It is a sweet pickle for sure, I cant wait to hear your report on the TB drives and overall performance of the board.
1 Like
Took a short ride today. Itās got some torque haha. Iāll report more this weekend
1 Like
Iām fusing balance wires on new builds now.
https://www.arrow.com/en/products/0251003.nrt1l/littelfuse
3a fast blow. Size and shape like a resistor. Was 25 pennies at the time I bought them. I solder 1 leg to the p-group, and splice the other end to the balance wire.
If you had these in, youād have had less mental trauma, but more surgery to do. 
As for what to do now, maybe you can try estimating wire resistance on each balance lead by applying a load and measuring voltage drop. If the suspect wires have a much larger voltage drop, cut your baby open and replace those lovely ribbon cables?
Bonus content, DoctorBass from ES.
Skip to 8:25 if youāre impatient.
(I want one of those power supplies!)
4 Likes
Good idea. Iāll do that. Thanks!
1 Like
Specifically I mean applying a 1A (for example) load across each pair of balance wires and comparing voltage against no-load voltage. Something like that. Iām thinking, you can use a cheap electronic load like for USB power banks, or a resistor of appropriate value.
Do you worry about nickel slicing into your balance wires when you run the wires across the tops and bottoms of the cans?
1 Like
No, the nickel lays low on the top of the cells 2 layers ( double insulators on top due to the + button being higher on the 20700) and the flat braided copper wire + solder fills in the top well, I have found a great silicone wire 22g with a high temp/ abrasion resistance. I know splitting is a concern, but that is usually a formula issue in the silicone. The braided copper wire is the key tinned and tightly braided, low resistance, low heat. The wires (all) are silicone covered and thicker sleaved and every test we did all shorts stayed inside the casing.
I do see your point and we inspect the pack and place fish paper where itās needed. Everything is done by hand and tested. The intent is to build a high-performance pack that doesnāt hold any heat. In our experience under built batteries that are too insulated create situations that shorts and meltdowns can occur. Fish paper holds heat, heat tape as well, hot glue does to and can melt and run onto your electronics and fry them like this:
What you dont see is something we borrowed from @DerelictRobot ā¦
2mm bullet connectors for all balance leads and temp sensors.
Small heatsinks on the MOSFETs (both sides) for charging and discharging (if used)
A bigger fuse for the bms 100A and a 10A fuse for charging
The quality of the materials used from cells to solder, the quality of work and patience in doing it well makes the difference and to be honest the only difference that mattersā¦
9 Likes
I did a quick experiment this morning. I took a 50 Ohm resistor and used it as a load for each cell, one by one. I measured the voltage drop across it with my multimeter. This put about 80 mA through the resistor. These are the values I measured:
Thereās no significant deviation. I believe that the length of the balance wires is a factor, which explains the differences. When measuring voltages on the cells directly, all values were 4.090 Ā± 0.001V.
I would have used more current but I only have shitty resistors below 50 Ohm. I could maybe go to 25 Ohm and 160 mA with those but no more than that.
1 Like
Quick question. My smart bms allows me to choose how much precision I want in the balance voltages. How much would you suggest? I love precise stuff so I decided on 0.001 . Could this be a cycle killer?
Not sure if the bms can measure that accurately. If I check mine while balancing that the values jump a bit from time to time, so I assume itās more in the 0.01V area. If you set your bms that exact you should switch on active balancing and not only balancing while charging. The time while charging is too less to get your cells that close balanced.
One thing I could imagine than is that the bms try to get as close as possible to that value but never reach as the measurement is not that accurate and with it drain your pack till cut off.
I think itās more than enough to set it to 0.01V.
1 Like
Which BMS is this? Might want to give one a try myself
Itās a 12s xiaoxiang smart bluetooth bms
I guess i couldāve read the name in your photo 
thanks!
Xiaoxiang is the company that made the app. You can you use the app with any bluetooth bms. I just went for one of their bms because that way I was certain of the app working properly
1 Like
Upgrades have arrived, final decisionā¦paid wayyy to much for a battery I thought was ready to ride.
Positivesā¦learning how to repair will help learn to build next one for waaaaaay cheaper
3 Likes
I would recommend not leaving it on active balance charge balance should be enough to properly balance your battery and active balance will āhideā any issues with your pack. Also balance of 0.005 is more than enough.
If you charge with 8-10A like I do itās not enough time for the bms to balance the pack probably. Even if there is only a bit drift. The balance current on this bms is not really much and need ages.
Also if you check the voltage when ever you switch on charger you will see if one group goes bad, plus the bms will not balance while charging, so you will definitely find out that something is wrong when one group goes bad as your range will become less.
1 Like
If you set the balance charge current low enough so that It starts balancing at 3.0V 100ma is more than enough for balancing IMO.
As far as active balancing goes isnāt it on all the time? doesnāt matter if youāre charging or discharging?
Itās just unneeded Imo Iāve heard people have had the bms go into a balance loop and drain an entire pack because of it.