So I’ve got most of the cells charged balenced overnight playing with a lower charge 1a hopefully for less vdrop. Found using balence while charge on and off and changing the start value for it to slow down the 6-10 groups that are well balenced from reaching overcharge.
It’s a bit tedious though to constantly do it delta now less than 0.6v
Could I prehaps just unplug the postive lead of that group? I’ve had weird BMS backlash on me doing that before. Wasent well designed in general though in fairness
From the BMS?
Delta of 0.6V is definitely not balanced. You should aim for at least 0.1V, preferably 0.03V or lower.
And do not disconnect the BMS wire permanently, they should all be connected.
What are the voltages of the cells? (A screenshot of the BMS screen would help)
I’ll trry and screenshot each pack with the LTT and with the FlexiBMS
This is exactly what I do lol
This was before I wrapped them lol
This is probably my worst pack - had it for over a year it’s been my abuse pack for extreme testing (it’s seen water damage and eveyrhing under the moon) recently lost group 10 fully and 9 is only recently acting up (discovered on the oringinal enertion bms that was delivered with this pack that the pgroup 9 pin broke off (the BMS side)
Plugged into LTT to check and see the above.
(Two other packs in question are new. Never used or used 1-2 cycles. But I shorted one of the groups trying to remove the pins to swap them to a 13s connector for the flexibms) I now know how to remove pins correctly lol… will add screenshots.
It does seem to me that a BMS that has direct access to + - and also the positives of the balence wires is much more capable than one that only has postive from the balence s and is otherwise charge only…
In the case of the LTT BMS seeing as it technically is a discharge bms but only connects to the negatives (other than balences) seems to work most of the time but has shortcomings.
Aka if you have 1-10 and group 5 is unreable/dead/ or something is wrong.
Groups 1-4 will balence and 6-10 will balence. But getting 1-4 to balence with 6-10 seems near impossible as it seems to try to push all the current into the faulty group (probably to revive it)
IMO, something else is going on.
These “passive” balancing BMS’ just discharge each cell, when required for balancing, using a resistor and don’t move charge from one cell to the other. They can’t bring any cell’s voltage up, just drop it down.
Some BMS controller chips do divide their cells into groups of five though. But the BMS setting will apply to all cells so even if there are separately monitored groups of five they should all be brought to the same range of voltages. Unless that BMS is all software controlled, via a MCU, and doesn’t use one of the BMS controller chips
Weird what that BMS is doing though.
What are your balancing voltage threshold and allowable balancing voltage delta settings?
I’m not a pcb expert but this BMS seems to be on par with flexi lite. Flexi seems better both can do the “active balence feature both during charge and outside of charge.
But I can say for sure watching it do what it does. It literally is taking say 10-30mv from a high group and feeding it into a weaker group which seems to work well. On my perfect pack this BMS needs only a minute or two to make sure eveyrhing is within 10mv or less
Again though I think there’s something to do with if a BMS has that + - main lead access. And of course is software controlled (which the LTT is)
Just took out that really old pack. Still can’t get group 9 to go above 3v for very long. Did manage for once to see it hit 3.078v. ( but it required me to have 4.35 as cell overvoltage which then the cooldown is 4.15. )
Once I got over 4.20on all cells the last group finally started to rise
I’m still wondering how to use these features correctly
Is this happening during charging?
If so then the charge is not being shifted from one set of cells to another. Charge is being drained from one set while the other set is still being charged up (increasing its voltage).
No it happens outside of charging, in addition I can do it during charging as well. (As in there’s nothing at all connected besides a BMS and A Pack
Something weird is going on then because I checked the circuit board for a LLT BMS and it has none of the components needed to shift charge from one cell to the other but does have the components to only bleed charge from individual cells.
The Flexi only does passive cell balancing too, just discharging individual cells as needed to balance the pack. Unless by “active” you mean using a MOSFET to switch the individual cell balancing discharge current on/off versus using true charge transfer balancing between the cells?
Could be a different model, i’ll admit ive tried to buy more of these and the perticular 60A discharge model i have, I cant even find a photo of it. Basiclly it shares the same charge discharge port - where most have a seperae port (makes no differeance)
Also it has a UART out port, insead of more Potential Balence pinouts.
To answer your Mosfet question - It Does Balence with MOSFETS on, and when they are switched off . Though the operation does seem slightly different between the two which i think you have explained possibly why.
Any ideas on the above?
Irony aside. the LTT is balencing well…
The flexi BMS is killing cells, but with the same pack i orginally set it up with it balences just fine.
Makes me wonder if there is a benfit to wiring each set of 4 cells +/- So a 10s4p would have an insane amount of leads… so i could just route those outside of enclosure for charging/bal.
Clearly going to have to learn more in the field of battery pack design.
Side note: I do notice some BMS come with Sheilded Balence wires. is this simply to ensure the readings are acurate?
Which thing are you referring to?
Huh? What is it doing wrong? Which pack is it doing this with?
Some people do bring out leads for each cell for balancing.
I don’t know why they did it but keeping out electrical interference (which can affect voltage reading accuracy) would certainly be the most logical reason. You can still get conducted interference coming down the wires though so filtering and other ways to handle that need to be in place.
That makes sense, I generally also either keep my wires from overlapping or i twist them if the fist option doesnt work.
Though with flexi i did use a extra long silcone 22awg bal leads… folded neatly, wondering if i use sheilded cable if that will improve.
The Sheilded Cable is rated at 105c 600v… to my memory only a few types of insulation go to 105 (i thought it was FEP/and irradiated PVC. I’ll check the cable for the Unique ID and see what it is.
The flexi works great but of course if your pack is 10s but actually 9s… it dosent like that. which is fair.
But i did drain 1 cell from 3.7 to 2.0v. (could be reading issue)
Bought some LCD’s to basiclly wire up to my packs leads so i can watch the progress on a few packs at once. 300 cells 40 at a time takes to long.
Question: Would it be best to make an adapter that splits the balence leads on the male connector plug that goes into the BMS? (can imagine that might kill the lcd’s?) which would mean id need to somehow access the leads on the packs?
Sorry, I’m not sure I understand the question. If you are accessing the balancing leads then I can only recommend doing it in the easiest, most reliable way that jeopardizes the pack the least,
which would be :D? talking to a fella thats blown more cells in the name of ignorant science 
Quick learner once i get it though.
All i can see going wrong is as charging is via bal once at a certain voltage. if the connector is split… the current would be split on both wires…as far as i know?)
Sorry, I just don’t understand the scenario you are describing. An image might help.
You want to take the existing balancing harness connector and put a y-connector on it so you can run the balancing leads to an external charge/balance port in addition to the BMS?
You’ll probably want to disable a lot of the BMS’ functionality so it’s not “fighting” your external balancing.
