uneven battery discharge rate LiFePo4 packs in series

[timelovelo]

Newbie, on my new rig, so bear with me.

Setup: 2 x 24v 15a/h LiFePO4 batteries, wired in series. Batteries are loaded onto bike 'pannier' style for even load balancing ie one 24v batt per side.

Problem: the battery on the LHS, where the negative feeds out into the controller consistently discharges at a greater rate than the battery on the other side- where the positive feeds into the controller.

Symptoms: - Both batteries charged independently of each other to full capacity. Tried rotating the batts around (batteries are less than 10 cycles old); still get the problem of one overdrained battery, one less-drained battery ie uneven discharge rate.

- The overdrained battery is always on the LHS- where the negative feeds out into the controller.

- If I use the batteries individually, I get up to 12ah life out of each battery, but only at 24v. When wired up in series ie 48v, I barely scratch 10/ah of drain before I get the problem of one battery being overdrained (it beeps incessantly), while the other one still with enough charge for me to limp home on 24v (after disconnecting the in-series wire).

- problem of uneven discharge has happened three times now, so it must be a feature of the circuit.

Solution: no idea! is there some kind of electronic load balancing I can insert into the circuit to evenly discharge the batteries?

Here's hoping there's a simple fix out there.

Tim

 

[JUICED]

Hello!

Do you have a BMS for the charging of this?


Juiced

 

[amberwolf]

- The overdrained battery is always on the LHS- where the negative feeds out into the controller.

If they are literally only in series, then this should make no difference, unless there is a fault in the wiring that is causing current leakage from the positive end whatever battery is on the LHS.

However, this would only be the case if when you use single packs you only use the LHS - and RHS + connectors to hook them up, and do not use the series connection wire at all.

If so, you could examine the cable that connects between LHS + and RHS -, to ensure it does not in any way have damage or insulation problems, and the same for the LHS - wire.

One slight possibility is that if the BMS in the packs is somehow being triggered into shunting by the fact there is a positive voltage above it in the RHS pack, and if the design is such that this causes extra pack discharge within the BMS.

Does the BMS get warm or hot on the LHS pack? If so, the above is likely to be the case. I'm not sure how to fix it though.

 

[oldhaq]

My first thought was the connecting wire between the two packs, it may be too small a gauge and/or too long to conduct the current well, and drain the neg side pack faster than the pos side. Try doubling up on the connection, shorten or replace with heavier gauge wire.
I had a similar problem on my first pack (2x24v), connected everything up, had power but applied throttle and nothing. Measured voltage and found LVC on controller was cutting in, shortened the thinnish connecting wire by about 2 1/2ft (it came from the infamous jimmywu!) and it was ok. I subsequently rebuilt the pack with heavier wire all-round.

 

[amberwolf]

My first thought was the connecting wire between the two packs, it may be too small a gauge and/or too long to conduct the current well, and drain the neg side pack faster than the pos side. Try doubling up on the connection, shorten or replace with heavier gauge wire.

All a too-small gauge wire would cause is insufficient current to flow to the controller. It wouldn't cause extra drain--it would actually *decrease* drainage from both packs.

That said, when I had a pair of U1-sized 31Ah SLAs in my CrazyBike2's cargo pods, I used a car battery cable to connect between the LHS and RHS units.

 

[oldhaq]

I still wonder if a thin wire linking the two pack halves affects the rate of drain on each pack?
I was thinking If electricity travels from neg to pos, and there is resistance halfway between the two pack halves, would the neg side pack's current flow into the circuit easier than the pos which has to go through the thin linking wire first, would the neg side pack get more drawn from it than the other pack, I'm not referring to "extra" total drain for the pack overall though, except for heat loss through resistance.
Another reason I'm thinking along these lines is that when I have had a cell failure it has most often been the negative end cell, BUT,that said have also had the + end cell go.

Another thing to consider is if the peukert effect is in effect here, somewhere. What are the type of cells are in your packs? What current is your controller rated at?
I suppose that still wouldn't explain why when either pack is on neg side it drains more than the other though.

 

[amberwolf]

I still wonder if a thin wire linking the two pack halves affects the rate of drain on each pack?
I was thinking If electricity travels from neg to pos, and there is resistance halfway between the two pack halves, would the neg side pack's current flow into the circuit easier than the pos which has to go through the thin linking wire first,
would the neg side pack get more drawn from it than the other pack,

No. It's all in series. Meaning, everything on the line is affected equally by the resistance of a wire in the path.

Now, if you were tapping off of the pack(s) for something, above or below that connecting wire, then sure, it would affect stuff on one side but not the other. But only then.


here: you can test this: Take two cells (or batteries without BMS) of any kind, identical, and identically charged. Wire them in series using a *single strand* of wire out of a regular insulated wire. Hook up a load across both batteries that will draw significant current from the pair of them, enough that you will feel that little wire get hot. Do this just long enough that they'll both be drained noticeably. Measure them both. Now recharge them to the identical state as before, and swap them so the more positive side one is now on the more negative side. Redo the test. You'll see that the cell that has a lower state of charge when done will end up being the same actual cell each time, regardless of position on the pack. It might require a large number of these tests to give a statistically valid sample to prove this, depending on how closely the two cells are matched in capacity.

Another thing to consider is if the peukert effect is in effect here, somewhere. What are the type of cells are in your packs?

Peukert affects everything at once if it's in series. If all the cells are the same size it'll be the same effect on the whole pack.

 

[oldhaq]

Ok, I won't do that test but I get it now, thanks.
Actually I've just imagined a simple circuit with two cells and a resistor between them, they're gonna drain evenly. Dunno why I had a mental block on this.

I'd have to confer then, the BMS('s) or associated wiring could be the suspect, or an insulation issue.

Does each pack have isolating diodes? may help when each pack has a dedicated bms if you run them right down.
Are you running 2x8 or 1x16 channel bms?

Also check the voltages for each cell, the bms may not be balancing all cells.

If it were my packs, to test if the bms is causing the problem, I would full charge then measure individual cell and total voltages, disconnect the bms and run the bike over a distance known to be well within range of the battery, then check each pack again and how much charge each then takes. They should hopefully take the same amount and can then lay blame squarely on the bms. That's me though, you may not want to run the risk of pack damage at all.

If you disconnect the in-series wire wouldn't that break the circuit, unless you have quick dis/connects everywhere to complete the circuit again. How do you run your bike on 24v? if it is a 48v or even 36v controller, I would've thought the controllers' LVC would be in effect.

 

[timelovelo]

here: you can test this: Take two cells (or batteries without BMS) of any kind, identical, and identically charged. Wire them in series using a *single strand* of wire out of a regular insulated wire. Hook up a load across both batteries that will draw significant current from the pair of them, enough that you will feel that little wire get hot. Do this just long enough that they'll both be drained noticeably. Measure them both. Now recharge them to the identical state as before, and swap them so the more positive side one is now on the more negative side. Redo the test. You'll see that the cell that has a lower state of charge when done will end up being the same actual cell each time, regardless of position on the pack. It might require a large number of these tests to give a statistically valid sample to prove this, depending on how closely the two cells are matched in capacity.


Snip

Yes, you I think you are right, it is in fact the same battery. I'm unable (and unwilling) to check each individual cell of the faulty pack, and I don't have the diagnostic equipment to check the BMS (an in-built VMS). I will need to get the supplier to confirm this.

Thanks for everyone's input!

 

[amberwolf]

Yes, you I think you are right, it is in fact the same battery. I'm unable (and unwilling) to check each individual cell of the faulty pack, and I don't have the diagnostic equipment to check the BMS (an in-built VMS). I will need to get the supplier to confirm this.

That test I gave was not for your pack or problem, but for Oldhaq to be able to see what I was meaning. Only the info I gave in my first reply was intended to help solve your issue.