Because of the Thanksgiving holiday in the U.S., TI E2E™ design support forum responses may be delayed from November 25 through December 2. Thank you for your patience.

  • State
  • Locked Locked
  • Replies 4 replies
  • Subscribers 67 subscribers
  • Views 1202 views
  • Users 0 members are here
Support feedback
Options
Options
Similar topics

This thread has been locked.

If you have a related question, please click the "Ask a related question" button in the top right corner. The newly created question will be automatically linked to this question.

Mosfet Driver problem

Mike Kehrli
Prodigy 20 points

I'm using the UCC27322D mosfet driver.  It's output is rated for a peak amerage of 9 amps.  I've been bench testing one for a number of hours, and just now smoked it.  I mean it was glowing cherry red before I turned it off.  But it has worked just fine for about 6 or 7 hours so far.  When I put a laser temp sensor on it after a couple of hours, it was reading at like 85F or so.  Barely warm to the touch.

Now, this may be the stupid part, but I'm driving 6 mosfets with it.  Theoretically that shouldn't be a problem, but maybe that's the whole problem.

Here is the schematic. Vdd is 12 volts.  F1-F6 are fuses, which I'm using on the prototype just to be able to isolate 1 or more of the FETs during testing.  The FETS are International Rectifier IRFB4310ZPBF.

Now, what you don't see in this schematic is the fact that I'm getting some big pulses back to my mosfets.  I get 75 volt spikes across my source and drain pins on the mosfets, for every output pulse.  When I scope the power to the driver, or the input or output pulses to and from the driver, I don't see any of these spikes getting to the driver.  Both the mofet gates and the driver input/output lines are very clean.  So, while I don't see how the spikes can be bothering the FET driver, this data needs to be included here.

Is it just stupid to drive so many FETs with one driver?  On paper, it seems fine, but then, on paper my FET driver didn't glow cherry red.

  • 0
    TI__Expert 4375 points

    Mike,

     

    What is the switching frequency you are driving the FET's with?

     

    You should always ensure that the driver is as close to the FETs as possible and that the driver output to gate trace is short. I do not recommend an inductor in this trace. The whole idea of short traces is to lower the parasitic inductance which in turn reduces voltages spikes caused by fast di/dt in the trace. These spikes will damage any driver. So I would remove L2.

     

    Do you have any waveforms?

     

    Richard.

  • 0 in reply to Richard Garvey
    Prodigy 20 points

    Hi Richard,

    Thank you for the reply.  Sorry for the delay.  I had to make another prototype as I had smoked the traces to the FET driver.  I've got some scope images to add now. 

    This first two show the output from the driver chip.  The first one is 500 uS per div and the 2nd one is 5 uS per div.  This just shows that the signal is not being shut off cleanly.

    These next 2 show the situation on the source to drain pins.  As you can see, there is quite a capacitive affect.  The first image shows 1X setting and the 2nd shows the 10X setting on my probe.  The amplitude of the spike is apporx 75 volts + and 10 volts -.

    I will certainly try moving the FET driver down close to the FETs.  Is there anything else you can see that I'm mising on this?

    Note, I've removed the inductor from the circuit before taking these images.

    Thanks,  Mike

  • 0
    Prodigy 100 points

    With no resistance between the gate driver and the gates, wouldn't the current be theoretically infinite?  I think I would like to see at least a small resistance here.  Maybe an ohm.  Or maybe one 250 milli-ohm resistor per FET.  

    Also, you mention big pulses on the drains.  Could it be that you are coupling this back into your driver through the gate-drain capacitance?

  • 0 in reply to MARK SULLIVAN
    TI__Mastermind 18200 points

    Mark, there is an internal gate resistance of .7ohm per FET as well as an internal impedance on the driver. 

    Calculating the power dissipated driving the FETs comes out to very low, on the order of a few mW, so this should not be your problem. Furthermore, the fact that it ran for so many hours without issue seems curious. Again, I believe this to be a driver question.