Bootstrap power supplies?

yarisboy

Hi,
Can you please help me with info? I'm building an inverter. I have good IGBTs and IGBT opto-driver chips but the DC-DC converter chips on hand are a little weak to supply my drivers. My DC-DC converters can put out 1 watt each but my opto-IGBT drivers can put out 5 amps to the gate. Doesn't look doable to me. When I swapped e-mail with our counterparts down under they told me to just use bootstrap power circuits to supply my driver chips. What the heck does the circuit look like and how does it maintain the correct reference voltage for each individual IGBT?
Thanks,
Stan Cloyd

P.S. Using the Prop with three A/D feed back signals.

Phil Pilgrim (PhiPi)

An IGBT gate is a capacitive load, so it does not require a continuous current to drive it. The 5A you cite would only be applied briefly until the gate is charged, so a 1W supply with sufficient output capacitance may well be adequate for the driver. It all depends on the IGBT's gate capacitance and PWM frequency.

-Phil

yarisboy

Thanks. As suspected. Initial experiments involve IGBTs with a 30 amp continuous rating. Current through Rg really won't be anywhere near the driver chip's sourcing capacity. The project involves an EV traction battery and 60 hertz output at 120 VAC rms. Because so many EVs out there use a relatively low pack voltage I need top be able to ping-pong charge stacked electrolytic power capacitors with the first set of IGBTs, then feed my boosted DC to a second set of IGBTs to produce the PWM ac in accordance with user load. If I can keep my capacitor bank charged up to 177 VDC I should be able to keep the loads happy with the PWM capabilities of the Propeller.
I researched inverters and found them available up to 48 VDC. Grid tie/solar inverters are just too pricey. The small cheap vfds we can afford (used) put out, in general, 240 VAC three phase and are not set up to take DC directly as an input source. The goal is to be able to use any pack voltage above 90 and drive air conditioning (5,000 btu), vacuum pumps (power brakes), and hydraulic pumps (power steering). Those of us responding to fuel-nozzle-rage are in a cost hurt because this power supply just doesn't exist yet. Smells like a Propeller project to me. Charge pump circuits abound that use an inductor and gate to charge the capacitor bank but I suspect the inductor would add quite a bit of weight and may even waste more heat than the switching losses of an IGBT H-bridge boost network. Not having to run a sky high carrier frequency will tend to reduce the switching losses even more. My IGBTs can handle 20 Khz but I doubt I'll have to go over 6 Khz worst case.