Will this li-ion battery work for lights ?

[dieselsmoke]

I have access to quite a few of these batteries for free. They're from PC server RAID controllers, and advertised as cells made in Japan, 3.7v, 7wH (equates to almost 2000 maH). The first I took out of its case was showing only 1.9 volts, but after a minimal charge it came up to 3.89 and drove a Q5 emitter at .89 amps for a few minutes without issue.

So it seems ok, and I know it's not protected, but could there be anything else fundamentally wrong with using these cells and possibly building a pack ? At least as a test or backup pack ?

 

[The Dane]

Yeah, individual capacity!

Try imagine a 4 wheel drive with 4 different motors.

In series the weakest link goes down first, but the wihecle continues despite, at a slower pace.

In paralell the weakest link shorts the remaining 3 out.

So unless the cells are of the same capacity then ganging them up is a waste of time!

Either series or paralell will be termined by the weakest cell.

 

[The Dane]

Me is beyond 0,8 so sorry for crappy English.

 

[dieselsmoke]

Thanks for the reply Dane.

The batteries are all identical and have never been used, so "theoretically" they would have the same capacity. I guess the only way to know is by opening a few up and testing them individually.

One question that came into my mind is the discharge rate. Are some li-ion cells able to feed the emitters at a different rate than other cells, and are there possible safety concerns if these cells were to be discharged too fast by driving emitters ?

 

[Benson]

Try imagine a 4 wheel drive with 4 different motors.

In series the weakest link goes down first, but the wihecle continues despite, at a slower pace.

In paralell the weakest link shorts the remaining 3 out.

Not really; either I'm seriously misunderstanding your analogy, or you've got it almost exactly backwards. You can parallel cells of any capacities, they'll discharge at proportional rates to keep the voltage balanced. For series connections, you should monitor the voltage of each cell to ensure they don't overdischarge. The best way to accomplish this is with a multi-cell protection circuit, but you can obviously get by without it if you can be careful enough, as in a pack for testing purposes, on the bench where a voltmeter is always handy and you check the cells manually...

The risk of overdischarging one cell in a series pack without noticing the drop in overall voltage or (in unregulated lights) in light output goes up with the number of cells; there's no hard-and-fast rule, but my feeling is 2s is fine, 3s is pushing it, and 4s or more is dangerous. That's assuming unprotected LiCo cells, charged separately or balanced; protected cells let you get away with more (as each cell voltage is monitored then), series-charging without balancing is more dangerous (because they don't even start at the same state), and safe chemistries will just ruin a cell with no fireworks...

Discharge rates: Assuming these to be typical LiCo cells, something like 1.5C or 2C, so 3A or 4A, would be the safe limit. This does vary for different cells, so you really would need to find the manufacturer's datasheet to know for sure. And for a series connection, the whole pack has the same capacity, so the same max discharge; for parallel, the pack has the sum of the capacities, so you sum the max discharge currents.

From your question it sounds like you mean direct-drive, which is highly dependent on components. With cells that size, you're probably more likely to fry the LEDs than the battery, but the safe way is to use some driver to ensure you use the current you want, and don't fry either of them. ;)

 

[mdocod]

Your risk tolerance in the end will probably be the deciding factor.

Looks like those batteries might have like 3 cells side-by-side there eh?

sitting at 1.9V for a long time is hard on LiCo cells, if that's what they are (I'm assuming)....

Testing would be required to determine their viability for being put back into service. Most here will probably recommend against it and I would certainly have no argument with that position.

You would need to charge them up and see if they hold a charge above 4.0V pretty well.. I would suggest charging them up all the way for the first time in a place where a chemical fire with hazardous fumes is not a problem, (like outside in a large fire-proof container or something)... You'd want to see if they heat up during charging, as this can be an indicator that they are in bad shape...

Driving LEDs directly off li-ions can be tricky business and yes, some cells are going to drive LEDs with more current than others if all other variables remain equal, and yes, there are limits on how quickly you can safely drain a LiCo cell. You should probably just assume a 1C maximum discharge limit (1 hour discharge time no less) since you do not have more information about the cells, and since they are old. Equally important, is how much current the LED can actually handle. Usually a careful selection of LED Vf and a resistor in series with the circuit is required, but not always. Different cells have different effective resistance. In your case, it appears that you have a stack of several cells in parallel in each battery, which has the effect of further reducing effective resistance, which may be offset by the fact that these cells are old and probably have higher internal resistance than they once did. If you're getting 0.89A across an LED with the battery charged to 3.89V, then you can probably expect much higher potentially damaging current to flow when the battery is fully charged unless you introduce some resistance to that circuit.

-Eric

 

[dieselsmoke]

Benson, mdocod, thanks very much for the detailed feedback.

I hadn't even thought about the possibility that there could be multiple cells inside the pack. It's an aluminum casing with the red tape around it, so I fugured it was just some sort of odd form factor single cell.

I'll be charging a cell to full capacity today - outside on the patio, the cell in an old steel military ammo box with the lid latched but not closed. I think that should be a safe location in case of venting.

Based on your ideas and other reading, I doubt I'll use the battery for anything beyond bench testing. Good quality protected cells are just too cheap to risk this pack flaming in my camelback during a ride !