20-60A adjustable ~200vdc current mode buck design

Feb 19, 2007 22 Replies

That's what I meant..

That's sort of what I thought would happen if you wanted to maintain a ground/neutral connection.

May as well do it so you can.

Embrace the 'iron'. You can use the full/half bridge directly to control output current and provide your isolation. There is no need to generate an intermediary DC supply and then add a buck converter you can do it all in one go.

I've had a fiddle with a half bridge using average current mode control and it's looking funky.

It might get messy. I haven't got my brane screwed in yet so I can't draw what you've got.... However boosting neg to pos is going to place extra voltage stress on that part of your converter and its control electronics are going to be referenced to a different point which may well be going up and down.

Spikes are common in simulation. Sometimes it's because you've done something stupid. Sometimes it's because there are two answers, the one you want and the one that has a big spike. Some series and or parallel resistance in inductors can help. However, in this case, it is likely that......

Yes, the node you mention can and will cause stress. Try adding a capacitor from that end of the boost inductor to circuit ground. You'll need one in real life as well. By all means, send me a copy of what you have got. There is an e-mail icon on the front page of my webshite....

formatting link

Ideal switches are nice..... (ideal everything) stick with them until it's working the way you expect. Then you might start adding 'real' stuff.

Later

DNA

----- Original Message ----- From: "Genome"

Voltage stress ?? try CURRENT stress -- the neg boost is drawing twice the current for the same voltage out :)

probably because of the non optimal component values I'm testing with -- gonna have to plug all the math into a spreadsheet and work out what SHOULD be there :)

There wont be any point in the control that will be referenced to anything but earth ground except the gate drivers.. using those hall effect current sensors gets me COMPLETELY isolated no matter if I'm sensing high side or low side.. very important since with ground shared between 2 boosts, I cant get in the low side at all, nor do I want to !! (I'm using low side with a bit of a difference from common usage in SMPS design -- to me -- low side is always ground, high side is anything that is away from ground, no matter the polarity. )

and sometimes because its really going to generate spikes :)

Dropped in the cap and spikes go BYE BYE !! :)

Thats the point I'm at -- getting it all "real" so I can subckt this and start the control electronics.

welp, everything is real except the switches now --

ASC file available if anyone else wants to look -- just ask :)

let me know what you think :)

Join the Discussion

Have something to add? Share your thoughts — no account required.

Didn't find your answer?

Ask the community — no account required