Bizarre Li-ion cell voltages

[splin]

My neighbour just bought some NOS laptop battery packs for the cells. He came round today clutching a Sanyo cell and his mutimeter. He'd put 4 cells in series (I presume he'd charged them prior to the test) and used them to power a 12V LED floodlight. The cells were probably around 2.6Ah but I didn't note the model number. As it happens I'd measured that lamp a few days before and it draws about 30W - it has 16 LEDs and must have a buck regulator as current halved at 24V.

At the start the battery voltage presumably would have been 4 x 4.2 = 16V, so the current draw would have been approx 1.88A at the start, increasing as the battery voltage dropped. At the end of one hour he stopped the test and measured the cell voltages. To his, and my surprise, 3 of the cells were reading around 3.93V - this after drawing at least 1.9A hours!
 
Even more surprising wa the cell he brought round - with his meter it was reading 4.42V! Clearly his meter was faulty so I checked with mine and measured 4.43V!

What is going on? I'm baffled.   

I'm fairly sure these aren't protected cells so not seeing the ouput of faulty or unstable protection circuits. Are they in imminent danger of bursting into flames? Is my neighbour about to be become a multi-bilionaire having discovered cells that increase their capacity the more you draw from them? That would present an interesting philosophical question - would you allow mankind to draw on this apparently limitless power source, risking ultimate cataclysmic destruction of the known universe when the battery exceedes its (unpredictable) capacity limit at Giga/Terra/Peta/Exa or whatever Joules?

[EDIT] fixed a few typos.

[llkiwi2006]

There are LiPo cells capable of operating at 4.35V or more. See https://en.wikipedia.org/wiki/Lithium_polymer_battery#High-voltage_LiPo_cells_with_silicon-graphene_additive

But you should check what kind of cells you have and make sure you are not overcharging them.

[splin]

There are LiPo cells capable of operating at 4.35V or more. See https://en.wikipedia.org/wiki/Lithium_polymer_battery#High-voltage_LiPo_cells_with_silicon-graphene_additive

But you should check what kind of cells you have and make sure you are not overcharging them.

Interesting reference, thanks, but do you know any Li-ion cells which could have a terminal voltage of 4.43V after nearly 2Ah of discharge - no matter what voltage you managed to charge them up to?

[llkiwi2006]

There are LiPo cells capable of operating at 4.35V or more. See https://en.wikipedia.org/wiki/Lithium_polymer_battery#High-voltage_LiPo_cells_with_silicon-graphene_additive

But you should check what kind of cells you have and make sure you are not overcharging them.

Interesting reference, thanks, but do you know any Li-ion cells which could have a terminal voltage of 4.43V after nearly 2Ah of discharge - no matter what voltage you managed to charge them up to?

Didn't catch that the 4.43V was after discharging. That's sounds pretty impossible, I have no idea what's going on. Must have been using a batteriser

[Siwastaja]

Most likely a test setup/measurement error.

Low batteries on poorly designed multimeters cause readings too high, although it would be a strange coincidence to have two multimeters read similarly wrong...

[The Soulman]

Did he wire that cell backwards in the string when discharging to the light?
The light would have running of aprox. 8 volts then, and would probably also use less current.. 

[Siwastaja]

Did he wire that cell backwards in the string when discharging to the light?
The light would have running of aprox. 8 volts then, and would probably also use less current.. 

A real possibility. In this case, the cell is severely overcharged, and possibly dangerous.

[Psi]

Did he wire that cell backwards in the string when discharging to the light?
The light would have running of aprox. 8 volts then, and would probably also use less current.. 

Not to mention that reversed cell being 'charged up' from the other 3 cells as they discharged,

Almost definitely what happened.

[splin]

Did he wire that cell backwards in the string when discharging to the light?
The light would have running of aprox. 8 volts then, and would probably also use less current.. 

Promote that man to Sergeant! I think you hit the nail on the head. At 8V the lamp consumes .9A so the cell would have had at least 30% over charge. The surprise to me is that you can even charge a Li-ion cell to 4.43V without really bad things happening.

I'm sure its lifetime has been seriously compromised but after discharging it (he left it at 4.43V for around 1 hour), he says it seems to be functioning ok, but he hasn't yet had a chance to do a proper capacity test on it. I have suggested to him that he shouldn't keep it in the house!

[amyk]

The surprise to me is that you can even charge a Li-ion cell to 4.43V without really bad things happening.

Look at the datasheets for an 18650 and you'll see the various safety tests which they pass. Here's a typical one:

Overcharge test
Test method: To charge with 20A-20V at 25? for 3hr.
Criteria: No fire, and no explosion.

[llkiwi2006]

The surprise to me is that you can even charge a Li-ion cell to 4.43V without really bad things happening.

Look at the datasheets for an 18650 and you'll see the various safety tests which they pass. Here's a typical one:

Overcharge test
Test method: To charge with 20A-20V at 25? for 3hr.
Criteria: No fire, and no explosion.

Is that for a protected cell? I can't see any way a normal cell wouldn't explode under this circumstance. Surely the thing would blow up from just the heating alone.

[Nusa]

The surprise to me is that you can even charge a Li-ion cell to 4.43V without really bad things happening.

Look at the datasheets for an 18650 and you'll see the various safety tests which they pass. Here's a typical one:

Overcharge test
Test method: To charge with 20A-20V at 25? for 3hr.
Criteria: No fire, and no explosion.

Is that for a protected cell? I can't see any way a normal cell wouldn't explode under this circumstance. Surely the thing would blow up from just the heating alone.

Sounds like a good experiment for someone. Note the test criteria doesn't say anything about the cell surviving, just not burning or exploding.

[llkiwi2006]

The surprise to me is that you can even charge a Li-ion cell to 4.43V without really bad things happening.

Look at the datasheets for an 18650 and you'll see the various safety tests which they pass. Here's a typical one:

Overcharge test
Test method: To charge with 20A-20V at 25? for 3hr.
Criteria: No fire, and no explosion.

Is that for a protected cell? I can't see any way a normal cell wouldn't explode under this circumstance. Surely the thing would blow up from just the heating alone.

Sounds like a good experiment for someone. Note the test criteria doesn't say anything about the cell surviving, just not burning or exploding.

There are plenty of videos on youtube of lipo batteries exploding from overcharging, e.g. https://www.youtube.com/watch?v=MtdH7FWKYSI,

[Nusa]

The surprise to me is that you can even charge a Li-ion cell to 4.43V without really bad things happening.

Look at the datasheets for an 18650 and you'll see the various safety tests which they pass. Here's a typical one:

Overcharge test
Test method: To charge with 20A-20V at 25? for 3hr.
Criteria: No fire, and no explosion.

Is that for a protected cell? I can't see any way a normal cell wouldn't explode under this circumstance. Surely the thing would blow up from just the heating alone.

Sounds like a good experiment for someone. Note the test criteria doesn't say anything about the cell surviving, just not burning or exploding.

There are plenty of videos on youtube of lipo batteries exploding from overcharging, e.g. https://www.youtube.com/watch?v=MtdH7FWKYSI,

But the goal of most of those videos is to make it fail with extremely high currents. This one started at 40A, and were clearly impatient enough to turn it up after a few minutes. Show me one that stays within the test parameters (and isn't a fake cell to start with) if you want to convince me.

[amyk]

Those are pouch cells, which are definitely far more volatile.

Look at this:

[llkiwi2006]

Show me one that stays within the test parameters

The first video was 2A at 20V, definitely within the test conditions.


Ok the 18650 cells do seem a whole lot more robust. They can still smoke and leak when shorted though, so it would be interesting to see how they react to being overcharged.

I found a video for overcharging 18650 cells, they didn't explode but they did get pretty damn hot, so it probably would've caught fire if kept going for longer.

[digsys]

Having spent many 100s hrs on Li packs for EVs / Solar cars, I find those figures VERY fishy !!
From ~ 3.9V - 4.25V (or higher) there's only ~ 2-5% of the cells capacity. This drops a lot when drawing >1C, definitely not the current / time you note.
A cell in poor condition MAY jump back up in voltage (surface charge / electron migration) when taken completely off-load ie for testing
Apply a 10mA load and see what you get. The highest voltage I've ever managed to charge Li cells is 4.35V, with a resultant 95+% loss of cycles.
Sometimes we do it to win a race. After that, they are totally unstable. Somethings fishy.

[NiHaoMike]

The highest voltage I've ever managed to charge Li cells is 4.35V, with a resultant 95+% loss of cycles.

That seems to be common with newer smartphones. I wonder if that's part of how they ensure the batteries don't last much longer than the typical 2 year upgrade cycle. I generally don't charge lithium batteries to more than 4.1V/cell or so.

[digsys]

The highest voltage I've ever managed to charge Li cells is 4.35V, with a resultant 95+% loss of cycles.

That seems to be common with newer smartphones. I wonder if that's part of how they ensure the batteries don't last much longer than the typical 2 year upgrade cycle. I generally don't charge lithium batteries to more than 4.1V/cell or so.

Naaaa, they don't need to screw around with charging profiles. Too risky anyway. They've been caught with more imaginative ways to "shorten"
battery life :-)  4.1V is perfectly fine, you only lose ~ 1+% max of capacity. Some systems use the 4.0 - 4.25V voltage profile to do a "top match",
but in all my research, it's pointless. The BEST matching is done down between ~2.4 - 3.1V. Biggest chemistry changes happen there.

[amyk]

I found a video for overcharging 18650 cells, they didn't explode but they did get pretty damn hot, so it probably would've caught fire if kept going for longer.

100C is burning hot to a human but not that hot in terms of electronics; you can see the voltage drop down to near 0, which means the fusible separator has melted and nearly stopped the ion flow. The temperature will decrease after that. If you look at the overcharge temperature graphs of most 18650s you'll see the same thing.

The 4.35V cells use a slightly different chemical composition, so that charging to that voltage doesn't result in a massive capacity degradation nor damage.