Peculiar Lithium Ion Charging Anomalies

[xyster]

So, with my battery pack's voltage at ~3.9V per 15-cell subpack (20s15p pack), I hooked up my 20 chargers to my bike's 20 jacks for an overnight partial charge as usual.

Before heading to bed, I checked to make sure all chargers were charging (each light red). But one light amongst the 20 was green -- meaning charging had ceased to one of the subpacks! It was too soon for charging to be completed, so most likely, I thought, one of the twenty charging plugs just needed a little wiggle, like sometimes I have to do.

So I wiggled the plug in the jack, but the charger's light did not switch back to red. So I checked the voltage of this misbehavin' subpack.

4.16!

That's done as far as these chargers are concerned. But this couldn't be right because all the other subpacks, starting at the same time and about the same charge-state, were only at 4.0 volts.

On a hunch, I kept the DVM's probes in the charger jack and switched on the bike's lights. Voltage immediately plunged to 4.0V -- right where it was supposed to be! I turned off the lights, and the subpack's voltage stayed at 4.0.

I plugged the chargers back in. The 20 lights on the 20 chargers turned from green to red as they should.

This is the second such incident with this particular subpack. It's never happened to any of the other 19 subpacks yet.
The problem is not in the voltmeter either, as the charger wouldn't have acted as it did if it was.

So folks, what happened?

Do some batteries or connections sometimes carry a little "hot spot" of extra charge that must be dissipated?

Should I be concerned about a bad cell or other subpack problem?

 

[NickF23]

Do some batteries or connections sometimes carry a little "hot spot" of extra charge that must be dissipated?

Its called surface charge, with nimh it happens at the end of the charge cycle and is thus not a problem and disapears within in minute or two of discharge.

Should I be concerned about a bad cell or other subpack problem?

Yes

It sounds like your subpack is damaged and has a high internal resistance, probably some cells are worse than others. A good start would be to test the subpack's voltage at the end of a discharge cycle. Better than this would be to test the subpacks internal resistance. halogen light bulbs are an easy way to this.

You might also want to disassemble the pack and test the individual cells.

I'd be interested to see what results you'd get.[/quote]

 

[fechter]

You might be able to test the sub-pack indvidually with some kind of test load. If it is damaged, I would expect the capacity to be low.

 

[safe]

You might also want to disassemble the pack and test the individual cells.

This is why I'm "obsessed" with the idea of packaging battery cells into "tubes". If something weird starts to happen I could always disassemble the tubes and look through the cells one by one and figure out if there is a failing cell.

The idea of having an expensive pack soldered together that could begin to show signs of trouble like this is very scary... the potential of a pack "melting down" is not good.

Those newer Lithium cells that are impossible to screw up with are sounding even better... this "caveman" doesn't want to be at the controls of a nuclear power plant style battery...

 

[xyster]

You might be able to test the sub-pack indvidually with some kind of test load. If it is damaged, I would expect the capacity to be low.

I recently load tested each subpack by connecting an extra XLR plug to my DVM's probes, mounting the DVM on my handlebars, inserting the plug into a subpack jack, then giving her full throttle to 35 amps, watching the voltage drop as me, on the bike, shoot full steam towards my garage wall. Then I repeated the process 19 times -- once for each subpack.

Hitting the breaks as soon as I got a reading, and before I pancaked the wall, it turns out all subpacks but one drooped ~0.25 volts, except one which drooped 0.5 volts. The bigger-drooping subpack is a different subpack than the subpack now in question.

After a long ride, all subpacks remain within 0.02V or so. One is not appreciably and reliably lower than any other.

I'm not surprised some cells are damaged after the hot over-discharge treatment I accidentally gave them a few months ago. I was thinking this the likely cause for the 1.1 C max output of the pack. Further study at rcgroups.com revealed other people who had found that li-ion cells damaged in this manner tend NOT to have lost capacity, but DO lose power capability.

 

[xyster]

Its called surface charge, with nimh it happens at the end of the charge cycle and is thus not a problem and disapears within in minute or two of discharge.

Ah, thank you....does surface charge mainly only appear then on damaged cells with high resistance?

Another factor: I'm charging overnight when temperatures are near or below freezing. But of course were this a problem, it ought be effecting all the subpacks.

 

[xyster]

The idea of having an expensive pack soldered together that could begin to show signs of trouble like this is very scary... the potential of a pack "melting down" is not good.

It's fairly easy for me to pull one subpack, then work on it while riding the bike four volts low. The main problem is getting in the box!

 

[NickF23]

...does surface charge mainly only appear then on damaged cells with high resistance?

I don't know. With nimh it happens with healthy cells at least.

I don't know very much about lithium batteries. My understanding is that when they're used outside certain parameters they can become quite dangerous quite quickly. I once saw an intersting paper about what happened to cobalt oxide cathodes as they deteriorated, it hads lots pics. When the cell was normal it had an intricate corel like structure where the ions would settle. The damaged cathode had a much smoohter less porous surface, the battery had a higher internal resistance. If I remember right, charging a battery in this kind of condition creates a buildup of free lithium metal which, when ignited creates the more spectacular kinds of lithium fires like this one

see http://www.youtube.com/watch?v=WeWq6rWzChw&mode=related&search=

 

[xyster]

when ignited creates the more spectacular kinds of lithium fires like this one

Not a concern. I carefully wrapped the batteries with tape and cardboard, securing them in wood boxes packed with styrofoam, wrapped with nylon, and mounted to the aluminum frame of the bike.

So in the unlikely event of a lithium fire, my bike should come out relatively unscathed.

 

[bobmcree]

"Not a concern. I carefully wrapped the batteries with tape and cardboard, securing them in wood boxes packed with styrofoam, wrapped with nylon, and mounted to the aluminum frame of the bike.

So in the unlikely event of a lithium fire, my bike should come out relatively unscathed."

who told you tape, cardboard, wood, styrofoam, and nylon would not burn?

if you threw this combination in a bonfire it would flare right up, and with the temperature of a lithium fire even aluminum will melt. i fear this packaging would vaporize quickly if your batteries decided to go into meltdown. i am not saying this is likely, but just that these materials are not at all fire resistant.

 

[xyster]

who told you tape, cardboard, wood, styrofoam, and nylon would not burn?

I was being tongue-in-cheek facetious -- sheez! How much more obvious could I be!?

I'm just getting weary of explaining in detail, over and over to different people, how I understand lithium's risks, have studied the issue at great length before leaping in, think the risks minimal, and within my budget have done all I can do to minimize those risks.

If I had the money for the equipment, I'd have taken a course in metal working, and then fabbed fireproof aluminum battery boxes.