Alarmingly hot lithium cells!

[iamdarkyoshi]

I have gotten into a bit of a lithium recycling hobby. I can get dead laptop batteries for free. I have been taking them apart, automatically rejecting cells under 3v (should I?) when I take the pack apart, and then I charge the rest at 650mA, spread across two cells. I currently just use some 2 cell powerbanks as chargers. They stop charging at 4.1v, oddly.

Anyway, I usually leave cell pairs together (don't break the spotwelded tabs) and charge them as a pair. However, this one pack seems to be weird. I have charged two pairs overnight, and in the morning, they were still charging, according to the powerbank. And they were HOT! About 50 degrees! I charge on a flamrproof surface in case things go a bit... flamey.

However, only one cell in each pair was hot, the other was just a bit warm.

So I broke the tabs off and then just let them cool down. The colder cells got to room temperature pretty much instantly. The hot ones, even after 20 minutes, still feel just as hot as they were before. But the voltage has only dropped to 4.05v. Wat.

Could these cells have been a different lithium chemistry in an 18650 form factor, or are they just royally dead?

[uncle_bob]

Hi

If the pairs of cells are in series, then the explanation is pretty easy. You don't say, but I'm guessing that your cells are set up in parallel. If so, the answer is simply that the hot cell has died.

Bob

[Gyro]

I have charged two pairs overnight, and in the morning, they were still charging, according to the powerbank. And they were HOT! About 50 degrees! I charge on a flamrproof surface in case things go a bit... flamey.

I think I'd be thinking of a bit more than a flameproof surface if charging unknown / old / suspect Lithium cells (particularly overnight!). You need protection from hot flames going outwards and upwards, not just burning the surface.

[iamdarkyoshi]

Hi

If the pairs of cells are in series, then the explanation is pretty easy. You don't say, but I'm guessing that your cells are set up in parallel. If so, the answer is simply that the hot cell has died.

Bob

They were in parallel. For gits and shiggles I threw the cold cell in my capacity discharge tester. Got 1.5Ah out of the cell. Not good enough for me to keep, but certainly not as bad as some of the other cells I have tested.

I have charged two pairs overnight, and in the morning, they were still charging, according to the powerbank. And they were HOT! About 50 degrees! I charge on a flamrproof surface in case things go a bit... flamey.

I think I'd be thinking of a bit more than a flameproof surface if charging unknown / old / suspect Lithium cells (particularly overnight!). You need protection from hot flames going outwards and upwards, not just burning the surface.

They were out in my garage on a cast iron tablesaw with nothing else within 5 feet of it. Tablesaw was dust free as I use a 1HP dust collector.

[Buriedcode]

Hmm yeah, don't charge lithium cells in parallel.  Doing so means that the differences between cells, and differing charge state cause different currents to run through each cell.  Its not only naughty and  frowned upon, its downright dangerous.  If you insist on charging them in groups, do it in series and monitor the voltages across the cells.

They could well have a 4.1V maximum charging voltage because some nickel flavours do indeed have that as the maximum cell voltage.  There are so many schematics for chargers floating about, and although some might think the 'constant current' phase is a luxury - it isn't.

I know you haven't paid for these batteries but give them some love, give them a proper charging algorithm!

[iamdarkyoshi]

Hmm yeah, don't charge lithium cells in parallel.  Doing so means that the differences between cells, and differing charge state cause different currents to run through each cell.  Its not only naughty and  frowned upon, its downright dangerous.  If you insist on charging them in groups, do it in series and monitor the voltages across the cells.

They could well have a 4.1V maximum charging voltage because some nickel flavours do indeed have that as the maximum cell voltage.  There are so many schematics for chargers floating about, and although some might think the 'constant current' phase is a luxury - it isn't.

I know you haven't paid for these batteries but give them some love, give them a proper charging algorithm!

I am using these cells in powerbanks. They are also what I am using to charge them, and they do have to be connected in parallel. I always try to leave them in the parallel groups they came from the laptop packs in so they are as matched as they can be.

I know it is not the most ideal thing to do, but if it worked in OEM HP/Dell/Asus batteries, I think it should work fine here. The weird part about the empty powerbank cases I bought is that they do actually only put out 4.1v. Maybe china FINALLY errored on the side of caution and made it so that manufacturing tolerances never go over 4.2v. But probably not 

[sleemanj]

Hmm yeah, don't charge lithium cells in parallel.

There is nothing wrong with parallel charging as long as the cells are healthy, are the same voltage when paralleled, and you preferably keep an eye on temperature.

This is exactly how laptop packs do it, thier cells (as shown by the op) are 1 or more parallel pairs (Ns2p) - each pair has a thermistor for monitoring.

[iamdarkyoshi]

Hmm yeah, don't charge lithium cells in parallel.

There is nothing wrong with parallel charging as long as the cells are healthy, are the same voltage when paralleled, and you preferably keep an eye on temperature.

This is exactly how laptop packs do it, thier cells (as shown by the op) are 1 or more parallel pairs (Ns2p) - each pair has a thermistor for monitoring.

And the powerbanks are no different other than the lack of temp monitoring. But they are charged at such a low current that it is not much of an issue (325mA per cell)

[CatalinaWOW]

Those cells were well matched when they left the factory.  That is not evidence that they are currently well matched.  You have already demonstrated that some are highly mismatched.   I know that it is a lot of work and time, but if you want to use these salvage batteries the best plan is to separate them, charge them separately and then do discharge curves on the cells so that you can pick matched pairs for your battery packs.  I strongly believe that if you go through this procedure you will end up with very few cells going back with their original partners.  I wouldn't even venture to guess if you can get a set well enough matched to use in your battery bank.

[iamdarkyoshi]

Those cells were well matched when they left the factory.  That is not evidence that they are currently well matched.  You have already demonstrated that some are highly mismatched.   I know that it is a lot of work and time, but if you want to use these salvage batteries the best plan is to separate them, charge them separately and then do discharge curves on the cells so that you can pick matched pairs for your battery packs.  I strongly believe that if you go through this procedure you will end up with very few cells going back with their original partners.  I wouldn't even venture to guess if you can get a set well enough matched to use in your battery bank.

I have been testing them individually. My tester spits the voltage and current through the COM port every second, so plotting a curve wouldnt be difficult. However, every "pair" I have done (aside from these oddballs) have had amp-hour ratings within 0.05Ah of each other. Not sure of the discharge curve though. But this is certainly not an excuse to just put any two cells together.

As for the "good" batteries, I recycle anything under 2Ah, so like 95% of them lol

Once the mail turtle delivers the rest of the crap I bought on ebay, I will be building a 10 cell charger that charges each cell individually at 1A. I probably should add a temp cutoff, but I won't. 

[iamdarkyoshi]

Well, I bought some temperature protection for my upcoming 10 cell quickcharger. Some bimetalic switches should do nicely. All I could find in a quantity of 10 were 50 degree or higher ones from a US supplier. 50 degrees is higher than I would like, but it is much better than no thermal protection at all, am I right?

[uncle_bob]

Well, I bought some temperature protection for my upcoming 10 cell quickcharger. Some bimetalic switches should do nicely. All I could find in a quantity of 10 were 50 degree or higher ones from a US supplier. 50 degrees is higher than I would like, but it is much better than no thermal protection at all, am I right?

Hi

Some is always better than none

It doesn't take much time to find videos of lithium battery fires. It's rare to find one that comes with a proper analysis of what the root cause actually was. I think caution is a really good idea in this case.

Bob

[Prime]

Hmm yeah, don't charge lithium cells in parallel.

There is nothing wrong with parallel charging as long as the cells are healthy, are the same voltage when paralleled, and you preferably keep an eye on temperature.

This is exactly how laptop packs do it, thier cells (as shown by the op) are 1 or more parallel pairs (Ns2p) - each pair has a thermistor for monitoring.

And paralleled cells should have even discharge and charge behaviour if they were well matched at the factory. They normally come in matched batches from the manufacturer. Well that's how our Boston and Panasonic cells for the solar car were supplied.

However it's possible one could have been damaged through welding.

Well, I bought some temperature protection for my upcoming 10 cell quickcharger. Some bimetalic switches should do nicely. All I could find in a quantity of 10 were 50 degree or higher ones from a US supplier. 50 degrees is higher than I would like, but it is much better than no thermal protection at all, am I right?

You can build a simple tester and charger with a quality RC battery charger and some cell holders from Keystone. Plug the 6 or 8 cell lead from your board into the RC charger and it will ensure they all get charged relatively evenly. We used a 4 port hitec on an external power supply to do 24 cells at a time.

And you should spot weld them back together after you are done pulling them apart. Solder tends to overheat them.

[iamdarkyoshi]

Hmm yeah, don't charge lithium cells in parallel.

There is nothing wrong with parallel charging as long as the cells are healthy, are the same voltage when paralleled, and you preferably keep an eye on temperature.

This is exactly how laptop packs do it, thier cells (as shown by the op) are 1 or more parallel pairs (Ns2p) - each pair has a thermistor for monitoring.

And paralleled cells should have even discharge and charge behaviour if they were well matched at the factory. They normally come in matched batches from the manufacturer. Well that's how our Boston and Panasonic cells for the solar car were supplied.

However it's possible one could have been damaged through welding.

Well, I bought some temperature protection for my upcoming 10 cell quickcharger. Some bimetalic switches should do nicely. All I could find in a quantity of 10 were 50 degree or higher ones from a US supplier. 50 degrees is higher than I would like, but it is much better than no thermal protection at all, am I right?

You can build a simple tester and charger with a quality RC battery charger and some cell holders from Keystone. Plug the 6 or 8 cell lead from your board into the RC charger and it will ensure they all get charged relatively evenly. We used a 4 port hitec on an external power supply to do 24 cells at a time.

And you should spot weld them back together after you are done pulling them apart. Solder tends to overheat them.

I am going to use ten of those 1A USB charger boards connected to a 25A 5V PSU. I bought a bunch of cell holders. I can individually charge cells on it and have a thermal cutoff on each cell. I don't need to put any connections back on them, the powerbanks have sprint contacts. If I use cells in applications where I need wires connected, I just leave half of the tab on them and solder to it. I have a tab spotwelder, but no tabs

[uncle_bob]

Hmm yeah, don't charge lithium cells in parallel.

There is nothing wrong with parallel charging as long as the cells are healthy, are the same voltage when paralleled, and you preferably keep an eye on temperature.

This is exactly how laptop packs do it, thier cells (as shown by the op) are 1 or more parallel pairs (Ns2p) - each pair has a thermistor for monitoring.

And paralleled cells should have even discharge and charge behaviour if they were well matched at the factory. They normally come in matched batches from the manufacturer. Well that's how our Boston and Panasonic cells for the solar car were supplied.

However it's possible one could have been damaged through welding.

Well, I bought some temperature protection for my upcoming 10 cell quickcharger. Some bimetalic switches should do nicely. All I could find in a quantity of 10 were 50 degree or higher ones from a US supplier. 50 degrees is higher than I would like, but it is much better than no thermal protection at all, am I right?

You can build a simple tester and charger with a quality RC battery charger and some cell holders from Keystone. Plug the 6 or 8 cell lead from your board into the RC charger and it will ensure they all get charged relatively evenly. We used a 4 port hitec on an external power supply to do 24 cells at a time.

And you should spot weld them back together after you are done pulling them apart. Solder tends to overheat them.

I am going to use ten of those 1A USB charger boards connected to a 25A 5V PSU. I bought a bunch of cell holders. I can individually charge cells on it and have a thermal cutoff on each cell. I don't need to put any connections back on them, the powerbanks have sprint contacts. If I use cells in applications where I need wires connected, I just leave half of the tab on them and solder to it. I have a tab spotwelder, but no tabs

Hi

Tabs should be a quick thing to get from China (for dirt cheap ...). It likely would take some quality shopping time to find the right ones.

Bob

[iamdarkyoshi]

Hmm yeah, don't charge lithium cells in parallel.

There is nothing wrong with parallel charging as long as the cells are healthy, are the same voltage when paralleled, and you preferably keep an eye on temperature.

This is exactly how laptop packs do it, thier cells (as shown by the op) are 1 or more parallel pairs (Ns2p) - each pair has a thermistor for monitoring.

And paralleled cells should have even discharge and charge behaviour if they were well matched at the factory. They normally come in matched batches from the manufacturer. Well that's how our Boston and Panasonic cells for the solar car were supplied.

However it's possible one could have been damaged through welding.

Well, I bought some temperature protection for my upcoming 10 cell quickcharger. Some bimetalic switches should do nicely. All I could find in a quantity of 10 were 50 degree or higher ones from a US supplier. 50 degrees is higher than I would like, but it is much better than no thermal protection at all, am I right?

You can build a simple tester and charger with a quality RC battery charger and some cell holders from Keystone. Plug the 6 or 8 cell lead from your board into the RC charger and it will ensure they all get charged relatively evenly. We used a 4 port hitec on an external power supply to do 24 cells at a time.

And you should spot weld them back together after you are done pulling them apart. Solder tends to overheat them.

I am going to use ten of those 1A USB charger boards connected to a 25A 5V PSU. I bought a bunch of cell holders. I can individually charge cells on it and have a thermal cutoff on each cell. I don't need to put any connections back on them, the powerbanks have sprint contacts. If I use cells in applications where I need wires connected, I just leave half of the tab on them and solder to it. I have a tab spotwelder, but no tabs

Hi

Tabs should be a quick thing to get from China (for dirt cheap ...). It likely would take some quality shopping time to find the right ones.

Bob

Yeah, I will buy some eventually. Not entirely sure why I made a spotwelder but didn't buy any tabs

[ez24]

I am going to use ten of those 1A USB charger boards connected to a 25A 5V PSU. I bought a bunch of cell holders. I can individually charge cells on it and have a thermal cutoff on each cell. I don't need to put any connections back on them, the powerbanks have sprint contacts. If I use cells in applications where I need wires connected, I just leave half of the tab on them and solder to it. I have a tab spotwelder, but no tabs

"sprint contacts"   do you mean spring contacts ?   I cannot find anything on sprint contacts

Can you link to your 1A USB boards

thanks

[iamdarkyoshi]

I am going to use ten of those 1A USB charger boards connected to a 25A 5V PSU. I bought a bunch of cell holders. I can individually charge cells on it and have a thermal cutoff on each cell. I don't need to put any connections back on them, the powerbanks have sprint contacts. If I use cells in applications where I need wires connected, I just leave half of the tab on them and solder to it. I have a tab spotwelder, but no tabs

"sprint contacts"   do you mean spring contacts ?   I cannot find anything on sprint contacts

Can you link to your 1A USB boards

thanks

Yeah, spring contacts. As for the charger boards, they are just the dirt cheap TP4056 boards.
http://www.ebay.com/itm/252335209689?_trksid=p2057872.m2749.l2649&ssPageName=STRK%3AMEBIDX%3AIT
Yeah I am going to charge these things in a flameproof environment

[ez24]

Keep us updated.  I have several used batteries so I think I will try the same thing.

[uncle_bob]

Yeah, I will buy some eventually. Not entirely sure why I made a spotwelder but didn't buy any tabs

Hi

The tabs really are not anything special at all. It should be a "pay a dollar and get a bunch in the mail (including free gift)" sort of thing. The ones you get may not last for a hundred years in a salt spray environment. I doubt that is a big risk with used batteries.

Bob

[iamdarkyoshi]

Looked up a datasheet for the hot cells. They were Sanyo cells, rated at 3.6v instead of 3.7, but still said to charge up to 4.2v. I got sixteen laptop batteries yesterday and tore them apart.

I have been rejecting cells under 3v out of the packs, is this a good plan or should I keep lower voltage ones as well?

Also, should I bother keeping cells that are between 1 and 1.9 Ah? I have been recycling ones under 2Ah. Would anybody be interested in lower capacity cells due to being used?

[uncle_bob]

Looked up a datasheet for the hot cells. They were Sanyo cells, rated at 3.6v instead of 3.7, but still said to charge up to 4.2v. I got sixteen laptop batteries yesterday and tore them apart.

I have been rejecting cells under 3v out of the packs, is this a good plan or should I keep lower voltage ones as well?

Also, should I bother keeping cells that are between 1 and 1.9 Ah? I have been recycling ones under 2Ah. Would anybody be interested in lower capacity cells due to being used?

Hi

I would not reject anything until you try a test charge at a constant current. If it's below 3V after maybe an hour at 1A  then reject it. If you have the time, I'd save the good lower capacity cells. You never know whey them might make a good trade for something you are looking for.

Bob

[juanfermed]

I think those cells getting hot could be well explained by an internal short circuit. The short circuit produces heat generation inside the cell. The cells are probably in a point in which the short circuit is still small enough to not make the cell enter in thermal runaway, and the the heat generation rate is still lower than the heat dissipation rate. You are still able to charge and discharge the battery because the short circuit might not consume all energy.  Does it makes sense with your experiments?

[sleemanj]

I have been rejecting cells under 3v out of the packs, is this a good plan or should I keep lower voltage ones as well?

Under 2.7v I would discard.

Under 3.2v I would charge at a very low rate at least the first time.

Also, should I bother keeping cells that are between 1 and 1.9 Ah?

Sure, if you have a use for them, it's 3-6 Watt Hours, it's probably better than most of the "new" Ultrafire, Trustfire etc ones that come from China.

[Siwastaja]

If you discard at 3.0V, you are discarding cells that are potentially just fine.

For voltages between about 2.5V and 3.2V, run a conditioning cycle - charge at C/100 until you get to about 3.3V, then let the cell rest for a few hours to measure the voltage again. If the cell is taking more than a few hours to reach 3.3V (or is not reaching it at all), or if the voltage drops back below 3.2V after this slow charge, discard the cell.

Cells slightly below 2Ah might be just fine, they may have been lowish capacity (think about 2.2Ah!) to begin with, in which case they may be in very good shape.

Old, recycled laptop cells, especially when paralleled, should be rated at low currents, something like C/2 absolute maximum, C/5 preferred. Build high-energy, low-power packs from these cells.

If you rate a 2Ah cell at 1A, you can't exactly rate two paralleled 2Ah cells at 2A because they don't share the load perfectly, but something like 1.5A instead.