• State Resolved
  • Locked Locked
  • Replies 5 replies
  • Subscribers 33 subscribers
  • Views 1897 views
  • Users 0 members are here
Support feedback
Options
Options
Related

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.

Breaking when driving the gate of an IGBT

Nahuel Manterola
Prodigy 55 points
Part Number: UCC21520

Hi all,

I'm using the UCC21520 to drive the gates of two IGBT. These IGBT's switch a high voltage capacitor to a solenoid. However, the module breaks after driving the gate a few times, even when there is no voltage on the source of the IGBT. I'm using 5V for the VCC and 18V for the VDD.

When the driving breaks, only one of the two channels breaks. The other channel can still be used until it also ends up breaking. When driving a broken channel, the UCC tries to get the gate to 18V, but the internal pull-down MOS also seems to get shorted. The voltage on the output pin and the 18V drops, until it goes into under voltage lockout. 

The input signal goes through an isolator, the ISO7241cqdwrq1. The input of this isolator is controlled with a button. Could a bouncing input signal from this button be a problem?

I've attached the schematic and the layout on the PCB. I highlighted the paths from the isolator to the driver and the output of the driver to the gate of the IGBT's. I believe these traces should be short enough not to cause any issues.

Thanks for your time. If you have any questions, I'll try to respond as soon as I can.

  • 0
    TI__Expert 3355 points

    Hello Nahuel,

    Thank you for your interest in UCC21520. I'm an applications engineer in the High Power Drivers group and will help you get to the bottom of this.

    First off, the layout looks good - and since it is working for a short period of time, I do not think layout would be the cause of the immediate issue. Is this issue repeatable after power cycling the whole system? These chips are definitely fast enough to send any bouncing signals to the output, so a debounce circuit on the input side couldn't hurt - especially when switching high power devices. I also don't see any external gate resistance or gate to emitter resistance, which we recommend to help combat output overshoot and ringing, as well as safeguard against induced gate voltage when switching elsewhere.

    By the looks of the layout, both VDDA and VDDB share the same 18V rail. How is this 18V generated? How fast does the 18V drop when the channel breaks? When the voltage is dropping, does the 'broken channel' still respond to inputs, or is it stuck on/off?

    Which IGBTs are you using? Some IGBTs need a negative Vge to turn off. Could you please share the IGBT portion of schematic so I can take a closer look at the grounds and supply of the HV?

    Best Regards,

    John

  • 0 in reply to John Geiger
    TI__Expert 3355 points
    Also, could you provide oscilloscope shots of the outputs when switching normally and when the channel breaks?

    Thanks,

    John
  • 0 in reply to John Geiger
    Prodigy 55 points

    Hi John,

    On your advice I tried to debounce the button input, and settled on just using a microcontroller to switch the UCC. This greatly increased the lifetime, so I think this was the main reason the device didn't work.

    We tried to manually add a 10ohm gate resistor, which didn't solve the problem. Since the device still worked we decided to implement this gate driver into our system, just to be safe.

    The 18V is generated by a flyback charger, which can deliver up to 300mA at a time. This is stored in a capacitor. Since in normal use the 18V doesn't drop, I assume this part is working fine.

    I'm running stress tests on Monday, so I'll let you know how stable the system is afterward.

    I've included the schematic of the IGBTs, if you need anything else, please let me know.

    Thanks for your help!
    Best regards,

    Nahuel

  • 0 in reply to Nahuel Manterola
    TI__Expert 3355 points

    Hi Nahuel,

    That's great that adding a debounce/MCU control helps to alleviate some of the issue. When you say that it greatly increased the lifetime, does that mean it still breaks eventually?

    Did you get any scope shots of what the output looks like after the 10 ohm gate resistor is added?

    Looking forward to hearing how the stress tests go.

    Best Regards,

    John

  • 0 in reply to Nahuel Manterola
    TI__Expert 3355 points
    Hi Nahuel,

    I'm going to close this thread as it's been open with no response for a while now. I hope you found your solution - feel free to comment again if you have anything else to add.

    Best Regards,
    John