Top balancing
This kinda deserves a good post on its own:
When I received the cells, as mentioned in the OP, they all tested to 3.295/.296. But, I got antsy and wanted to start playing around. I hooked up 1 string of 16s to my inverter, charged some, discharged some, didn't have bus bars torqued down well, and un-balnced that string from the other 16. Would not advice, but end of the day not a big deal.
Here is how I top balanced these cells:
1.) I made custom busbars for the 2p16s setup. Doing this with a handdrill was a 1/10 experience, could not NOT recommend enough. I actually ordered a drill press so I can come back and finish rounding out some of these holes a little bit as the fit is somewhat tight on the terminals. <please don't make fun of my lopsided holes on a few. This took way too many hours to do and kinda sorta figure out>
View attachment 47461
2.) I hooked up the cells on a 2p16s config, slapped on the trusted Renogy 48/3500 (Seriously the product is not *that* bad Will
, and it has a marked grounded spot for the AC out that you missed in the video)
3.) Charged them like this for @25A until 1 cell (set) hit ~3.3
3a.) Note: the 16s bms can't reach them in straight line, and I have yet to cut the bus bars to make this a proper rectangle, so I did the oblong ending so I didn't have to mill anything else until my press gets here.
3b.) I should have let them go all the way to 3.6 on 1 cell. I jumped the gun because "cell" was creeping out higher than the rest. This could have been due to non-copper busbar connecting them (see below), or maybe not. Idk i am not an EE.
4.) The 16th cell group (on the Red positive) hit high voltage
View attachment 47462
5.) I took the cells and split them (poorly) into 2 sets of 16 again, with "cells" 1-8 now being 1-16 on string 1, and 9-16 being 1-16 on string 2.
6.) I series charged string 1 more, while starting to parallel charge string 2 @ 3.65V
6a.) The 10A/30V charger would only sink like 4A of current (prob because voltage was close-ish). After literally 48 hours the cell voltage just hadn't changed a meaningful number.
7.) While String 2 was in the forever parallel balancing act on the slow charger, I manually tried to balance some of String 1 (the series string) by adding resistors to the high cells, led light sinks etc.
7a.) This actually worked quite well. I got this string to hit HV disconnect around with them all being pretty close.
7b.) Parallel'ing was still going slower than life
8.) Swapped the string series (1) into parallel, and parallel (2) into series
9.) Parallel charged string 1. This was making slow but noticable progress from ~3.4 to 3.63 over a few days
10.) Did Step 7 with string 2 to manaually balance while in series. I also used boost charged some cells as needed to help speed this along
11.) String 2 *back* to parallel. Step 9 with string 2 now. This only took ~12-18 hours to get to ~3.63
12.) Re-trickled string 1 to be very close to string 2 voltage
13.) parallel connected String 1 and 2 now into 1 giant 32P battery lol. I put on the 3.65 and let it charge for another day or so until the current sunk to ~0.2A, and the voltage across the entire battery was 3.621
14.) Let them rest for a few hours, then re-set them up as in step 4 (but used copper bars on the main negative, main positive). Hooked them up to the inverter and let the capacity test begin. The cells were really locked in voltage (imo) all the way down. Never got more than ~.007 apart
Fun facts. I did make sparks only once, and thankfully that was when messing with the parallel pack @3.5V or so. And boy did it rain sparks for a few seconds (I tried to connect 2 sets of 4 backwards (P --> N and N-->P)
Overall this took probably 5-6 days to fully do. Could have been ~2-3 if I knew what I was doing, and had enough trust in what I was doing to let the charging happen overnight the first few nights (I was too worried about un-intentional overvoltage, so I disconnected when I couldnt monitor)
Thanks for coming to my tedtalks