard feature of those ICs!!

So if one wants to maintain say 200 amps in the primary loop, shorting the IGBT bridge might really be needed... Of course if the loaded system Q is ten, then we have to input about 10% of the total system power too. As you stated before, we may need an oil cooled secondary. But some people do have those ;-)) I work with sustained 500 amp RF circuits and one also has to really worry about eddy currents and "lots of copper" issues too or it will just inductively cook everything near it. I suppose when we get to the point of "exploding" the secondary coil, we will have reached some sort of milestone ;-)))

Much much to ponder....  Almost too much %:-))