Those are, without a doubt, arcing marks. I’ve seen it hundreds of times with “mechanical” vaping devices.
The alcohol will clean off some of the carbon and debris but the “burnt” metal high points of the marks need to be removed because they just encourage more arcing. Scotch-Brite Heavy Duty pads can help some or emery cloth. The pits you cannot remove without taking off lots of metal.
The circular pattern is from the cells slowly rotating a little bit every time they bounce off the contact. As much as we might doubt that the cells are bouncing and moving the evidence is clear, these cells are not being held firmly in place.
Cells do not have a flat top or bottom contact and the NESE contacts are probably not flat either (convex most likely?). This results in just a few high points that are touching and when the contacts bounce this is where the arcing occurs.
The next time you use that setup place a mark on each cell when it’s in the holder and see if the mark is in the same position after you take a ride. If not, then you have the proof that they’re moving around.
Yeah, that is arcing alright. It’s an issue in most mechanical electrical switches/relays because when they engage the contact always bounces. I don’t know what’s the state of the art solution for it, but some switches back in the day had mercury wetted contacts so the bounce happens while the surface is submerged in liquid metal.
Sounds like something a small amount of copper foil (or any conductive metal foil) between the cell terminal and NESE contact would fix. Folded over itself a few times & shimmed in to conform to & bridge all high/low points.
Sounds like Winfly compression packs wouldn’t have this issue as it uses flexible braid instead of solid formed contacts
@rusins, have you checked your cell dimensions? Those 21700 are all over the place but all i know off are slightly longer. This is the first time i see this issue. I sold probably in thousands of modules and have nothing like this reported till now.
Could you post a photo with cells inserted of one module? Could you judge how much the black foam is compressed?
Well, I took a closer look at my modules, and noticed something crazy – 4 of the 6 modules are partially deformed! I’m guessing the heat from the arcing might have caused them to expand on one or both sides!
I measured a few of my cells – they’re 21.4mm wide, and 70.1mm tall. (although maybe they are exactly 70.0mm, since I had to put a piece of kapton tape on them to avoid shorting my cell as I measure with my calipers, lol)
(For the record, my brother’s 18650 nese pack also has what I think are small burn/arcing marks on the cell holder tabs. Just not as extreme. And no warping, or cell disbalance as of yet, thankfully)
The arcing creates excess resistance on the contact surfaces and you get ohmic heating during high amp draws. In time the gap will widen and arcing will increase.
Had the opportunity to test my board at the local bobsleigh track last week.
Youtube edit:
My motors should be set to 60A each, so I’m surprised that Davega is showing 62A max out of them. Also interesting that a hill climb like this only used 80A from the battery at its peak. If I had a robogotchi then I could look at the voltage to figure out if it was sagging or not, but it certainly didn’t feel like it was. Battery is now the 12s4p P42A pack that Simeon built for me. The XT-90 wires and balance wires are pretty short so it looks like a mess inside the enclosure, but I managed to get it to work.
Front truck with the reverse cone bushings is sketching me out, especially because I don’t have much grip with the current glass frit. Maybe new shoes will help, but I think it is what it is. Going to replace the front with Savage’s TKP truck to see if that feels better. (I’m 95% certain it will)
My VX1 keeps losing connection every now and then, seemingly on bumps. My connections should be secure in the board, so idk what’s up with that. Not sure if I should upgrade to the VX1 pro or just get a puck. Or, use my OSRR remote, but I need the 2nd UART port for that, which would mean using FW 5.3 Maybe…
Yes, I know the OSRR has a PPM mode, but there’s an upcoming feature I want that needs UART.
Anyway, I’m going on holiday for a week so I will check how much power the flexiBMS drains in that time. Only near-future upgrades for this board will be some sick lights
you might think theres a universal standard in tolerances for cells. Theres not.
Dimensions differ alot between manufacturers. Think we have three versions of cell holders for 21700 atm just to pass shake and vibration tests.
Great read @rusins ! - you have motivated me to get my switchblade running but I cannot give it the same level of attention to detail you nailed with this one.
Also cool to see that you can actually fit 12S3P in NESE cases under a board this size, but a bummer to see the issues you are having with the compression packs. Hope you figure out what’s going on with those. Is it possible the print was done improperly or with bad filament?
Took the board apart today to install a puck remote. I’ve given up on OSRR availability, and I don’t really need a remote with telemetry if I have Davega.
First things first, I wanted to check if I had a loose wire or something that would explain my VX1 FULL disconnecting. Well, it seems like I did an overkill job securing and insulating the receiver wires, so conclusion: not user error, VX1 is just crap. I also had the VX1 in my small board disconnect fully once, so I will be swapping that one out too.
I also wanted to make some space for the 12V buck converter, so I made an extender cable for the battery’s XT-90, that also splits off into a switch → fuse → buck → fuse → external accessory port.
Battery is secured with double sided tape and foam on the sides (it’s not closed cell, but it’s all I had), and then I hot-glued things like the flexibms and buck in place. This is the final enclosure layout: (it’s tight!)
Then I cut up a 12V side-emitting LED strip, and hot-glued it around the enclosure. It doesn’t seem to hold well, so there’s a big chance it’ll fall off. Will have to check up on it while riding. Worst case I’ll use epoxy, but that means it’ll be permanent
Once the threadlocker cures I’ll ride to a more scenic place for final pictures (still have to powder-coat the duck handle too!), but here’s a few pics in the meantime.
The puck color doesn’t quite match, but I might swap it out for a different shell anyway. I also bought the glow-in-the-dark shell, but it’s very underwhelming, like most glow-in-the-dark things are. Thanks for reading!
Maybe…
