High side FET driver supply won't work due to rapidly changing voltage?

[treez]

Hello,
I wish to do a two transistor forward converter and drive the upper fet from a fet driver feeding off a bias supply.
The bias supply obviously has to be referenced to the source of the upper fet, which is a rapidly changing square wave voltage (in this case a square wave from 0 to 28v AT 200khz).
I wish to use a simple isolated flyback for this. But will the rapidly changing nature of the reference point of this bias supply mean it will not work well enough?
..I mean, will their by problems with interwinding capacitance of the flyback transformer due to the rapidly changing nature of the reference point?

 

[FvM]

I don't what "bias" means in this regard. Gate driver supply by isolated DC/DC converters is standard in power electronics, particularly for switcher voltage levels that can't be achieved with bootstrap circuits, or if other requirements demand an isolated driver supply (100 % duty cycle capability, startup conditions, ruggedness).

I translate "rapidly changing nature of the reference point" with switching frequent common voltage. Yes, the DC/DC converter must be aware of it, with respective isolation strength and interwinding capacitance restrictions. That won't be a real problem for a low voltage switcher.

 

[treez]

Yes, high side drivers are standard on today's pwm controller IC's, but the primary has 12 Amps rms current flowing in it, and normal on-chip bootstrap high side drivers aren't powerful enough. And we don't want to use a pulse transformer high side gate drive, as they need a highly toleranced, low leakage gate drive transformer, therefore, we want a high side bias supply, with a fet driver, with the gate drive from the main controller being a mere signal from the main controller to the fet driver which sits in the high side circuit. (the one that's referenced to the source of the upper fet)

This way we can assure substantial gate drive to the upper fet.

 

[FvM]

Higher gate current isn't exactly a reason against bootstrap supply. You can use bootstrap drivers with current boosters or bootsrap supply for high current driver ICs.

 

[treez]

thanks, though we use a two transistor forward converter, and the bootstrap type high side fet drivers don't work in this case.

 

[dick_freebird]

This is one of the great problems in bridge driver design
and you have blanking, latching, etc. employed in order
to let the high side stick to its business despite all the
slewing and ringing. A wireball SSI scheme isn't so neat
and has a lot more stray C to impose slew rate currents
on circuitry that doesn't respond as you'd like.

If you have a lousy bootstrap diode with a long recovery
time, say goodbye to a big chunk of your reservoir charge.
You want to determine whether this is a supply quality
problem or a control spoofing problem, as first order of
business. This can be tough to do if the high side
voltage is enough to blow out your 'scope channel.
You may need to step into it by (say) substituting a
clean capable battery as the high side supply and
ditch the bootstrap diode, just to see what happens.

 

[treez]

thanks, I may add that our Vin is 28V.

 

[mtwieg]

the bootstrap type high side fet drivers don't work in this case.

Bootstrapping high side driver work fine for two switch forward converters. You just need to make sure there is UVLO protection on the boostrap diode, which is a standard feature nowadays.

 

[FvM]

Bootstrapping high side driver work fine for two switch forward converters.

In normal operation, the source of the high-side transistor swings to ground and can recharge the bootstrap capacitor. During startup (UVLO of the bootstrap driver active), only the low side transistor is switching and charges the bootstrap capacitor through the transformer primary winding.

Or precharge the bootstrap capacitor by switching the low side transistor statically on.