Short to GND after hooking up Battery to Aux

#1

Hi everyone!

I just recently purchased the 48V Version of the odrive to power a small electric vehicle.
Really love the feature to power two motors from one board. Closed loop yaw control really makes a difference in driving dynamics!

I used Benjamin Vedders VESC before and hope to get a little better control with the odrive, mainly due to the encoders used. Yeah, I know that the VESC also supports encoders, but they have always worked a little sketchy for me.

Long story short, I was hooking up all the electronics last night. I guess I was a little tired, as I hooked up my battery (~36V LiIon) to the AUX Channel where the braking resistor goes, polarity was not reversed though. Visible results are a blown 30A Fuse in my Battery leads and a hole in one of the DRVs.

Two Questions:

  1. What happened?
    If I understand the schematic properly, current should be going from the AUX+ through the body diodes of the High Side MOSFETS to the DC Link and should then be charging up the DC Caps.
    Should neither cause excessive current flow nor any other problems, right?

  2. Whats broken?
    I checked all current rails on their own:
    3V3 is fine.
    5V seems to have a short to GND with ~3Ohms
    DC Link, dont know, because the DRV 5V Buck causes problems obviously.
    Changed both DRVs, no change in symptoms.

I am currently lack Ideas what the Issue could be?

Bests and thanks,
Jakob

0 Likes

#2

That is unfortunate.
I agree with you, as far as I know your analysis about charging the DC bus with the body diodes should be correct. I don’t know what else could have happened.

Did you measure the 5V rail with the M0 DRV chip desoldered?

0 Likes

#3

Hey Oskar,

thanks for your input. Always good to get some confirmation on my theories :slight_smile:
I really dont know what happened. Maybe we messed up something else.

Will try to check the 5V Rail again with DRVs desoldered tomorrow or so.

Thanks,
Jakob

0 Likes

#4

Hi all,

did some further research :male_detective:

  1. Desoldering M0 DRV does not change anything on the behavior described above. All of the following measurements are with the 3V3 LDO desoldered.
  2. Current draw has the characteristic of a diode. No current till ~0,7V and then the current rises sharply. At 500mA we have around .8V.
  3. Heat seems to come from the DCDC area. I only got a rather cheap thermal camera so I can not tell for sure from which component.
  4. If I desolder the coil from the DCDC Converter, Current drops to around 130mA. Seems a bit much, but still reasonable?
  5. 5V Current drops to 90mA if I populate the 3V3 LDO again.
  6. Carefully started populating the DRV and the DCDC components again.
  7. As soon, as I place the coil, we have a short to GND again.
  8. WTF? Thinking about a short in the output cap, but then remember the diode-like behaviour of the 5V rail.
  9. Turns out that POS Diode of the 5V Buck was placed in reverse. Doh!

I can not tell for sure, if the diode was placed like this from the factory. I strongly think so, as I always placed it the way it was before and because the behaviour was always like this.

Really happy that we fixed the odrive!

Bests,
Jakob

0 Likes

#5

I’m glad you were able to fix it!

0 Likes