Red Sanyo 18650 Cells Getting Hot While Charging

jestronix

Member
Joined
Nov 7, 2017
Messages
54
I think there's a small market for thermal controlled testers like this. All chargers sold should by law have thermal cut per cell, its mad really that any battery charger has no thermal cut.

The thermal fuses in laptop batteries are set at 85c on my packs, they are also automatic reset ones too. They are not stuck to the cells but sit against some masking tape, so I'm thinking the cells could get a lot hotter before it would fail. Thermal runaway is at 120c ? I'll have to look at some lower temperature ones that could be hooked to a relay and simply fail the circuit. Keeping costs low per ccharge bay.
 

daromer

Moderator
Staff member
Joined
Oct 8, 2016
Messages
5,391
Most laptops i looked at have that attached to the cells.

Thermal runaway is depending on chemistry. 85C seem high? Its normally not recommended to go above 60c
 

DarkRaven

Active member
Joined
Sep 2, 2017
Messages
1,235
I don't know any examples, I did never really care, but 85C is believeable for me. This is still on the safe side of things. It's a bit high and you could even argue that it makes no sense to give the cells this much room as they shouldn't even reach 60C in this use case.
Most cells are specified to be safe up to 100C. Some even go higher without any danger, it depends on the cell. There is no general rule like "thermal runaway happens at xC".
 

BlueSwordM

Member
Joined
Jul 5, 2017
Messages
430
Thermal runaway can either be very violent, or very swift and not dangerous depending on the temp of the cell.

For example, most of our cells are composed of a lithium cobalt compound. This is a very volatile chemistry, and dangerous thermal runaway happens at 150C.
It can occur at 90C, but is a long and not too dangerous death.
 

alfu

Member
Joined
Jan 12, 2017
Messages
40
It's no secret there is a trade-off between capacity and charge/discharge current. For applications in 8-cell powerbanks where the charging is 1A from a USB port, so that each cell sees 1/8 amp, these excell (pun intended) because they have such high capacity.

Test them for capacity, charge them in a cheap Chinese spring-loaded powerbank so you can get them up to 4.2V (Opus won't do that), and TEST THEM FOR VOLTAGE RETENTION OVER TIME, and you will end up with a bunch of great little Li-ion cells.
 

jestronix

Member
Joined
Nov 7, 2017
Messages
54
So I decided to grab some of my toasters and cycle them at 300ma. I was quiet surprised to see these cells pull 90% of their stated capacity. They also were happy to reach their 4.21v top without any heat.

As we know low current gives better capacity but these results are impressive. in a low current cycle environment these cells could have plenty of life left.

However heat is an indicator that these cells are not as healthy as their cool brothers, yet when unstressed they deliver the same capacity. So next question is cycle life, will they happily punch out the same cycles ? I doubt it but what happens if they are 80% as good ? Do they still have a place, specially if they don't heat up.

I plan on running a couple of tests and maybe run a number of good and bad cells for a few hundred cycles at 300ma. many will say chuck them and they may be right, but I think a cycle test is called for first.

Ive had a few cells at 1000ma charge up and complete their discharge test and then turn into heaters when they try to charge back to 4.2.

Another issue I'm having is I can't cycle test these cells without constantly being around to detect the heaters, even though I run these in a fire proof environment I'm not keen on my chargers getting destroyed . At 300ma I'm not seeing any heaters. At 1000ma I see about 25%.
 

Herbi

Member
Joined
Sep 2, 2017
Messages
32
mike said:
1958greyhound said:
Free ;) since no one wants them.
You might be able to talk me into paying for the shipping.

I can get $0.50/cell selling them on eBay. And by that, I mean selling the honestly, blatantly pointing out everywhere in big bold red font that they got hot and are being sold for scrap.

Any idea what those buyers use these cells for? Thinking of selling my deselected cells too. Thanks
 

Geek

Well-known member
Joined
Aug 15, 2017
Messages
934
Herbi said:
mike said:
1958greyhound said:
Free ;) since no one wants them.
You might be able to talk me into paying for the shipping.

I can get $0.50/cell selling them on eBay. And by that, I mean selling the honestly, blatantly pointing out everywhere in big bold red font that they got hot and are being sold for scrap.

Any idea what those buyers use these cells for? Thinking of selling my deselected cells too. Thanks

They make excellent torch batteries. You just need to charge them slow or cutoff at 4.1v. Many of them still hold upward of 1500mah
 

rev0

Active member
Joined
Oct 3, 2017
Messages
369
I realize this is an old thread and there are many others out there, but I think this is the most relevant in terms of trying to find out *why* these cells are heaters. I decided to cycle test a cell I discovered was a heater (the infamous Sanyo red ring, UR18650F I believe) to see how it behaved, since there were reports of these cells being fine after a few cycles. I am using a custom built tester that can graph current/voltage/temp/capacity and measure DC internal resistance (~250ms pulse method) with 45 deg. C temperature cutoff for safety. The results so far seem to indicate that the cell does indeed stop being a heater after the first (or second) cycle, depending on your choice of cutoff current (I'm using 50mA), and the coulombic efficiency (that is, how much capacity you get out over how much you put in) is increasing with each cycle, though total capacity seems to be dropping off quickly.


image_hcwynb.jpg


This is still a heavily degraded cell, and it takes about 3 hours to charge, which would be unacceptable for any reasonable use. If anyone has a less degraded heater they'd like me to try I would be happy to give that a go. I'll continue cycling this one in the mean time.
 

Geek

Well-known member
Joined
Aug 15, 2017
Messages
934
rev0 said:
I realize this is an old thread and there are many others out there, but I think this is the most relevant in terms of trying to find out *why* these cells are heaters. I decided to cycle test a cell I discovered was a heater (the infamous Sanyo red ring, UR18650F I believe) to see how it behaved, since there were reports of these cells being fine after a few cycles. I am using a custom built tester that can graph current/voltage/temp/capacity and measure DC internal resistance (~250ms pulse method) with 45 deg. C temperature cutoff for safety. The results so far seem to indicate that the cell does indeed stop being a heater after the first (or second) cycle, depending on your choice of cutoff current (I'm using 50mA), and the coulombic efficiency (that is, how much capacity you get out over how much you put in) is increasing with each cycle, though total capacity seems to be dropping off quickly.

This is still a heavily degraded cell, and it takes about 3 hours to charge, which would be unacceptable for any reasonable use. If anyone has a less degraded heater they'd like me to try I would be happy to give that a go. I'll continue cycling this one in the mean time.

Do these self discharge if left to sit? I find that most of my heaters, while they won't charge above 4.1v they don't tend to self discharge below 4.05v. As stated before - they do make excellent torch batteries.
 

rev0

Active member
Joined
Oct 3, 2017
Messages
369
Geek said:
rev0 said:
I realize this is an old thread and there are many others out there, but I think this is the most relevant in terms of trying to find out *why* these cells are heaters. I decided to cycle test a cell I discovered was a heater (the infamous Sanyo red ring, UR18650F I believe) to see how it behaved, since there were reports of these cells being fine after a few cycles. I am using a custom built tester that can graph current/voltage/temp/capacity and measure DC internal resistance (~250ms pulse method) with 45 deg. C temperature cutoff for safety. The results so far seem to indicate that the cell does indeed stop being a heater after the first (or second) cycle, depending on your choice of cutoff current (I'm using 50mA), and the coulombic efficiency (that is, how much capacity you get out over how much you put in) is increasing with each cycle, though total capacity seems to be dropping off quickly.

This is still a heavily degraded cell, and it takes about 3 hours to charge, which would be unacceptable for any reasonable use. If anyone has a less degraded heater they'd like me to try I would be happy to give that a go. I'll continue cycling this one in the mean time.

Do these self discharge if left to sit? I find that most of my heaters, while they won't charge above 4.1v they don't tend to self discharge below 4.05v. As stated before - they do make excellent torch batteries.

I haven't checked this. I've got 2 other cells that were in parallel in this pack so presumably they're in the same condition, I'll do some more testing on the other 2.
 

rev0

Active member
Joined
Oct 3, 2017
Messages
369
Just an update, after 43 cycles the coloumbic efficiency is still high but the capacity has continued todropoff rapidly.


image_scaikz.jpg



image_snmron.jpg
 

Geek

Well-known member
Joined
Aug 15, 2017
Messages
934
rev0 said:
103 cycle update. Still heading to nil capacity, ETA around cycle 165.

Fascinating. That is handy to know. Pete cycled some 18650s100 times and they only showed a small loss in capacity. Looks like those Sanyo Heaters are just no good for anything. Aside maybe a torch battery.
 

Cherry67

Active member
Joined
May 13, 2018
Messages
495
I agree that is a very very useful investigation.

I would like to ask (sorry, i didnt read all), if the IR has been measured at any time ?
 

Wolf

Well-known member
Joined
Sep 25, 2018
Messages
1,282
Chiming in here with some research data.
My chart of known heaters shows that IR has something to do with it.
Here is the chart again.

image_uodlsk.jpg

and the box of(Notice the Color inequality)

image_jogrti.jpg


Also a spreadsheet of tests that I have been doing with predictions of mAh rating with just measuring IR(Some are wrong predictions some are pretty close)with a 4 wire Kelvin testerover the last week this is just a snip of it but you notice the Sanyo on the bottom with 70.4m? it is defiantly a heater even at 500mAas I check and charge all my cells at 500mA.
So this Sanyo cell overnight had to get to 4.2 for the LiitoKala to do a discharge test and record the mAh of the cell and then recharge the cell.
This morning the cell was hot not scalding but definitly way hotter than the others. I did not go to the shed to get my Temp gun which I will have next to me in the future do document the temperatures. From the collection of data that I have accumulated albeit not a huge dataset I have come to the preliminary conclusion that a cell with over 70m? IR esp. Sanyo has a high potential of being a heater.

image_cvpozi.jpg


Picture of the Lab

image_axellg.jpg


Wolf
 

Wolf

Well-known member
Joined
Sep 25, 2018
Messages
1,282
rev0 said:
103 cycle update. Still heading to nil capacity, ETA around cycle 165.


image_ysfuky.jpg

Cherry67 said:
I agree that is a very very useful investigation.

I would like to ask (sorry, i didnt read all), if the IR has been measured at any time ?

Yea Cherry the IR shows in the chart on the left it starts at 120m? and is heading to 150m?

Wolf
 

Cherry67

Active member
Joined
May 13, 2018
Messages
495
Sorry....

And matching.
 

stefan1025

New member
Joined
Dec 12, 2018
Messages
3
As someone working for a company that is using 4s3p battery packs with the Sanyo UR18650A in theier product, i can tell you we face to issue of overheating cells too. After investigating the issue for entire year now, i like to share me insight with you.

As we know, the cell voltage starts to drop after the cell hits 4050 mV to 4200 mV and it releases a condierable amount of heat.


image_wkidbx.jpg


Despite the BMS cuts off the charge current once the temperature hits 60C, the temperature keeps rising for several minutes. The reaction that was triggered by charging the defective cell is self-sustaining even if the charging is supended. The highest temperature I recorded is 93C.

It's also clearly visible, that not all cells reacs this way:


image_yqilyd.jpg


First we let some defective cells analyze by a specialized company. This is the original summary:
"The answer is that there are several micro-shorts appears from the negative electrodes through the separator to the positive electrodes. It is clear that negative electrode, separator and positive electrode are glued together. The reason behind may be due to long time charging and discharging at high temperature. However, it is very hard to obtained any evidence. In the same pack, several cells has the similar micro-shorts appearance when they opened the cells and checked under the microscope."

Yes, internal short circuits where kind of expected as this is the only way the cell can produce any heat on it's own. Also the self discharge rate of defective cells is about 3 times higher than the one of new cells.

So, the UR18650A just doesn't like high temperatures. If you look at old datasheets, the cell was rated for a charge temperature up to 45C. In the newer once only 40C. Panasonic knew for several years, the cell gets damaged when it is charged at high temperatures (but they didn't tell us, so we run into the issue and had to investigate it by ourself).

As a test, I took several brand new battery packs and charged them to 4.2V/Cell and 0.6C at an ambient temperature of 50C to 55C with one or two cycles per day. And sure enought, after 280 days the self-discharge rate of some cells rised so high, the BMS struggled to keep the battery pack balanced despite active balancing build in.After just 334 days and 440 cycles the fist cell showed the exothermic effect.
We also run a second set of battery packs at the same ambient conditions, but only charge them to 4.1V/cell, and they are still fine even without cell imbalance.

But after seeing a lot of reports from failed battery packs, I think not charging at high temperatures is the problem, but keeping the cell constantly at 4.2V at high temperatures (50C to 55C). The batteries typically fail after 18 to 36 month and as we use the batteries only as back-up, so they have a low number of cycels, usually around 50. I even saw a battery pack failing after just 15 cycles (and 2 years age)!
Oh, and the capacity is still fine, even 80% and more for some defective packs.

Time for a summary:
- Don't keep the UR18650A charged to 4.2V at high temperatures.
- The cells develop a internal short circuits. This effect is permanent and can not be reversed with conditioning cycles. Treat cells that come from a high temperature environment with extrem care.
- You can avoid the triggering of the exothermic effect of defective cells by reducing the charge voltage, charge current and ambient temperature. But the parameters are different for every cell and there is no "safe" area.
- Despite a saw a lot of defective batteries, non of them reached the termal runaway (charge current always cut off at 60C, pretty sure if you keep charging it will eventually catch fire). But don't trust your life on my experience. The cells are very dangerous and we immediately take the failed battery packs out of service when we detect an anomaly.

I hope this information helps you.

What I'm interested in, have you ever had a cell that hit the thermal runaway?


Geek said:
Do these self discharge if left to sit? I find that most of my heaters, while they won't charge above 4.1v they don't tend to self discharge below 4.05v. As stated before - they do make excellent torch batteries.

The self discharge rate of defective cells is about 3 to 6 times higher compared to brand new cells.


Wolf said:
Chiming in here with some research data.

Great, thank you.
 

bogptrsn

New member
Joined
Oct 27, 2017
Messages
29
DarkRaven said:
I never had any cells that got warm, let alone hot, during charging, if charged at max 1A. That means no noticeable heat when touching and hardly any measurable temperature difference in relation to the environmental temperature.
Is this still true for you?
 
Top