LiFePO4 Batteries: Discussion Thread for Those Using Them as House Banks

[JmanC]

Originally Posted by Cpt Pat

That's what I've been doing for the past three years. Every 30 or so charge/discharge cycles, I'll charge to 100% SOC (charge current taper to 0.03C) and measure each cell voltage at the end of charge with my Fluke 117 (that has a NIST-traceable calibration). If I see more than a 0.05 volt mismatch, I connect a 2.7 ohm 10 watt resistor across the high voltage cell(s) until I achieve balance. That resistor cost me $0.27.

I would check more like every 100-300cycles if they are well matched, would you agree?
Does that simple balance method get the same end result as an active BMS unit? Sounds like you will always keep leveling your cells AH capacity down just shy of 100% vs. 100% with active bms.

My non BMS idea was to have some pos & neg wires running to all the cells in parrelle that connect to a plug, with an isolator switch on the end of the strings, so I can simply hook up a benchtop PS and quickly top balance every few years. If that makes sense.

[witzgall]

I was once one of the "Non-BMS" crowd. Here is why I now think this is not a good/safe approach.

This method assumes that there will never be a failure or degradation of a cell that does not happen gradually. If a cell starts to go bad quickly enough, these "Check it now and again" methods will not catch the issue. A highly observant user MAY catch it, or may not. Having a BMS in place to keep balance and let you know if they are not is crucial, IMHO. Having a 24/7 sentry to protect when a cell goes bad is paramount. Think about it. The fact that the battery cells will fail is not an "If", but a "When." How do you know when?

Chris

Originally Posted by JmanC

I would check more like every 100-300cycles if they are well matched, would you agree?
Does that simple balance method get the same end result as an active BMS unit? Sounds like you will always keep leveling your cells AH capacity down just shy of 100% vs. 100% with active bms.

My non BMS idea was to have some pos & neg wires running to all the cells in parrelle that connect to a plug, with an isolator switch on the end of the strings, so I can simply hook up a benchtop PS and quickly top balance every few years. If that makes sense.

[CatNewBee]

Buy an addequate BMS., or go gambling.

[newhaul]

Originally Posted by CatNewBee

Buy an addequate BMS., or go gambling.

to much money on the line not to get at least a cheap one .

[mbartosch]

I agree. Get a BMS with cell leveling and ample cutoffs. (Balancing in the BMS is optional IMO, though)
We actually had a bad cell in our 4P4S pack after last season’s winter. The BMS detected the problem and prevented a very expensive chain reaction of other cells deteriorating as a consequence.
Will probably report about this incident sometime later.

[newhaul]

Just thought I would post this here
My cheap bms with passive top balancing
150 for my 200 amp in and out bms

https://www.ebay.com/itm/4S-12V-3-2V...edirect=mobile

[mikedefieslife]

Originally Posted by mbartosch

I agree. Get a BMS with cell leveling and ample cutoffs. (Balancing in the BMS is optional IMO, though)
We actually had a bad cell in our 4P4S pack after last season’s winter. The BMS detected the problem and prevented a very expensive chain reaction of other cells deteriorating as a consequence.
Will probably report about this incident sometime later.

Did you just replace it re-balance and carry on as normal?

I would have thought the LFP crowed would be even more wary than the FLA crowd as adding a new battery/cell to an older bank.

[newhaul]

Originally Posted by mikedefieslife

Did you just replace it re-balance and carry on as normal?

I would have thought the LFP crowed would be even more wary than the FLA crowd as adding a new battery/cell to an older bank.

define older
For Fla. Older is 3 to 5 years .
For Lfp that changes to 10 to 20 years old

[mikedefieslife]

Originally Posted by newhaul

define older
For Fla. Older is 3 to 5 years .
For Lfp that changes to 10 to 20 years old

I suppose. In which case there is no need to bother even attempting to get cells from the same batch with sequential serial numbers which had been the advice in the past.

[newhaul]

Originally Posted by mikedefieslife

I suppose. In which case there is no need to bother even attempting to get cells from the same batch with sequential serial numbers which had been the advice in the past.

not as critical as it used to be thanks to modern electronics .
But still good idea to get the bank as close as possible to begin with .
Much less chance of an issue to begin with.

[nebster]

Originally Posted by witzgall

I was once one of the "Non-BMS" crowd. Here is why I now think this is not a good/safe approach.

This method assumes that there will never be a failure or degradation of a cell that does not happen gradually. If a cell starts to go bad quickly enough, these "Check it now and again" methods will not catch the issue.

Here's the real question you/we have to answer when applying your logical reasoning: if a cell goes bad "quickly," what actually will happen, in the scenario where there is no BMS present?

The answer has to be something sufficiently "bad" to justify the cost and risks of adding more management.

What might it be?

[CatNewBee]

Negligence of the user.

[nebster]

Originally Posted by CatNewBee

Negligence of the user.

Lolwut?

[CatNewBee]

How many LiPo have you wrecked in cellphones, tablets, computers, camaras, drones, power tools, blue tooth headphones and gadgets?

No protection, no alarms, they either die from permanent overcharging or silently in storage by deep discharge.

Most boat batteries, especially LFP die in storage by leaving them almost empty for longevity by self discharge and small currents, or by overcharging, leaving them permanently on the shore power.

You wont notice an empty LFP under way either, because the voltage will stay most of the time above 12V, when your alarm goes on, the cells are already depleted and critical.

LFP works so great, you just forget about it in a few weeks, and next season you kill them. The biggest enemy of a LFP is a negligent ignorant user, and a BMS can help protect the system from human errors.

[nebster]

Originally Posted by CatNewBee

How many LiPo have you wrecked in cellphones, tablets, computers, camaras, drones, power tools, blue tooth headphones and gadgets?

They wear out pretty quickly. That's not "wrecking," that's just 500 cycles at high rates.

No protection, no alarms, they either die from permanent overcharging or silently in storage by deep discharge.

Most small batteries have very active management and protection. At a minimum, they'll be throttling charging to prevent fires. They have BMSes. And yet, paradoxically with respect to your argument, they supposedly are being "wrecked" all the time. ???

Most boat batteries, especially LFP die in storage by leaving them almost empty for longevity by self discharge and small currents, or by overcharging, leaving them permanently on the shore power.

I don't have the data to show that one way or another, but I certainly won't disagree with you. However, I have no idea what this has to do with my question or Chris's concern about rapid cell failure.

LFP works so great, you just forget about it in a few weeks, and next season you kill them. The biggest enemy of a LFP is a negligent ignorant user, and a BMS can help protect the system from human errors.

Now I know what "negligent user" means, but it seems that either you're not understanding the question I posed, or you are replying to someone else.