This issue can be completely avoided with good circuit design. There is no need to try and source MOSFETS capable of handling those spikes when easier methods to fix that exist. There are also PCB design options that can help mitigate the issues. (those fixes are exclusive to when you are using a PCB as they’re made by taking advantage of the fact that PCBs have layers)
1 Like
@GnomeMeansNo From the video it looks to me like the stormcore is off. Did you turn it off before you got in the uber or got home?
If so this would seem to imply something went wrong with the switch?
1 Like
Regardless of the size of the spike? Do you also work in electronics?
I don’t currently work in electronics but I do have a M.S. in Electrical Engineering.
I’d have to double check, but aside from extreme spikes caused by external sources, the methods should work for large spikes too. What counts as extreme varies based on the device, but, generally, it would be higher than anything the device could ever generate on its own. As well as higher than anything that could be produced by the planed environment. So, for example, PLCs can handle a lot more noise because they are made for factory environments.
1 Like
This interests me as well since I have a 60D on my desk with burnt precharge resistors and solder blobs around it.
I’m thinking the same thing happened in the case described in this thread. Could be that the precharge resistors were stuck active when the unit was on, and still active when turned off… heating up until desoldered itself… causing a short.
Mostly speculations though… pictures would be interesting.
1 Like
In my experience this is almost always a symptom and not the root cause. Probably phase or input mosfets, the BQ chip or in rare cases the small t1 and t2 chips. Have you checked resistance between phases? All the same?
You should design a controller 
4 Likes
60D in my case
1 Like
I didn’t check if it was responding to the power switch but it stopped responding to the remote before I ubered so I think it was well and truly dead already at that point.
I’m not in the habit of turning it off I just let the 10 minute timeout turn it off after I turn off the remote so wouldn’t have done that.
2 Likes
I’ll try and pull it apart this weekend.
3 Likes
Yes you are right. I think my controller has kept the precharge active for too long. There are burn marks in the thermal pad above the resistors.
The values are still fine though. The controller seems ok when bypassing the precharge fets.
I might just bridge those… will see. Don’t have the need for it at the moment so will delay the repair
2 Likes
The board might have been on, it was last used like a few days before and I’m not sure why all I know is the fire started at the capacitors.
I’ve seen this exact failure lead to the stormcore blowing up inside of a board while turned off with my own eyes.
Now if I see a stormcore go into that unresponsive state I unplug it, consider it dead, and send it back to LaCroix. They can deal with it.
4 Likes
I used to work as an ASIC designer, its the same as designing PCBs. Its just micro/nano level instead. Sure, everything you design is pretty much made of transistors but the CAD principles are pretty much the same. I’m sure in a decade or so, silicon fabs will be widely avaible for anyone. Just a matter of harnessing the knowledge and someone besides fucking Cadance to make a software to CAD wtih that doesn’t cost hundreds of thousands per licence/year to use…
3 Likes
Time to add a loop key to my vanguard I guess.
This is a good question. 5.3 is questionable to me
1 Like
Whatever they were coming with in Sept of 2021.
5.something
IIRC that was FW5.2
1 Like
