Flyback Voltage Suppression

[poflynn]

Hi All,

I'm pretty new to all this stuff so please go easy!

A friend of mine is having trouble with her supplier for a BMS (Battery Management System) module used in an electric tuk tuk. 45% of the 200 units in the field have so far failed. I (A lowly final year student) went to take a look at the BMS module and there appeared to be no transient suppression at all on the board. I setup a simple test where I spun the motor up and jumped on the brakes to try and induce an inductive spike. I measured close to -1kV across the battery terminals and all the chips on the BMS were immediately fried. These tuk tuks are used in fairly rugged terrain (In Africa) so I figured this was likely to be the cause of the issue where the motor gets locked up when the wheel hits a pothole. (There is a higher rate of failure in areas with worse terrain).

I spoke to the Chinese suppliers about it and they agreed that some sort of transient suppression was required. Fast forward a few months and they have proposed this: They want to use the BMS MOSFETs to disconnect the the negetive side of the battery when a transient is detected. I don't really like this since surely that doesn't allow the transient to go to ground and could cause an even higher spike? Also would that not cause the high voltage to appear across the MOSFET terminals? I don't really understand why they can't just use a flyback diode or something that will send the transient to ground.

I wanted to check here before I said it to them because maybe I am missing something?

Thanks in advance for any comments, suggestions!

P.S. my degree is mostly based around digital design so this is not my area of expertise!

[jonpaul]

what is a BMS? Is module a motor controller or mains charger?
We need to see   the design, schema and waveforms before a solution can be suggested,

Jon

[T3sl4co1l]

Just use a bigger battery?  One that can handle the inrush current during regen braking.

It's not clear how you made the measurement.  It can only be transiently, but a DMM will not provide indication of that; you might see a number flash by but you have no idea where along the peak that value was sampled.  A high frequency chart recorder (available in more advanced DMMs) can plot this, or an oscilloscope proper.

Or what "-1kV" means.  Was it positive beforehand?  What's the default value anyway? (240V? 400V? 48V? 12V??!)  Was it negative beforehand (probes swapped) so you're measuring an overvoltage?  Or did it actually fully reverse?

Or what the measurement is.  Is it a known-good brand of meter?  Or one of those off-yellow crap boxes?  Do you know that it is immune to RF interference?  (A BMS and/or drive without basic braking protection does sound like it might not pay any attention to EMI filtering either.  A couple hundred volts of switching noise will challenge even name-brand meters!)

Even so, a "large enough" battery is not required, but then one needs a braking resistor, and system to control it.  (Perhaps this is already present, I don't know.  The supplier not suggesting components to support it, sounds like no.  But who knows; the supplier probably doesn't know anything about the module, and the manufacturer only slightly more than nothing.  Lots of cheap units like this are made with little design consideration.  Scarily, many products manage to work despite that; or work for long enough to be useful, before succumbing to what would otherwise be critical design flaws that happen to trigger rarely.)

Tim