new lithium bank install conundrum

[Nightsky]

Hi All. I just installed new, drop-in LiFePO4 batteries. By way of background info, I had 3 AGM 8D batteries in my original bank. I have left all the previous wiring to/from battery bank in place, took out the 3 8D's and dropped in 3 12v Lithiums of similar physical dimensions to the 8D's, so everything should just connect without to much fiddling. I first charged each new battery to the manufacturers spec (14.1-14.2v), before paralleling (positive to positive to positive to the red, house lead, and negative to negative to negative to the shunt) each battery in to a bank. I made new cables for the new bank, ensuring each cable was the exact same length to allow for equal charge and discharge of each battery within the bank. Here is the conundrum, my bank did NOT discharge equally from each battery. In fact, it discharged solely from the battery that had the negative going to the shunt, and did not draw any amps from the other 2 batteries. At first I thought the bluetooth info from each battery's internal BMS was giving faulty info, but I broke the bank apart and checked each battery separately with a meter for voltage and checked against what the bluetooth was saying for each battery. I confirmed that only the one battery was discharging from the bank. Totally unbalanced! Obviously I have to find a solution to this before we can get underway again. Anyone in the braintrust out there have any clues what I may have done or assumed incorrectly? Since I got rid of my old 8D AGM's, I can't go back to that, even temporarily, I have to sort this out, and soon. Right now I am "living" off one battery at a time, while charging up the battery that we drew down the night before, not a tenable situation.
I did go to this website,
SmartGauge Electronics - Interconnecting multiple batteries to form one larger bank

and read up on their methods for paralleling a bank and will try Method #3 and "hope" I get the desired result, but first I have to get caught up and make sure all 3 batteries are equally charged before ganging them to a positive and negative bus. My first method was #2.
thanks in advance

[RoadRacer]

^^following....

[Gtstricky]

Method two should work well. Which LiFe batteries did you put in and what BMS are you using?

[Steve_C]

More information would be helpful. It should be impossible for a cell to draw down without affecting the others. It sounds like maybe the internal BMS of some of the batteries are preventing them from being connected when installed.
Please provide pics of the connections, a diagram of the wiring, and brand of the batteries !
Also have you contacted the seller and/or manufacturer of the batteries? It pretty much HAS to be the BMS's. Drop-ins sound like a great idea, but without a thorough understanding of what BMS its using and how they are programmed its a gamble.

[john61ct]

Many makers of "drop-in" batteries state that there are limitations on how many can be grouped in parallel, or strung in series.

Their internal BMS may not even be designed to support such usage at all.

Better to either get a packaged LFP system installed by a professional, or DIY using bare cells and protective electronics that is specifically designed & selected for that particular use case.

The whole myth of "dropping in" is usually based on the assumption that all you need is one unit.

[Paul L]

If you swap battery positions and retest you should be able to determine if the issue is the wiring or the specific batt/bms.

[Nightsky]

Originally Posted by Gtstricky

Method two should work well. Which LiFe batteries did you put in and what BMS are you using?

That's what I thought too. Will try method 3 and see what happens.

[AndyEss]

Originally Posted by Paul L

If you swap battery positions and retest you should be able to determine if the issue is the wiring or the specific batt/bms.

This procedure will clarify what the problem is

[Nightsky]

A previous reply didn't go through so will try again. These are Kilovault 3600 HLX batteries. Similar physical dimensions to 8D batteries but half the weight. Tried tech support and still waiting for call back after 3 days.

[john61ct]

Says 4 parallel strings are OK

Have you tested them as individuals?

Are they all higher than the 11.5 BMS cutoff?

Are they all within 0.1V of each other?

Who did you buy them from?

Or are you trying to contact the manufacturer?

Maybe try AltE, usually pretty responsive.

[alansmith]

Following also. Seems I spend an hour per day reading up on lithium or watching videos on YouTube. Please keep us posted about your outcome. I would be nervous as heck without customer service not returning a call.

Did you prebalance the three together? It sounds like a BMS software issue... but what do I know. Feel your pain though.

[Nightsky]

Here is an update. Did finally get a call from tech support at AltE, at 5:15, so they are working late to get caught up. Short version, he didn't know why initial attempt at paralleling using method 2 (see above) didn't work and when I mentioned that I was almost finished reconfiguring using a common bus for all the neg terminals and a common bus for all the pos terminals, he was happy to hear that as that is what they prefer. He was able to show me how to access data to each of the 4 individual cells in each battery and what paramaters they should be at. After getting it all done and hooking back up to house loads, each battery is now drawing down (somewhat) equally, with only 0.4A difference between them. So I am calling that a success. Next step is to get the new alternator and Balmar external regulator to charge the bank, but that will have to wait for morning. Next week, 2 new Victron Multiplus's, and solar controller get installed. Final step will be adding more solar panels.

[thirds]

I also endorse method #3, though it won’t give you perfect results unless the wires are perfectly identical, #1 and #2 are guaranteed not to give you perfect results.

Some variation should be ok, and the biter so BMS is the key feature if these batteries.

Glad you have good visibility inside them to monitor their internal state.

Drop in is more expensive but as someone who has built a system from scratch I think it’s very often worth it (vs managing cells yourself.)