For lifepo4 I’d need to stop at 3.65, right?
We’ll do this exercise today. I’ll get the pack slowly down to 2.5 and charge it back up.
Well, looks like I may still have to take one of the 6S packs apart.
I was discharging slowly (0.6A) and about when all cell groups got under 2.6v one of them started dropping fast. Within 10-15 seconds it was under 2v. Stopped the discharge and it slowly climbed back up to a little over 2.
The other 6S half seems fine, all the groups stayed close to a little under 2.5v.
Is there a point in charging the pack back up to see how much goes in or having one group lose voltage so low is bad and I’d have to replace the cell/cells from that group first?
1 Like
Arrgghhh….my apologies…yes, for LiFePO4 you stop at 3.60V or 3.65V. You’ll have to check the datasheet.
So one of the 6s packs seems fine. I’ve been charging it at 1A (0.2C) for over 5 hours now.
A bit surprised that we have already put over 5Ah in… this is 6s2p and the cells are rated at 2500mAh. How are we getting more capacity?
I’ll let them get up to 3.6 and we’ll see about the other 6s pack tomorrow. I’ll probably charge it up once but I am a bit worried about that p-group I mentioned above that discharged too low.
Slightly different topic. I got in touch with the manufacturer of the spot welder I used and it looks like I may be able to get a refund on it. It’s advertised as capable of welding up to 0.2mm but it won’t even do properly 0.15 pure nickel strips.
Anyway, the point is, I’d like to get this right. What spot welder should I get for this and future pack builds? Ideally it should do up to 0.3mm strips and it can either come with its own battery or be able to work with 6S LiPos (I have a few spares). I don’t have a spare car batter so lead acid or similar options are out.
Only does up to .2 and can self-destruct if you don’t modify it before the first weld. It’s easy to fix up but not for everyone. I had to add a cap and some copper to beef up the traces. Probably use someone else’s suggestion lol.
It might work, I don’t mind tinkering with it and I do have some copper bars…
On the other hand, I might just order a kWeld kit but not sure how long it would take for shipping from DE. That and I was hoping to use some 6S Lipos I have… I read that kWeld works better with lower voltage input. We’ll see, I’ll do some more research.
Back to our packs… slow progress. Yesterday we decided that one 6S half is good and today we’re charging up the 2nd one. This 2nd one I’m a bit more worried about, it’s the one where we saw that one cell group get charged much faster initially and then here we also have another cell group that discharged much lower, under 2v while the rest were at 2.5v.
Right now it’s charging back up and we’re at about 3.35v. Once it gets to 3.6v I’ll discharge it one more time all the way and if that one group gets that low again, I’ll have to replace it.
2 Likes
ok, results are in.
The 2nd pack charged up fine, we ended up putting in almost 5600mAh and the p-group which ran high first time, was fine now. It probably needed a good balancing…
Moved on to discharging it.
The p-group I was worried about, the one that discharged too low, died completely this time. By the time the rest of the p-groups were down to 2.8v, the bad group was already below 1v.
Surgery it is! “Luckily” it’s group 6, the one from the rear end so it was easy to remove without damaging the rest.
Now, I do have a bunch of spare cells but I want to make sure I choose good ones. What’s the best/quickest way to test individual cells? Should I just make a 1s2p group and discharge it to see how it behaves?
1 Like
If you plan on doing more battery building you should get a cell charge/tester that can do a capacity test. Quick and dirty check would be to check the voltage of any cells that have been charged and let to sit for a week and see if they are self discharging.
Side note: What’s funny to me is that smaller packs with fewer cells in parallel are actually harder than larger packs. With 4 or fewer cells in parallel you have to be more precise and careful about matching cell capacity and IR characteristics. Also the connections and nickel strips need to be perfect as any defect in placement or welding doesn’t have the structural advantages of larger groups. If there are any cell mismatches the increased load placed on lower pgroup builds per cell causes them to wear even faster and get into unbalanced or over charged/discharged states. Rant over lol
I would suggest fully charging the prospective cells if they aren’t already. Checking them after a week of sitting for self discharge.
Another good test is voltage drop over a common load which is a way of testing IR without testing IR. Separate your FULL (important that they are fully charged) cells by exact voltage - test to the hundreth of a volt at least but you want as much accuracy as you can easily manage, then take some of the closest matches and put them on the same load for the same amount of time and check the voltage while under load. I usually load my cells for a full second or two but not much more than that with a clock running so I don’t have to count and record at the same time. You want this first and second voltage measurement to match as closely as possible. Use a resistor or non fluorescent or non led light bulb, a fan, a heating element. Just so long as it will only load the cell below its rating.
Take two that match best and make them kiss, they p group buddies now
3 Likes
Indeed. This thing is turning out to be quite a pain but we’ll get there.
For sure I plan on building more packs in the future so I’ll get a decent welder and tester eventually. Now I just picked two random cells from the left overs and I’m going to charge them up.
I’m starting to feel like in those time loop movies where same things just keep happening over and over. Now, while I was messing with the spare cells, I put those halogen bulbs to discharge the “good” pack again just to double check. Low and behold, now a p-group from that went too low. All the rest were at 3.0 and one group started going below 2.5 quickly. I stopped the load and it started to recover but this is not a good sign… I need to look into this now again 
Recycle batteries are way more work than new cells and require significantly more equipment and effort to make work well. I just like doing it fsr it doesn’t seem cheaper than new cells to me.
I wont use a used cell I haven’t tested throughly, been burned (figuratively) too many times.
For just cell charging and basic capacity testing i use the opus 3100 and 3400 chargers, takes time to do the discharge test but it is repeatable within about 5% in my experience but don’t trust the IR test for any of the smart chargers. You can set the voltage on this charger so it will do lifepo batteries, there is a switch on the pcb you can flip to make it work correctly BUT it will kill cells if it is set wrongly. (Probably why it isn’t easy to access or a published feature).
Good ir testing results can only be obtained with dedicated higher voltage test equipment but the test I described in the above reply will work with just a few dollars in parts - just remember to test multiple times: cell 1 then cell 2, then cell 1 again and cell 2 again. That way a bad or accidental high resistance connection doesn’t throw off your measurement.
1 Like
One more question on these, trying to understand. Before taking apart the “good” pack because of that p-group which started discharging too low.
Is it possible this happened because the pack was never completely fully charge and balanced? I did stop charging at 3.6 so is it possible that one group was not actually fully charged and when I did the discharge exercise, it just ran out of juice faster than the others?
Trying again now to fully charge up the pack and I’m going to leave it on the charger until it’s happy. We’re doing it at 1C now so it shouldn’t take long… the balancing part, I’m not sure.
I’ll be honest, I’m getting a bit tired of messing with this pack. I did charge/discharge the random replacement p-group cells and they seem to behave so I welded it back into the pack and wrapped it back up.
I connected back the LLT BMS and I’m going to leave it alone to do its thing. I know there’s a bigger delta but that’s because I was separately charging/discharging the two 6S packs.
If it looks ok by tomorrow, I’m going to move on to connect the vESC, remote and motors.
2 Likes
The balance current is like 50ma so it might take a day or two for initial balancing. Let it do its thing and then look for problems. It can’t be rushed easily so relax a bit, i had assumed you were working with a balanced pack. You can’t tell if discharge is un balanced or wonky until it is first balanced and this can only happen at the top of the charge cycle. An unbalanced pack in the middle of its discharge is going to look scary bad and part of why balancing is so important. My recent battery build took a week to balance iirc
1 Like
Is it better to leave the charger connected or not until it’s all balanced.
The way the BMS is set up now, it’ll stop charging when any cell reaches 3.65 and resume charging when it’s below 3.55
It’s looking much better now. I’m still going to leave it alone for some time, I need to prep the board and the vESC mount…
That is looking good, i set my first balance on a pack to .01 delta and then after a few cycles set it to .025v delta. It takes longer to get the smaller delta but let’s me see exactly what I’ve got going on and i take my good multimeter and check to confirm the voltages. You can calibrate in the bms from there but if it’s close I don’t worry about it and my readings have always been close enough.
I’ve been switching it to .025 delta after a few cycles because I don’t see any need to be better matched with how im riding and charging (sitting in the middle 85% of my capacity usually). Seems to be working fine so far. I charge to 4.17 then it balances and settles to about 4.15ish, I live at the top of a hill so i keep head room so i can break hard on the downhill heading to the trails around me and usually get home with 15-20% minimum… also because hills
1 Like
While taking a break from watching the pack balance, I’m tinkering with the board.
We have to extend a bit the channels for the leads connecting the pack to the vESC and make new holes for new threaded inserts for the battery box. The vESC will be mounted to the original boosted ESC backing plate so that’s going to be quite neat.
Oh yeah, and I fixed the logo 
2 Likes
Slow progress. Sometimes I think I enjoy the whole process more than reaching the end.
Loopkey day!
I wanted to keep things as modular as possible so this is how we’re going to connect it. The reason is the XT90 connector part is going to be glued in the box and I’d like the pack removable.
Found a nice piece of solid copper bar:
For now I’m going with a bunch of layers of conformal coating. Will probably print a little case for it or something:
6 Likes
As much as I wasn’t looking forward to this part, time to pack the …pack.
Again, in hindsight, I would’ve designed the enclosure with significant differences but at least for now this will work. It fits but space is quite tight. Better lead management would’ve also helped but we’re not going to worry about that for now.
To be honest, I’m going to be surprised if all this works but for now the BMS is not complaining.
I wanted to connect the vESC for initial setup but I can’t find the right bullet connectors to match the ones already on the vESC leads. Might have to take those off or just solder the motor wires. We’ll see.
5 Likes