Reposting in this forum because it is more active.
As the Subject suggests, I appear to have smoked my VESC today. It is a VESC 6 from TRAMPA that I have been using to balance my custom onewheel build. I have just recently got to the point where I can actually move my motors and what not, but I have always been having the balance app go into 'fault' mode, which basically just disables outputs to the motors as far as I can tell. I haven't been able to get very good logging yet to figure out what my actual current draw is and what not, but based on some tests I did, I think for the majority of the time I run this, I am only drawing ~20-30 amps, but I have large current spikes that go above 100 amps.
As to what happened, Yesterday I added a 'fault dealy' to try to see if that was part of my issue. The onewheel worked a lot better and I was able to ride it maybe 50 yards continuously. After those 50 yards it was running way worse than before. I decided that was enough testing for the day and let it go to sleep in my trunk. Today when I tried to turn it on, it didn't. I tried to connect via USB, and it also wouldn't connect, and said I needed to make sure the vesc had power etc. I could visually see none of the LED's on the VESC on. At that point I tried disconnecting and reconnecting the battery. When I started to reconnect the battery there was a spark along my quick disconnect. My 60 amp automotive fuse didn't blow but I could smell some magic smoke(no visual smoke). At this point I checked the battery volage and it was in the middle of its range at 23V, and I disconnected the motors. I tried one more time to connect power, again sparks and I smelled a little more magic smoke smell(but no visual smoke). I took the VESC apart a bit and the smell is coming from the fett/shunt area. Has anyone seen this before? Is it likely to be fixed with just a resoldering? and what do I need to do to stop it from happening again?
I can provide some information on what currents/duty cycles I was using if you request specific numbers. I am very new to all this VESC stuff so I don't know what all the numbers mean yet.
I also have a followup question. I want to verify what parts are bad. The easiest way to do that is to measure the resistance across the outputs of the mossfet to make sure they are shorted in the correct places and disconnected where they are supposed to be. The mossfet has solder pads underneath the chip currently that I cannot probe, are there other locations on the board I can probe the same points? Basically I want to find places I can probe for the gate/source/drain of every mosfet, if possible on the vesc 6 MKIII.
Do you already contacted TRAMPA support? Sounds like a hardware issue that should be covered under their warranty.
For me your first problems before the smoke sounds power switch related.
I think the MK III schematic is not released yet. But I guess the older VESC 6 schematic can be used to find annotations for the gate resistors for an example and they haven't changed in the MK III version. https://vesc-project.com/node/311 Not sure if they haven't changed, but that would be my first try because gate is hidden under the DirectFETs.
I am talking with them now, but they say it's unlikely to be a hardware problem, as those are very rare on this product line. I tend to agree that it doesn't feel like hardware issue, im not ruling it out but it also is incredibly hard to test for that. So for the time being I am ignoring broken hardware, and trampa support doesn't sound like they would replace it anyway.
I found that I could probe the drain and source of each mosfet by probing the motor channel outputs a/b/c, power, and ground since those should all be equipotentials. What I found is that all 6 mosfet's are broken, or my test setup was wrong. That is even more strange because I wasn't even using the B phase. I am running 2 brushed motors in parallel off of the A and C phase.
Does the VESC have built in protections in the firmware or in the VESC tool to keep the user from overstressing it? I know a lot of hobbiest stuff enables the user to relatively easily break itself in favor for more customization, but the VESC tool seems polished enough it may have had some of that built in. For example if I set the current to 200 amps continuous, would the vesc just cook itself, or would it shut itself down before it actually broke? If it would shut itself down, then I would start to consider hardware issues again.