• State
  • Locked Locked
  • Replies 8 replies
  • Subscribers 31 subscribers
  • Views 267 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.

DRV8770EVM: U1 Component Failure

Danny Turner
Prodigy 30 points
Part Number: DRV8770EVM

Tool/software:

We have been using the DRV8770EVM Evaluation Board as a pulse generator. Our setup consists of a Keysight Waveform Generator, with the + connected to INXA (both INHA and INLA shorted together) and the - connected to INXB (both INHB and INLB shorted together). The MODE pin is connected to GVDD. The MOTA output goes to the high side of our load and MOTB to the low side of our load. 

We are using the external GVDD source and apply 18V to the board. 

We are constantly grounding the INXB side, which is enabling the low side FET (connecting MOTB to GND), and we pulse the INXA side, which enables the high side FET to deliver the supply voltage to the load. 

The waveform generator is outputting a single 0 to 5V square pulse. There are a wide array of pulse widths that we use, but they are on the order of milliseconds.

We have seen failures of the U1 component on 4 out of 5 of the DRV8770EVM boards that we have. The component has visible damage, near the center of the part. 

Any ideas what could be causing the failures of the U1 component? Have you seen this in the past? 

We noticed the logic input voltage is a function of GVDD, with a maximum voltage of GVDD +0.3V listed in the datasheet. If GVDD were not applied (0V) and our waveform generator outputted 5V to the INXA pins, could this cause damage?

  • 0
    TI__Genius 11651 points

    HI Danny,

    Thanks for your question.

    We have a couple of follow up questions

    1. Is there a load connected to the board?

    2. Have you checked on what kind of damage shows up in the device(like does any pin shorts to ground or open up compared to the good unit)

    Thanks,

    Ibinu

  • 0 in reply to Ibinu Alaudeen Nadeera
    Prodigy 30 points

    Hi,

    1. Yes, there are various loads connected throughout our testing. The case where this board failed, the load was drawing about 30mA. The load was a zener diode that has a 500 ohm series resistor. The supply voltage was set to ~30V and the Zener voltage is 16V and a ~20msec single pulse was sent. There are other cases where the load is just a resistive load, drawing a few amps for tens of msec.

    2. We performed a lot of resistance measurements and the one that sticks out is INHA appears to be shorted to GND. It measured 0.9 ohms. There were some other anomalous measurements, like INHA to INLB measuring 171 ohms (as opposed to ~2 MEG on the good board), INHA to INHB measuring 273 ohms (as opposed to 1.6 MEG on the good board), among others, but I'm guessing the root of the problem is the INHA being shorted to GND.

    Thanks,

    Danny

  • 0 in reply to Danny Turner
    TI__Mastermind 19060 points

    Hi Danny,

    How long is the dead time? As you stated above, the low-side FET is left ON while sending PWM to HS FET. The deadtime needs to be longer than PWM pulse width. If DT pin is connected to ground, then the fixed deadtime is 200ns when high and low side inputs are both logic high. 

    Best,

    David

  • 0 in reply to David Medis
    Prodigy 30 points

    Hello,

    Looking at the schematic of the DRV8770EVM board, the DT pin is tied to ground through a 0 ohm resistor. We did not modify this at all, so the deadtime is 200nsec. 

    In my above statement, the low-side FET is left ON only on the B side (INHB, INLB) while the A side is pulsed (INHA, INLA) with a signal. We are controlling per Table 7-3 in the datasheet. There shouldn't be any case where the high and low side FETs are both enabled on a given side (A or B), since we are in inverted mode. 

  • 0 in reply to Danny Turner
    TI__Mastermind 19060 points

    Hey Danny,

    Please allow me some time to recreate this issue in the lab and I'll reach out when I know more. 

    Best,

    David

  • 0 in reply to Danny Turner
    TI__Mastermind 19060 points

    Hi Danny,

    Apologies for the delayed response.

    I was unable to recreate the issue in the lab. If possible, could you provide or capture the input and output waveforms on an oscilloscope? This would greatly help in diagnosing the problem.

    It does seem like the input resistance of INHA is low. It's possible it could have somehow been shorted to ground.

    Best,

    David

  • 0 in reply to David Medis
    Prodigy 30 points

    Hi David,

    I will try to get some oscilloscope plots for you, from our working unit.

    For some clarification, the boards have been working for a long time (many months) before we started to see them fail one by one. I suppose it is possible the same error was repeated across multiple units in that time period, causing them to break, but I am not exactly sure what that error could be. Our test procedures have been unchanged, so we are not altering the setup. 

    Thanks,

    Danny

  • 0 in reply to Danny Turner
    TI__Mastermind 19060 points

    Hi Danny,

    Sounds good, looking forward to analyzing the waveforms. The waveforms will help to simulate your setup and hopefully find the issue. 

    Best,

    David