| Electronics > Projects, Designs, and Technical Stuff |
| Panasonic NKY467B2 36V 15AH 540Wh. Ebike Li ion upgrade, burning my father'ass? |
| << < (9/9) |
| Siwastaja:
IIRC the IC markings were not visible in the teardown photos. You could try to google for them, they were widely available a few years back. It's possible it's some COTS management chip, or it might be custom. Single cells wouldn't be replaced in a paralleled bank of cells for numerous reasons. It would be very expensive at least, and a slow process since the swapped cell must be charged/discharged to the same SoC. Packs also tend to be glued, dipped or sprayed with some kind of goo. The takeaway here is that if the cells are not reliable enough to build reliable packs, then just don't use them at all, you won't have a business with the failing cells. 1 serious cell failure within a TESLA pack probably is a showstopper for that pack. They can't remove it from the equation. If it shorts out completely, sure, it'll blow the fuse wire. If it starts to leak a tiny little bit, it's not going to matter. But anything else, and the pack is unusable as a whole. So, if you want to have returns below 1%, you need to have cell failures below 1/700000. This is very well possible given Panasonic's quality control. The failure rate must be low enough that you can replace the complete pack - maybe one series module in some cases, but probably not with TESLA. I'm sure they'll closely analyze any failures and keep the process in control with Panasonic (and their new plant). |
| Zucca:
I installed the chinesium parts... well one was good, the other one failed on me. :horse: First this one: very pleased with the device, here the look-up table: 31,76V 0% 34,04V 10% 35,29V 20% 36,27V 30% 36,88V 40% 37,50V 50% 38,10V 60% 38,82V 70% 39,49V 80% 40,30V 90% 41,40V 100% there is also a "secret menu" which let you set the back light, standby and cycle between % and V display plus other two settings which I did not understand (of course where I can find some pdf on it???). Anyway it work(ed)(s) well. Double finger crossed. This one: was at the end a complete fail. First the heat sink was 0,25mm away from the mosfet... I had to rework that. Then I tested it with about 10A load and it was working ok, also the charging was ok. It even cut the output at about 30V in the discharge test. Also the self discharging test was ok with that BMS: Of course after the final soldering actions and after closing the battery case, I did the final test. Boom, failed. I got only 17V open (0V with 1 to 10Aload) on the output with a 41V battery. Tried to disconnect, reconnect. Nothing. Toasted. I asked a refund to ebay, will see. I just applied a protection spray for wet environment for electronic before the final test: but I don't think it could kill that stupid board. Well it failed once also before (same shit) but when I connected the charger, it fixed it. Anyway there were signs of hand soldering rework on that board.... it did not inspired any good from the beginning. Finally the B- cable on the sense cell connector was a straight short to the big B- pad... why then a cable there? To externally connect what is already connected on the board? :palm: (PS: I used that cable to connect the battery gauge, so I was happy for that no sense... >:D) Trouble-shooting a 8€ chinesium board? #FORGETABOUTIT I got this one: eBay auction: #273019581068 please God send me something which works fine this time. Anyway I solved the mechanical problem with some bad ass hot air melting process, here some pictures of the Frankenstein battery. |
| Zucca:
So I did it! Not a pretty job, but holy cows that bike now flies. The battery is so much lighter and equal if not more powerful. Li-Ion for president! My mom is scared to ride it >:D ... |
| Navigation |
| Message Index |
| Previous page |