Re: Frustrating Problems with Gate Drive Transformers on SSTC - Ahhghhghghg! !! ! !

["Tesla list" <tesla-at-pupman-dot-com>]

Original poster: "Jan Wagner by way of Terry Fritz <teslalist-at-qwest-dot-net>" <jwagner-at-cc.hut.fi>

Hi,


On Tue, 27 May 2003, Tesla list wrote:
 > Original poster: "by way of Terry Fritz <teslalist-at-qwest-dot-net>" 
<dhmccauley-at-spacecatlighting-dot-com>
 >
 > I'm having some seriously frustrating problems with the gate drive circuit
 > on my full-bridge SSTC.
<big snip>
 >
 > If the TC442x chips are in the circuits by themselves, they are
 > fine, but as soon as you mix a pair of TC4421 and TC4422 together
 > (outputs of these go to gate transformer), then one get SMOKIN HOT!

Sounds like you should add some small freewheeling/clamping schottky
diodes. The TC442x have internal diodes, but they're ~1.8V V_forward, and
the chips do heat up surprisingly much.

There's no heating with TC4421+4421 or TC4422+4422 because then you aren't
driving anything (no voltage accross the load).


 > The only thing I could think of was now that there is 75% deadtime per
 > output (as opposed to 50%) before, that maybe i have some DC bias or
 > something on the transformer primary winding.

Only with 100% dead time there's no DC bias... ;-))


 > I tried adding a 0.1uF capacitor in series with the primary winding of the
 > gate transformer and it fixed this problem.  Now the TC442x run cool and
 > seem to work.

100nF sounds small - it may resonate with the gate drive inductances and
the output looks "rounded off" at some parts. Maybe an additional larger
cap (t.ex. some "high voltage" >50V electrolytic 10uF) helps?

16V electrolytics aren't good for this, low voltage rating<=>high ESR...

 > HOWEVER!  Always a however . . .   Now my gate drive signals are all messed
 > up - basically probably because of the series capacitor now.
 > They look a bit like this:
 >
 > http://users.snip-dot-net/~liche/waveform.jpg

Non-50% duty and gate drive transformers don't mix very well.

You could try http://dsms.ajusd-dot-org/~fritz/AN1.pdf (D2,C3,R1 not
absolutely necessary, and xfmr ratio in your case 1:0.5 because of the
dual-ended drive)

Here the coupling caps and gate driver output impedance also act a bit
like a low-pass filter for the PWM'ed audio, so they shouldn't be made
too large either... Unless you want to build a subwoofer ;-)


regards,

  - jfw

--
*************************************************
  high voltage at http://www.hut.fi/~jwagner/tesla
  Jan OH2GHR