diy solar

diy solar

Li Ion 14S (barely used) @ 7V - is it recoverable?

teotwawki.je

New Member
Joined
Feb 9, 2021
Messages
27
We have a 7P14S (6,720WH, prismatic in a restricting box) Lithium Ion battery pack which due to COVID travel restrictions, has run down to 7V total. It does have a BMS but apparently that failed to do low voltage disconnect. There was nothing else connected other than via the BMS. These were new cells shortly before COVID. Is there any chance of recovery, even if lower capacity? If so, how? It's big and heavy so in-situ would be preferred.
TIA.
 
Are you sure the BMS is not already in low voltage disconnect mode? Have you checked the individual cell voltage? Are these supposed to be 3.7v Li-Ion cells or are they 3.3v LiFePO4 cells or other)?

Once you confirm the cells are dead, you can follow a tutorial like this one to jump them enough to get them to take a charge again...

(it's not letting me post the video link, can search You Tube for the title: ' DIY: How to revive a dead 18650 (or any) Li-ion battery cell ')

They use a good cell to briefly jump a dead cell, but you could also use a constant voltage power supply to jump the cell enough to get the voltage up high enough for using on a cell charger (or just charge the cell using the power supply)...

Then of course you would have to top balance all of the cells again prior to putting the pack back together in running config.
 
Last edited:
Definitely Lithium Ion Polymer, actually: Li[NiCoMn]O2 based cathode & graphite based anode.
7V directly across the 14S battery pack, and also 7V across the BMS output. I assume a BMS in LDV would show 0V across the output.
So an average of 0.5V per cell. Haven't metered each 7P cell set individually yet. Found the video OK, will watch it now. Thanks.
 
Definitely Lithium Ion Polymer, actually: Li[NiCoMn]O2 based cathode & graphite based anode.
7V directly across the 14S battery pack, and also 7V across the BMS output. I assume a BMS in LDV would show 0V across the output.
So an average of 0.5V per cell. Haven't metered each 7P cell set individually yet. Found the video OK, will watch it now. Thanks.

The reason I mention to confirm at the cell voltage, is because many BMS don't actually have a mechanical relay to cutoff the B- but use solid state transistors and such, so it does not seem that the idea isn't impossible, to imagine the 7v could just be 'leaky' volts or something silly like that...
 
I've now done the maths: 140AH pack over 9 months (6570 hours) is only 21mA! So the BMS would only need to drain/leak 21mA (1Watt) to have flattened the fully charged battery pack in that time. Do cheap Chinese BMS drain/leak that much? Seems believable to me.
From the many videos, we now have a plan of action. Fingers crossed.
 
Last edited:
I gotcha, but I didn't get the 6570 days part hehe...

BMS do require an idle power spec to run them, especially if they have bluetooth.

Anyways, good luck on it!
 
Oops you're right - 6570 hours, not days. Corrected. It was only the word I got wrong, sums still right as far as I can figure.
 
The Battle Born BMS when over loaded to charge caps in inverters shows low voltage till you disconnect it so it resets. And it is then right back at full voltage.
 
Our circumstance is a bit different from all the YouTube demos of jumping one 18650 from another & it is still on-going but looking promising. So for others & future reference I'll continue to note progress & results. What's different is as follows:
1.) The physical size of our pack: 2 boxes of (7P7S x 20AH), connected by metal links with 392 screws. So we've not dismantled the battery pack into individual cells.
2.) Our jump battery is 2P x 20AH so only 2/7ths of the capacity of the 7P cell sets we are jumping. So it's taking a lot longer than the 10 - 30 seconds mentioned in some of the videos. So far about 5 minutes into each 7P cell set.
3.) The charger we are trying to get to recognise the pack is a "48V" All-in-one expecting 14S Lithium Ion. Last night we reached nearly 38V & it still wasn't recognised yet.
4.) Also these are barely used cells, so we're not battling with the effects of aging.
 
Success!! The All-in-one recognised the battery pack once it reached 43.5V and is now charging happily.
Having gently reached a pack voltage of 38V and with the 2P jump battery not much higher V than the target cells, we changed method.
Now using 2S AGM batteries (at 12.7V) we started charging half of the pack at a time, softening the brutality by going via the ceramic power resistor we use for pre-charging the inverter capacitors. After a round or two of that, then the same without the resistor - a little sparky but OK.

So, what works with a single 18650 also works with a 6.7kWH pack in-situ without having to dismantle it!
 
Do you have a strategy to check if the cells are all back in balance once it's fully charged? And to balance them if necessary without disassembling the pack? Does the BMS have balancing (on charge) built-in or anything?
 
Do you have a strategy to check if the cells are all back in balance once it's fully charged? And to balance them if necessary without disassembling the pack? Does the BMS have balancing (on charge) built-in or anything?
The BMS is a cheap one but it does have passive balancing - we just don't know the algorithm or parameters. While manually charging each 7P cell set, we were logging all cell Vs regularly and adjusting the charge pulse for each (by estimate) to not get any wild differences. After letting it settling over-night we'll measure every cell V again in the morning and make a plan if we need to. We could help the BMS bring any high cells down by holding the ceramic power resistor across them where necessary.
I am greatly impressed with how robust these cells have been & glad of all the pack design effort we did up-front.
 

Attachments

  • batpack.jpg
    batpack.jpg
    589.6 KB · Views: 4
Back
Top