LiFePo4 Battery balancing/charging

GroupDIY Audio Forum

Help Support GroupDIY Audio Forum:

This site may earn a commission from merchant affiliate links, including eBay, Amazon, and others.

madriaanse

Well-known member
Joined
Apr 6, 2006
Messages
530
Location
Long Beach, CA
I need to charge and maintain a 180Ah pack of LiFePo4 batteries (40 cells). I don't want to let the individual cells charge past 3.6V (or below 2.9v), and I need to keep the cells balanced within the pack. Is there any device out there that does this (without breaking the bank of course)? Any schematics or ideas? I haven't come to any good conclusions on what would be the best way to do this.

Any help greatly appreciated!!

M.
 
> a 180Ah pack of LiFePo4 batteries (40 cells).
> without breaking the bank of course


Seems to me you already broke the bank?

> I don't want to let the individual cells charge past 3.6V (or below 2.9v)

Put them in parallel. They WILL be the same voltage.

Of course your currents will be huge. And a shorted battery will burst violently.

> I need to keep the cells balanced within the pack.

40 individual smart-chargers?

The only ways to charge many-cell battery arrays are parallel and series. The only sane way to charge high-current cells feeding high-voltage loads is series-charging of MATCHED cells. "Chain is as strong as its weakest link." For a minimum-weight maximum-strength chain, EVERY link must be the same.

There may be some inexpensive way to add 40 monitors to indicate that a cell is unhappy. Just an R and an LED per cell will give a visual indication of gross unbalance. But then what do you do? Cut it out? Leak charge current around it? Isolate it and have 40 small boosters to increase charge to individual cells?

The cost, weight, and complexity soon lead to a contraption which is unreliable and unmaintainable, and maybe a bit heavy. Whereas a ton of dumb old lead-acid cells from the same lot can usually be charged in series without individual cell monitoring, for several years, and then can be replaced for much less than Lithium prices.

What is really happening is that lead-acid is built "robust", has the heavyweight market locked up. The only niche for Lithium is light-weight, and the tradeoff is much less robustness, therefore much fussier maintenance. In a 6-cell laptop, same-lot Li cells match well enough for simple series charging with acceptable life at modest cost. In a 40-cell electric car, matching is not so easy, and the bank is already broke, so there is a real problem.

My next car may be a wood-fired steam engine. Zero electronics.
 
I just saw in the paper that they are making a run at the steam car land speed record (127 mph set in 1906).
======
This is not going to be cheap but DPDT relays could alternately insert a cell into a series chain for charging (with current source), or remove it and shunt the current around it to the next cell in line.

To detect the voltage of individuals cells would require multiplexing the 40 inputs down to a more reasonable number (inexpensive micros can easily have 8+ A/D inputs. Reading the voltage of two adjacent cells and subtracting would give the individual cell voltage. Since these voltages can be sampled several times per second you don't need 40 individual A/Ds. Mux'ing each A/D input x8 or maybe 1 A/D input by 8x8. Note: the micro A/Ds are typically mux'd internally so six of one, half dozen another.  The voltages would also need to be divided down to fit the input range of the micro, but even cheap ones have 10-12 bit resolution so it should be good enough for this task.

I haven't researched charging different kinds of cells but IIRC they like to be charged with pulsed DC or sometimes even a modest AC component to prevent cell deterioration. A decent amperage current source capable that much voltage compliance will require some heat sinking unless you go switching for that. The charging source would be driving a lower and lower voltage as cells reach full charge and drop out of the path.

Sounds like an interesting project. Have fun..

JR

Note: you might be able to bypass cells a little cheaper using power mosfets, but relays will be less likely to fail in mode that shorts across a cell. The micro could also detect if a relay isn't working and shut itself down.
 
Back
Top