Red Sanyo 18650 Cells Getting Hot While Charging

Jon

Member
Joined
Aug 20, 2017
Messages
115
I think it would be a very game person to recommend a safe temperature, everyones situation is different.

Quoting from Battery University; Lithium-ion cells with cobalt cathodes (same as the recalled laptop batteries) should never rise above 130C (265F). At 150C (302F) the cell becomes thermally unstable, a condition that can lead to a thermal runaway in which flaming gases are vented.

http://batteryuniversity.com/learn/archive/lithium_ion_safety_concerns

I am NOT RECOMENDING 130C as a safe temperature, obviously you want to keep a very large margin from there!!
 

SemaJG

Member
Joined
Sep 6, 2017
Messages
45
50c is my absolute hottest temp when testing for inclusion in the powerwall when charging at 1A. I may try giving the Sanyo's a charge, a rest then a second charge as suggested by several on the thread who say many only heat badly on the first revival charge. 85c was my hottest so far.
What do you think is my 50c temp too high or too conservative?

Jimmy
 

DarkRaven

Active member
Joined
Sep 2, 2017
Messages
1,235
Each manufacturer has its own threshold, but generally most, if not all, cells are specified safe to 100C. To become unstable they have to get far hotter than 100C. However, that doesn't mean that 100C is a "normal" operating temperature. It might be under extreme loads right at the end of their specified charging or discharging currents, but it certainly shouldn't be under small loads like 1A charge/discharge. Everything around 40/45C I would consider pretty normal for a used laptop cells and new cells probably won't heat up at all under such circumstances.

jestronix said:
Tested another 20 or so, no hot ones, however i noticed that these cells are far happier at a 500ma charge rate. Not one has heated up. similar on the discharge, 500ma test. Think these red cells are old chem, and perhaps were never designed for 1000ma use. In my packs these guys will never see 500ma draw, so im happy to use them. outside my house in a steel shed :)

This has nothing to do with it. Of course they are rated for much more than 500mA of current. To find cells that are rated that low you have to go very far back in time.
While charging they don't reach the end of charge voltage. This isn't affected by current, at some point they will stop charging and turn excess power into heat. No matter how big the current is. If you turn down the charging current they will heat up slower, that is the only difference.
While discharging they behave completely normal, just like any other cell. If yours don't and you see a significant difference between a 500mA and 1000mA charge/discharge then your cell has other issues. This is then another matter entirely. Probably just a very worn out cell.
 

Korishan

Moderator
Staff member
Joined
Jan 7, 2017
Messages
6,321
50C is about as hot as I'll go, as long as I notice it. Sometimes the temp will rise rapidly withing 10mins or so. That's how I ended up with an 80-ish C cell one time. I had to re-glue my TP charger to the cell holder after that.
 

Oderus420

Member
Joined
Oct 1, 2017
Messages
243
50C is my max when you consider that a mass majority of cells are only hitting 32C.
 

daromer

Moderator
Staff member
Joined
Oct 8, 2016
Messages
5,384
Many cells say that you should stay below 60C when in use so i would say 50C is a good limit.
 

jestronix

Member
Joined
Nov 7, 2017
Messages
54
So tested another 20, no heaters :) 500ma cycle is giving me 1600mah on average from 4.1 to 3.0, these are 2000mah cells , so no bad really. I have enough for 15 to 20kwh, oh the fun starts now. Glad I got all these cells for free :) insane hours ahead. Anyone in Oz built their powerwall and looking to offload some opus ? I'm thinking 6 or so.
 

DarkRaven

Active member
Joined
Sep 2, 2017
Messages
1,235
jestronix, I'm sorry to be blunt, but please take a look into spelling the units correctly. These are case sensitive. It is mA, mAh and kWh. Reading it like this with all lower case letters sends a shiver down my spine :D
 

jestronix

Member
Joined
Nov 7, 2017
Messages
54
Getting some nice capacity tests from my cells 2100 mAh :) low iR too :). They get warmish but no firecrackers yet. I'm thinking these cells are 2200 mAh new ? Not sure if the liitokala 500 is being generous.

On further reading looks like these were 2400mAh new, they are green tops. 2013 - 14 I think.
 

howiegrapek

Member
Joined
Sep 25, 2017
Messages
59
Korishan said:
50C is about as hot as I'll go, as long as I notice it. Sometimes the temp will rise rapidly withing 10mins or so. That's how I ended up with an 80-ish C cell one time. I had to re-glue my TP charger to the cell holder after that.

Maybe I'm too cautious, but I wont keep any cell which gets hotter than 45 degrees C for any reason. AND, as of now, I won't use any individual cells in my walls which get hotter then 37 (the same temperature as human normal 98.6 F). With the number of good cells over 2200mAh, it is easy to be safe and smart. Of course, I also take into account capacity, internal resistance, time to discharge/charge and so on.

Kori, that is funny (albeit a bit disconcerting) that you had to 're-glue' after a cell got that hot. When I started down this journey, I quickly realized that I hated the idea that I have to monitor the first set of cells on the first test to weed out the heaters.So I built a smart tester. My "first test" bench uses the "charge-test" (just as opus defines) - After harvesting a new cell, assuming it has voltage and looks good, I log it and run first test suite. Amongst other things, the tester tracks the temperature through all cycles (checking every 30 seconds) and logs in the database for each cell.

Anyway, The tester is 16 cells, each in individual 1 cell holders. I automatically turn off the battery holder for any cell reaching 45 degrees during the test and send myself a page. I also send text pages to myself for every temperature rise of 3 degrees over 35 on any individual cell. This way, I don't have to continually monitor the cells during this test (which may take 12 hours or more per cell) - I can start them before bed, when I go to work, etc, and not worry too much. Yes, unattended battery testing (certainly for the first test) can be dangerous, so all of those are done in my detached garage, in a fire resistant area.

So, when I find hot cells (and yes, interestingly, most of them are red in color :), I write "heater" on them and put those in a separate container. Maybe I'll recycle them, maybe I'll go back and retest, whatever. Either way, with the number of cells I have which don't heat up at all, I don't waste my time on anything which might cause problems down the line.

I have discovered that some of the Sanyo's are perfectly fine. Some are not. Maybe they had a manufacturing issue (thus their recall). It is our job as DIYers to be smart - and to identify at the initial testing phase which ones are safe to use in our own walls.

Just be smart and be safe.

H.
 

BlueSwordM

Member
Joined
Jul 5, 2017
Messages
430
My personal threshold for cells is a delta T higher than 10C at 250mA. If it exceeds that threshold at that current, I consider it dead for almost all purposes.

I'm still doing some research about these Sanyo cells, but I have a hypothesis that still needs some more research and testing, as I'm waiting for a constant current battery tester from China.

My current hypothesis is that the the lithium concentration inside of the LiCo2 anode is higher than usual, and leads to a phenomenon called passivation layering, along with the lower purity of the cobalt compound, caused the problems we encountered in these cells. It's still not certain, but I could come to a conclusion in a few weeks.

Basically, this usually happens in lithium primaries due to their very high molar concentration of lithium, forming a passvation layer as the battery isn't used. This isn't really bad for them, as they aren't rechargeable, meaning it isn't a problem. In fact, it lowers their self discharge rate significantly.

I came to that statement, because I noticed that with my Sanyo cells from that time, their self discharge rate is very low, even for a lithium ion battery. Even my best new cells didn't have that low of a self discharge rate. So, I did some more research, and apparently, that behaviour is quite bad for lithium ion batteries on the first cycles, as by clearing that passivation layer, the internal resistance is quite high in the beginning, resulting in much more heat than usual if a battery was stored for a while. With the lower purity of the cobalt used, it resulted in an actual lower thermal runaway threshold.

That, along with charging a possibly months old pack along using the computer in a high heat environment, lead to the thermal runaways, and the massive recall.

This is what I've compiled for now.
 

jestronix

Member
Joined
Nov 7, 2017
Messages
54
On further testing i have discovered that the cells that get real hot, if you take them off and let them sit then attempt to charge a second time, they don't get hot. however voltage drop was huge. 4.2 down to almost 4v on most of them. on a full cycle they only got 1500mAh . So essentially the heater ones are good indicators that they are shit and can be tossed. Charging at lower rate may reduce the heat and let the test complete.

My thoughts are , rather than having to charge all the way up to 4.2, is there a quick test i can do. I will test if a large amp charge will induce heat in the shit cells and not the good ones. Gradually ramp up and see what i can get. Resistance isnt a give away in this situation.

As i push further into processing these cells, im seeing 25% toasters so far. i charge right up to 4.18 no heat yet, then i manually hang around for them to reach soak at 4.2 where they produce heat. quick check by hand and i throw them on the spot.

The non toasters are producing impressive mah compared to new, seeing 2200 from a 2400mah new cell from 2013! so not bad.

overall im mixing the 1900 to 2200 as a mix and program my use for 1700 mAh.

these red sanyos still have a place and you can salvage some great cells if your prepared it. im charging inside besa bricks with tin backing and screened off front, if one cell goes off, its only the charger and cells that get taken out. its all outside anyway.

thinking i might make a dedicated temp controlled charger for this job, that stops charging when heat reaches toaster level.
 

jestronix

Member
Joined
Nov 7, 2017
Messages
54
Some more testing tonight, I'm pretty sure I can now pick heaters out before waisting time with a capacity test. I watched the voltage on known heaters vs prime cells. The heaters voltage drops far quicker underload and rises far quicker on charge. Literally seconds of testing, after this the cells settle down. This will speed up my sorting way quicker, note on resistance I could see no difference on anything over 40. Anything under 40 was always a good cell, but I have the majority are over 40. So I need to still sort those.
 

Korishan

Moderator
Staff member
Joined
Jan 7, 2017
Messages
6,321
Personally, I'd test every cell. There are those here on the forums that had those heaters, and have had good cells after their first charge. But, if with heaters you're getting 25% bad rate, that's pretty good. As long as you have the needed cells to build without the need for them.
Now, you could use those post-heaters as flashlight cells or something like that. 1500mAh will go a long with an led flashlight.
Nice return on the good ones, 2200+ mAh on reclaimed is really good. Are they from new-old stock?
 

howiegrapek

Member
Joined
Sep 25, 2017
Messages
59
Yup, 25% is pretty good. I see much worse (My testers have temperature sensors on each cell and log/alarm accordingly). It depends on the batch of recycled packs I guess. I've previously seen more then 75% good tests on these. This weekend, I tested 100 red cells from sanyo (and also 16 orange ones) newly harvested - the orange ones were all good. The red ones, only 7 of them took a charge/discharge/charge full cycle without rising higher then 100 degrees f (38c). 9 of them stopped before 45. I stopped the test on all the others with temps higher then 45. One observation - the discharge cycle (or course) generated higher heat, and most didn't get hot until after 3 hours of discharging. All this done at 500ma. Like Kori said, it is worth testing everything because you just never know, but if testing time is tight, you might just skip the red ones until a later date.
 

jestronix

Member
Joined
Nov 7, 2017
Messages
54
For me there's no skipping the red cells :) mine are all red all 3000 ish of them. How do you have any details around your tester ? I was thinking of using compressor ebay controllers to throw a relay cutting power to both the pack and the controller. Was thinking of mounting temp sensors under each cell holder to get reasonable heat reading.
 

howiegrapek

Member
Joined
Sep 25, 2017
Messages
59
ya, I have also close to a thousand red cells sitting in a bucket. They all need to be tested. As for my testers, they are pretty custom, but the easiest thing to do is to modify them like daromer did (see his videos). This works and will give you temperature readings as you move forward. Of course, you'll have to remove hot ones manually, but It is easy enough to write a little script to read the temps and text you at a threshold. I can provide that if you'd like.
 
Top