hmmm...yup..keep that dead time to as much as you can afford. You know,
60KHz is awfully high for the IR2133 with it's low 200ma/420ma current
specs. You might want to look at Intersils HIP4086 Series which have a
beefier output current spec or a FET with a much lower gate charge.
Regards
Madhu

-----Original Message-----
Hi Madhu,

Thanks very much for your well-considered answer!

The reason why I think cross-conduction is the problem is that the problem
happens even with no load. I.e., there is a current drawn from the drain
supply (not the same as the IR2133's supply) even when just one leg of the
bridge is being PWMed and no load connected. The current is proportional to
the PWM frequency, and is about 0.7 amps at 60kHz, 0.3 amps at 30kHz. Also,
when I delayed the turn-on of the high side after turning off the low-side
(or vice versa) by about 700 nanoseconds, the current went way down to an
average of only 0.05 amps at 30kHz.

Sean

At 09:51 AM 5/20/2003 -0400, you wrote:
>Sean:
>The IR2133 can source 200mA to turn on the IRF1405S. Assuming that you have
>no gate resistor the FET is going to take 1.3us to turn on
>(gate_charge/current) and 0.62us to turn off. I doubt cross conduction is
>the problem here. The overheating is most probably caused by the length of
>time the FET is spending in the active region.
>To do a rough estimate, I have assumed a VDS of 50V, a Load current of 5A
>and a switching frequency of 20KHz. With a rise time of 1.3us, and a fall
>time of 0.62us, you have turn on losses at 33watts..with a D2-pack you have
>to have a huge copper area on your board to get rid of this heat. When you
>are doing h-bridge designs, pick the FET you want and then pick a driver
>that can drive that FET; not too fast as you will have Di/Dt induced spikes
>on your gate and not too slow or your FET will be spending too much time in
>the active region. Then design your gate drive properly to get the correct
>rise and fall time from the FET. You can use a gate resistor or a resistor
>in parallel with a schottky to fine tune the rise and fall times of your
>FETs. There is a lot of thought that goes into well designed H-Bridges and
>you might want to think about over current protection, short circuit
>protection (phase to phase and phase to ground) etc.
>Regards
>Madhu
>
>-----Original Message-----
>Hi all,
>Does anyone here use any of the floating capacitor FET gate drive ICs from
>International Rectifier? How about the IR2133? If so, I'm having some
>trouble interpreting the deadtime specification that they give in the
>datasheet. Specifically, they say that the maximum on-off and off-on
>switching time is 750ns and the minimum deadtime is 100ns. However, they
>also show a graph which indicates that the deadtime is measured from the
>50% off time to the 50% on time, so that, it would seem to me, that given
>that the switching time is much longer than the deadtime, you would expect
>that both high and low side FETs might be on at the same time.
>
>I have a circuit using an IR2133 and IRF1405S mosfets in three half H
>bridges and I'm having cross conduction problems, seemingly because of the
>above situation. I think I'm having conduction of 100 amps for a few
>hundred nanoseconds each transition, which I'd like to avoid on principle
>but also because it is on the verge of causing thermal problems even
>without any additional load!
>Any ideas or suggestions?
>Thanks,
>Sean
>

--
http://www.piclist.com#nomail Going offline? Don't AutoReply us!
email listserv@mitvma.mit.edu with SET PICList DIGEST in the body