Salvaged 18650 cell: Good voltage, low capacity

[etc]

I have a number of 18650 salvaged from various laptops. I keep the ones that charge to about 4.1V-4.2V and don't get hot during charging. I use the Pila charger with success.

I tried a runtime test with 1x18650, about 2300 mAh, and that Nichia GS 5mm P60 "battery vampire". I know Nichia takes 3.7V max but it doesn't appear to mind being overdriven a bit. I got only about 8 hours out of it.
According to my estimate, it should have a good runtime of at least 20 hours or maybe much more due to the very low drain, I think about 25mAh? I ran it overnight on a fully charged 18650 cell and after about 8 hours, the cell was completely depleted, down to 3.7V.

I know something is obviously wrong here as you can take some P60 dropin module like Malkoff M60LL and get close to 15-18 hours with it. And it's much brighter at 50 lumens vs 10-15 lumens of the Nichia GS. Something is inconsistent.

In fact, I ran M60LL on 1x18650 overnight, on AW 2600 cell which I know is OK. I got it down to 4.0V after about 6 hours. It was either 4.2V or 4.1xx to begin with. It's 1 year old so it's possibly it lost a little capacity but generally works well.

I will try running the test of Nichia GS on 1x18650 AW, or maybe unprotected Panasonic NCR18650 which I also know are good and see if I get much better runtime.

Conclusion: You can take a salvaged laptop 18650 cell and despite it charging fully to 4.2V, it may not have the full capacity. These old cells are usually 2200-2600 max and given how often they are several years old, it would not surprise me that the actual capacity is much lower, below 2,000 mAh, maybe considerably lower.

 

[45/70]

Depending on what charger is used, just because a cell charges to 4.20 Volts, doesn't really tell you much of anything. While not ever having owned a Pila, I know from the algorithm that it uses, that typically, older cells likely won't charge up to 4.20 Volts, so that is a bit odd.

The easiest way to evaluate the condition of a LiCo Li-Ion cell, is to charge it up to 4.20 Volts (not always possible, depending on the charger) and see how much the voltage drops after the cell rests a bit. New or nearly new cells will not drop in voltage at all, or very little. Deteriorated cells will drop in voltage considerably. When a cell drops 0.2 Volts, or more, it is considered to be at EOL.

The best way to determine the health of a cell, is to do basically what you're doing, a "real life" test, or discharge them on an analyzer. You didn't mention if cell voltage drops after charging. If it does, this may help to determine the condition of your cells.

Dave

 

[PCC]

The easiest way to evaluate the condition of a LiCo Li-Ion cell, is to charge it up to 4.20 Volts (not always possible, depending on the charger) and see how much the voltage drops after the cell rests a bit. New or nearly new cells will not drop in voltage at all, or very little. Deteriorated cells will drop in voltage considerably. When a cell drops 0.2 Volts, or more, it is considered to be at EOL.

I just learned something new. All of the 18650 cells that I have are salvaged from used laptop battery packs and almost all of them drop to 3.9V after a few days of resting. The thing is that they all work fine but I've never tried to drain them down to a resting voltage of 3.7V to check their overall condition. Maybe I need to make a rig for such testing and start to weed out the bad cells?

 

[alpg88]

Maybe I need to make a rig for such testing and start to weed out the bad cells?

when you do let us know how it works.
i too have around 50 cells from laptops, i build a holder for 1 and just hooked to 1A load while measuring V drop over time, i can't say it is efficient or accurate way to measure cell condition.
i would really like to measure true capacity of my cells, may be you will find a better way, than i did.

 

[flatline]

I just learned something new. All of the 18650 cells that I have are salvaged from used laptop battery packs and almost all of them drop to 3.9V after a few days of resting. The thing is that they all work fine but I've never tried to drain them down to a resting voltage of 3.7V to check their overall condition. Maybe I need to make a rig for such testing and start to weed out the bad cells?

Personally, I'd pitch those cells in the recycle bin.

When I evaluate a salvaged cell, I charge it, measure the (open circuit) voltage. Then I let it set a couple of hours and measure the voltage again. If its below 4.0v, I pitch it. If it's still good, I wait a couple of days and measure the voltage again. If the voltage continues to creep down more than, say .02 volts, I pitch it on the assumption that it has internal shorts.

Internal shorts are the enemy since they can cause your cell to explode when charging.

--flatline

 

[etc]

I discard any cell if it drops in voltage even a little bit. None of my existing salvaged 18650s drop in voltage at all, in a few weeks anyway. I check them all every few weeks. That's the interesting part.

Yes, Pila can and does charge these cells to anywhere between 4.10-4.20V, depending on the cell. If it charges to something like 4.06V, I use it once and discard it. Apparently the voltage is not the sole criterion to determine how good the cell is and how much capacity it has. I suspect my cells are really low in capacity, somewhere between 1,000-2,000 mAh.

I have a lot of 1800 mAh cells and they must be 5 years old. They do light any powerful module I have, the voltage is good and doesn't drop. However the runtime will suffer. Still, I got them for free, so no complaints there and they make excellent emergency / power outage lights. Run them down to nothing and discard them if you have to.

Also keeping in mind the load of that 5mm Nichia is non-existent. So it should have performed especially well. I will re-run the test tonight with a known good cell, such as Panasonic NCR18650 cell, of 2900 mAh capacity. I have just a few of these cells (versus 100 salvaged cells) and they both charge to almost 4.2V. One charges to 4.14V and the other 4.17V. I recall they used to charge to 4.2V when new. So I wonder if they are too getting a little tired after 1 year of usage.

 

[etc]

When I evaluate a salvaged cell, I charge it, measure the (open circuit) voltage. Then I let it set a couple of hours and measure the voltage again. If its below 4.0v, I pitch it. If it's still good, I wait a couple of days and measure the voltage again. If the voltage continues to creep down more than, say .02 volts, I pitch it on the assumption that it has internal shorts.

Internal shorts are the enemy since they can cause your cell to explode when charging.

--flatline

In my case, it it has been obvious. A cell will usually go down drastically, in a day or two, from 4.1V to 3.9V or something like that.
Still if it doesn't, it doesn't complete validate that it's good.

Although it's conceivable I somehow missed the cell I used in the test and it somehow discharged without me knowing it. I will go over all of them one more time. Unfortunately my multimeter is not fine enough and cannot check hundreds of Volts, just 4.2, 4.1V... etc.

 

[jasonck08]

Almost every 18650 pull I've tested is ~60% of the rated capacity and should be tossed. I did open up a pack that had 2400mAh Samsung cells, 2-3 cells were dead (due to water damage) and the other cells still had about 90% of their original capacity. Generally speaking 18650 salvaged cells are useless. Not only is the capacity bad, but I've noticed very high internal resistance in cells causing voltage drop to be a big problem. If the cell is under about 80% then recycle it.

 

[flatline]

I've got 11 salvaged cells, all around 5 years old or so. I don't have any idea what any of their capacities are, but they all last a month or so in my L-mini's before getting below 3.8v with daily use, so I figure I'll keep them.

I don't have any high-drain uses. Maybe that's why they perform so well for me.

--flatline

 

[etc]

I re-run the test. This is P60 drop-in moded with Nichia GS 5mm, a very low powered module obviously. Think Gerber Infinity Ultra type lumens, maybe twice as bright. 15-20 lumens max and maybe not even that much.

I used 1x18650 AW cell, in FiveMega host. Ran it 9PM-9AM. The cell went from 4.2V - 4.0V or to 50% of the capacity in 12 hours. I would guesstimate that it would go for another 12 hours, meaning the total runtime should be 24 hours.

Keeping in mind my AW 18650 2600 mAh cell is about a year old and probably lost 20% of its capacity as I always store it fully charged.

This runtime is good, but not great. Malkoff M60LL can go almost as much (20 hours) but it is several times brighter. I suppose the real reason for the 5mm 'battery vampire' is that it can run below 3.7V down to 2.5V or below. So if willing to discharge the cell down to 2.5V, it can go well beyond 24 hours of runtime. I did test it at 2.8V and it's awfully dim. Think 1/4 of Gerber Infinity Ultra lumens.

This proves to me that: There is nothing wrong with the module but something really wrong with the salvaged 18650 cell. It's a 2350 mAh cell. It appears to have just a fraction of its intended capacity, maybe as low as 1/4 to 1/3.

 

[45/70]

It sounds to me like your cells are suffering from degradation, for sure. As you mentioned before, voltage is not the only method of evaluating a cell's general condition. Actually running them to see how they perform works better. The voltage drop after charging method is much simpler of course, and has always been sufficiently accurate for me.

......The cell went from 4.2V - 4.0V or to 50% of the capacity in 12 hours. I would guesstimate that it would go for another 12 hours, meaning the total runtime should be 24 hours.

Generally, with LiCo Li-Ion cells in reasonably good condition, an OC voltage of 4.00 Volts corresponds to about 75-80% remaining capacity. This would mean your runtimes would be quite a bit longer than your estimate.

Dave

 

[etc]

I see, then that means good news.

I like used 18650 cells. The capacity is apparently vastly reduced, but they were free and I think I have about a 100 of them. Surefire 6P or 1 x 18650 is the ideal EDC format for me anyway.

Nevertheless I will buy some new high capacity 2900 mAh cells.

 

[KiwiMark]

Almost every 18650 pull I've tested is ~60% of the rated capacity and should be tossed. I did open up a pack that had 2400mAh Samsung cells, 2-3 cells were dead (due to water damage) and the other cells still had about 90% of their original capacity. Generally speaking 18650 salvaged cells are useless. Not only is the capacity bad, but I've noticed very high internal resistance in cells causing voltage drop to be a big problem. If the cell is under about 80% then recycle it.

I pulled 8 cells from a ~5 year old Sony battery pack that was not performing well in the laptop. From my tests on the hobby charger it seems that the capacity is anywhere from 50% to 75% depending on how high a discharge rate I test them with.

They DO hold their charge though - I charged them and read 4.18-4.19V on each cell after taking them off the charger, after several weeks they still read 4.18-4.19V - in fact not one of the 8 reads under 4.18V.

They are supposed to be 2Ah cells but at 2A discharge I get little more than half that and the voltage drops pretty darned quick right from the start of the test. At 1A discharge they run for longer and the voltage doesn't drop so quickly - but they still only give around 1400mAh. They are probably fine for running single cell flashlights that typically run for over 2 hours on a good cell and would definitely be OK on lower modes.

I would NOT be keen to trust these in multi-cell applications though - a little caution with old & unprotected Li-ion cells couldn't hurt!