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

[senormechanico]

Bob,
I balanced my pack using an automotive headlight.
I just soldered the high and low beam leads together and used alligator clip leads to the high cell.
The light was dim, but it served the purpose of a load.
After a few sessions of 30 minutes, then ten minutes, and a couple of minutes another cell became the "high" one according to my MiniBMS.
A digital voltmeter checking all cells confirmed it.

After that, I just let the BMS do its thing.
Cheap way to get it balanced.

[deckofficer]

Originally Posted by senormechanico

Bob,
I balanced my pack using an automotive headlight.
I just soldered the high and low beam leads together and used alligator clip leads to the high cell.
The light was dim, but it served the purpose of a load.
After a few sessions of 30 minutes, then ten minutes, and a couple of minutes another cell became the "high" one according to my MiniBMS.
A digital voltmeter checking all cells confirmed it.

After that, I just let the BMS do its thing.
Cheap way to get it balanced.

Since there was a small improvement, I'll keep cycling for a bottom balance by applying a load to the high cells. If I can't get it balanced I'll buy the variable power supply and un-band and charge in parallel as per Maine Sail.

[diugo]

Bob, you shouldn't have to unband the batteries to do a parallel bottom balance. First, remove all the series busbars. Then use a resistive load to deplete all the cells to 3.1V or less, in order to reduce the potential energy to a safe level. Try to get all cells to nearly the same voltage.

Then use your big jumper cables and briefly connect each of the six combinations of cell pairs (AB, AC, AD, BC, BD, and CD) in parallel, to distribute most of the remaining energy evenly. Finally, tie all four positives together with #10 wire. Do the same with the negatives, and leave the wires in place a few hours.

[NahanniV]

Originally Posted by diugo

Bob, you shouldn't have to unband the batteries to connect the cells in parallel. First, use a resistive load to deplete all the cells to ~3V, in order to reduce the potential energy to a safe level.

Then use your jumper cables and briefly connect the six combinations of cell pairs (AB, AC, AD, BC, BD, and CD) in parallel, to distribute most of the remaining energy evenly. Finally, tie all four positives together with #10 wire. Do the same with the negatives and leave the wires on a few hours.

+1 but just to be absoloutly painfully clear please disconnect your series jumpers first

[diugo]

Originally Posted by NahanniV

+1 but just to be absoloutly painfully clear please disconnect your series jumpers first

Eek, thanks for pointing that out! I edited the orig post.

[deckofficer]

Thanks guys, I'm sure I'll get there. Battery #1 was so close to balance it charged just fine with no cells going above 3.75 before the charger rolled off to 0 amps.

[Maine Sail]

Originally Posted by diugo

Bob, you shouldn't have to unband the batteries to do a parallel bottom balance. First, remove all the series busbars. Then use a resistive load to deplete all the cells to 3.1V or less, in order to reduce the potential energy to a safe level. Try to get all cells to nearly the same voltage.

Then use your big jumper cables and briefly connect each of the six combinations of cell pairs (AB, AC, AD, BC, BD, and CD) in parallel, to distribute most of the remaining energy evenly. Finally, tie all four positives together with #10 wire. Do the same with the negatives, and leave the wires in place a few hours.

I just ordered 6 jumpers from Balquon, instead of three, so I can do a parallel balance anytime I need to. As you said they do not need to be un-banded to do parallel.

Just connect all + posts together and all - posts. You will now have a 3.8V 1600Ah pack, if for example using 400Ah cells, so charging from dead will take quite a while..

Letting it sit for a few days in parallel will allow them to self balance. As the voltage potentials between cells get closer and closer less and less current flows between cells to "balance". This is why leaving them in parallel for a couple of days, as suggested by Balquon, makes this process quite easy..

Personally I can't imagine trying to "balance" cells by applying loads etc.. Wiring them in parallel works and according to Balquon this is one way they do it. They will self balance in parallel easier than you can do trying to balance individual cells based on voltage. Even without a power supply, that can charge at sub 4V, you will have a very closely balanced pack by just leaving them wired in parallel for a good period of time...

[Maine Sail]

Originally Posted by deckofficer

Instead of spending $275 for the 30 volt, 30 amp variable power supply that your using, any problem other than longer times with using a $50 18 volt, 3 amp variable ps?

No problem at all but you might want to opt for a power supply that has "over voltage" protection. I would however limit the current to perhaps 2.5A so as not to cook the power supply. If I run mine at 30A for long periods it does get pretty hot but if I cut it back to 25A it runs considerably cooler..

When in parallel these banks, in terms of Ah, become very large....

The Mastech VoltEQ series are specifically designed for charging batteries & Li cells... They are designated by the EX at the end of the model number and have an extra over voltage dial to lock in the max voltage. In theory all power supplies should control voltage pretty well but some do not... For the price of the 3010EX, below, it is pretty tough to beat and 3X the current output of the 3A model......

They make a 3010EX which is 10A for $149.00

[diugo]

The goal with parallel cell balancing is to get all cells to the same state of charge (SoC)---NOT the same voltage, since each cell has a slightly different charge curve due to manufacturing variances.

Paralleling in the middle "plateau" region of the charge curve is problematic because there, cells of the same voltage could have significant differences in SoC. This is why cells are never "middle balanced".

Best to do the paralleling in the lower knee region below 3.1V where SoC varies by only a few percent at most.

[deckofficer]

Originally Posted by Maine Sail

No problem at all but you might want to opt for a power supply that has "over voltage" protection. I would however limit the current to perhaps 2.5A so as not to cook the power supply. If I run mine at 30A for long periods it does get pretty hot but if I cut it back to 25A it runs considerably cooler..

When in parallel these banks, in terms of Ah, become very large....

The Mastech VoltEQ series are specifically designed for charging batteries & Li cells... They are designated by the EX at the end of the model number and have an extra over voltage dial to lock in the max voltage. In theory all power supplies should control voltage pretty well but some do not... For the price of the 3010EX, below, it is pretty tough to beat and 3X the current output of the 3A model......

They make a 3010EX which is 10A for $149.00

That one at 10 amps and $149 works for me, thanks for the link.

[Lagoon4us]

Originally Posted by mcarling

A solid-state DC-DC converter will do that but you want one that is designed, built, and marketed as a battery charger with a specific setting for LiFePO4 batteries. It would output whatever voltage is appropriate depending on your State of Charge and accept DC input voltages from very roughly about 8V to very roughly about 20V.

Thanks i'll google and see what i can find.
Cheers

[mrm]

Originally Posted by Lagoon4us

Thanks i'll google and see what i can find.
Cheers

Likely you will find that your solar charge controller *already is* a DC to DC converter (esp. if it is a model accepting input voltages much higher than bank voltage). Your concern was how to protect a bank against over voltage in case of a charge controller failure. Two possibilities:

- controller fails 'open' - no charging but no further protection is required;

- controller fails 'short' - full (almost) solar panel voltage is present on output - severe damage to the bank possible. In this case about only sensible protection is a circuit breaker isolating bank from a charger, activated when bank OVC is detected.

[Lagoon4us]

Originally Posted by mrm

In this case about only sensible protection is a circuit breaker isolating bank from a charger, activated when bank OVC is detected.

Thanks heaps for the advice i will have 4 different charge currents available and was concerned of over voltage, murphy's law etc.

Cheers

[diugo]

Lagoon4us, my plan is to use my existing Morningstar TS-45 solar charge controller---and its high voltage disconnect feature---to disconnect charging whenever 13.5 volts is reached, and reconnect charging when the bank falls to 13.1. (I will disable all the normal lead acid features, ie absorb, float, equalize, and temperature compensation.)

For cell safety, I'll use a separate cell voltage monitoring circuit to disconnect the battery (via a shunt trip breaker) if any cell goes outside the safe 3.1-3.5 volt range. Plus audible alarms to alert if outside the normal 3.2-3.4.

[mrm]

Originally Posted by Lagoon4us

Thanks heaps for the advice i will have 4 different charge currents available and was concerned of over voltage, murphy's law etc.

Cheers

In that case a bit more than a simple breaker may be required.

Guessing that one of the other sources will be a wind generator. These will have dangerously high output voltage when not loaded, so in addition some dump load will be necessary. A water heater may be a good choice. In addition keeping to a make-before-break switching principle is important in this case (first connect a dump load, then isolate bank).

With alternators it is field control (to avoid damaging rectifier diodes and possibly other attached electronics when load is disconnected..).

You probably know all of this already, so just mentioning this stuff for other readers..