• State TI Thinks Resolved
  • Locked Locked
  • Replies 10 replies
  • Answers 1 answer
  • Subscribers 29 subscribers
  • Views 378 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.

Original question:

DRV8844: Part blew up when output is connected to +48V while VM is not powered

DRV8844: Continuation: Part blew up when output is connected to +48V while VM is not powered

Tor Sovel
Prodigy 160 points
Part Number: DRV8844

This is a continuation of investigation about this issue:

https://e2e.ti.com/support/motor-drivers/f/38/t/837153

So I added an external 2Amp diode (Ifsm=150A) in parallel to the internal free wheeling diode from OUT to VM to have the high Inrush current flow through the external diode. The DRV8844 output still gets damaged. My initial theory that high inrush current that charges the bulk VM capacitors is not the problem here. I removed all external capacitors that are connected to VM and with or without the external diode, the DRV8844 output is still getting damaged.

I don't think that the outputs are protected for case like this, GND=0V, OUT=48V, VM=open. Can you please verify and let me know what is causing the damage?

We have a customer ready to order production samples but we can not move forward with this problem unresolved.

Thanks,

  • 0
    TI__Guru** 114905 points

    Hi Tor,

    Sorry to hear you are still experiencing damage. We have contacted the design team for comment. Please expect a reply by Friday.

    Can you provide some additional information?
    What is the fully charged battery voltage?
    Is the battery connected directly to the output without a load?
    If there is no load, what is the wire length from battery to DRV8844?
    Can you provide a full schematic? If needed, we can make arrangements to share privately.
     

  • 0 in reply to Rick Duncan
    Prodigy 160 points

    The fully charge voltage is expected to be 56V. Nominal battery voltage is 48V.

    We are using a 48V 12.5A power supply for our bench testing. https://www.digikey.com/products/en?keywords=SE-600-48

    The 48V power supply output is directly connected to one of the outputs. There is no load attached and VM pin is floating. 

    Wire length is about 5 feet between power supply and DRV8844 pins.

    Schematic is provided from previous thread but here it is.  https://e2e.ti.com/resized-image/__size/1230x0/__key/communityserver-discussions-components-files/38/DRV8844-SCH-48V.JPG

    Thanks.

  • 0 in reply to Tor Sovel
    Prodigy 160 points

    Rick,

    Also as a follow up, do you know the internal diode characterisitics of the DRV8844 like forward voltage? I used STPS2150 Schottky diode which has a Vf ~ 0.8V and I was thinking that maybe if the Vf of the DRV8844 is lower than 0.8V, majority of the inrush current is still going through that internal diode causing it blow up.

    Thanks,

    Tor

  • 0 in reply to Tor Sovel
    TI__Guru** 114905 points

    Hi Tor,

    Sorry for the delay. Design has been contacted and requested that I perform a few experiments on my side to help identify possible causes.

    We should have another reply by Wednesday.

    Regarding your question about the Vf of the diode, I was expecting the schottky to have a Vf of less than 0.8V. The internal body diode is about this same value.

    If you have schottky diodes with a Vf of 0.3 to 0.4, please try those while we perform our experiments here.

  • 0 in reply to Rick Duncan
    TI__Guru** 114905 points

    Hi Tor,

    Have you been able to locate and try schottky diodes with lower Vf?

    I ran a quick experiment on the DRV8844EVM this evening. When plugging 8V into OUT1, there is a current spike of ~18A. This spike is expected to be much larger at 48V.

    There is an instantaneous power of over 12W (18A * Vf of .7V).

    I did find some schottky diodes that I plan to add tomorrow. I plan to start with two MBRS360 in parallel and measure the current going through the schottky diodes. If I am able to re-direct the current through the schottky diodes, I will then raise the voltage at OUT1 to mimic your setup.

    Please expect the next update by Friday.

  • 0 in reply to Rick Duncan
    TI__Guru** 114905 points

    Hi Tor,

    As mentioned above, I took two MBRS360 schottky diodes placed in parallel from OUT1 to VM. The forward voltage of these schottky diodes are ~.35V at 25C, with a peak current of 125A.

    Using these two diodes I was able to raise the voltage of OUT1 up to 20V with no problems. I stopped there because I can no longer measure the current flowing the body diode using the preferred method.

    At 20V, the inrush current is 46A, with less than 4A flowing through the body diodes.

    Next I removed the current probe showing the total current, and measured the body diode current only.

    I was able to power the device multiple times at 48V, with only 7A passing through the body diodes.

    So assuming the inrush current through the body diodes is proportional at 48V, the current through the schottky diodes is ~ 80A. If the 80 Amps is passing through the DRV8844 body diodes, this would be a transient power surge of >50W. By putting the two diodes in parallel from OUT1 to VM, the transient power through the body diodes is reduced to ~5W.

    Your thought that a high Vf  for a schottky diode could have a negative impact is valid. Please try a schottky diode with lower Vf.

    Also, please be careful when testing. There should be some cooling time between current surges through the body diodes and other components.

  • 0 in reply to Rick Duncan
    Prodigy 160 points

    Hi Rick,

    I was not able to test with a lower Vf diode since we don't have this diode currently available. We have to order some in the next few days.

    Do you think a die design tweak is possible to solve the problem? We really have a very limited PCB area and adding an external diodes may not be our best option for this particular product. The biggest external diode I could use is a 1x SOD123FL per output because of this space constraint.

    We have similar design using Infineon BTS723GW high side driver and we don't encounter such problem. We tested this Infineon part at 56V connected to its Output and the part still operational after the test.

    I really like the DRV8844 since it can be High side, Low side or HBridge driver using the same chip.

    Thanks.

  • 0 in reply to Tor Sovel
    TI__Guru** 114905 points

    Hi Tor,

    Sorry, there are no plans to tweak the die.

    The BTS723GW is rated to higher overcurrent, which probably is the reason that the BTS723GW did not experience the problem.

    There is another thing you could try assuming you have an unused output. You can try connecting two outputs in parallel, which would reduce the current in each diode to ~50%. There could still be a risk to other parts of the device.

    That is why the external schottky is the preferred approach.

  • 0 in reply to Rick Duncan
    Prodigy 160 points

    I did the 2 outputs in parallel plus 1 external diode (Vf~0.8V) during my initial test before and still did not solve the issue. I will try it with the lower Vf diode when it arrives next week.

    Thanks.

  • 0 in reply to Tor Sovel
    TI__Guru** 114905 points

    Hi Tor,

    Please keep us posted with your results.

    If possible, please monitor the current amplitude and duration through the diode and the device when experimenting.

    In my experiments, the current was less than 8A through the DRV8844 for less than 50us.