Quick 18650 question.

[Apex]

Newb here, first post.

I'm currently harvesting some 18650 cells from laptop battery packs. As I test them most are above 3.0 volts, but occasionally I'm running across some around 2.6 or 2.3v. When I charge them for 10/15 min they then show well above 3v. I was told some of these packs have been sitting for 3 to 4 years maybe more.
Of course I will cycle them and retest before they get used in a pack.

So, if a cell shows 2.3v but can be briefly charged to hold 3.5v is there a chance it could be used?

Btw, great site, glad I found it.

 

[Hillhater]

Most likely, but you will need to test capacity, voltage retention, etc..
Read this...
https://endless-sphere.com/forums/viewtopic.php?f=14&t=26383

 

[parabellum]

I would recommend to not use them in a pack, the cell has suffered a permanent damage and may have formed some nasty dentrines, waiting to cause internal short with every cycle, compromising rest of the pack and integrity of your home.
You could use them in flash lights and other less critical application though but handle them as little igniter that can go off any time.

 

[Apex]

Thanks.
I read a lot of that link provided, still working thru it.

I definately planned on doing further testing before any cells were "approved" for use in a pack.
That said, is 2.8v after years in storage is the lowest voltage acceptable for a ebike pack build? Because I'm seeing quite a few at 2.6ish. I do understand that damaged li-ion cells can be a risk, I have years of experience with RC lipos.

If any cells below 3v are dangerous, and it's going to be a crapshoot with no exact test to weed out the dangerous ones then screw this, I'll go back to buying new lipos.

 

[Hillhater]

Some 18650s have a normal operating range down to 2.5v ( though not much to use down that low "). So low voltage is no indication of a cells "safety"
If it charges normally, can hold a charge , (4v for a week or more) without dropping more than 0.1 volts or so), and discharges 75% or so of its original stated capacity,..... Then it should be useable.
..But you can never tell how long "old" cells will last. You have to monitor and check frequently.

 

[Apex]

Thank you, Hillhater.

I guess I'll pick up an intellegent charger or two for these cells, I see the b6 is used a lot and economical.
I'll keep pouring through that battery info link, good stuff. Hell of a learning curve, for sure.

 

[Skrzypas]

I don't see a problem with cells at 2 V. It's important to check if they heat up during normal charging (with a rate you plan to use) and discharging. Standard laptop cell that you are looking for should be completely cold at 1 A charge and discharge and holds min. ca. 80% of its original capacity.

 

[Tats]

I've cannabalised one of those packs and am using the cells in my bike light at 2s3p. Seem to be holding up well but all were above 3v when I started

 

[redilast]

A while ago I discovered some Fujitsu new old stock packs with Sanyo cells in them. All the cells were in the 2.1-2.6v range. The cells had 10 year old date codes. I've designed a small charging circuit in the past so I know that any time a cell is <3v its best to soft start the charge cycle and charge at <100mA until the cell is above 3v. I did that to several cells, fully charged them then tested their capacity. Capacity was 98-99% of what the datasheet maximum was. Then I used my internal resistance tester and the cells were just under 50 mili ohms AC 1Khz Pulse which is roughly the same as new fresh Samsung 26F's.

I believe the guys at battery university say that recovering a cell that has been under 1.5v for more than a week is a no-no. And like <2v for a long period of time is not a good idea. But 2v+ is probably ok, and that has been my experience. Zero noticeable or measurable cell deterioration from a cell sitting in the low 2v range for who knows how long.

 

[Apex]

This is good news then. Thanks for the replies.

I'm marking down initial voltage at harvest, then I will be sure to gently recharge them all for the first capacity tests. In the name of safety, I am converting an old barbecue into my charging station with a fire extinguisher nearby, all initial charging of the questionable cells will be done outdoors.

Now I need to figure out my C rate for the controller/motor combo I'm using, the cyclone 300 kit from sick bike parts. I had initially planned on a 48v pack, most of the cells seem to be rated at 1.5C discharge.

 

[Overclocker]

these 17s6p were taken from 3M projector packs. out of 38 packs only 3 duds, the bms tripped i.e. at least one cell in the pack went low.

the duds all charged up fine and had capacities practically the same as the good ones. same internal resistances too (icharger).

but the duds didn't go into the pack. i just didn't want to risk it. there's a reason they dropped in voltage

 

[Hillhater]

wow ! Nice score on those packs.
Do you know what cells, make and model, capacity etc are in them ?
17s ?.... The packs appear to have printed on them "10.8 volts, 2100mA, ??

 

[fechter]

I've had cells that were reading zero volts work well after charging. Sitting around the BMS in the pack drained them slowly over years. These low cells need to be charged at a very low rate (a few mA) until they get above 3.0v or so, then can be charged at the normal rate.

 

[999zip999]

To charge at low rate make a quick big parelle string in 2p for slow charge with same voltage cells to 3.1v. I wouldn't use lower then 2.5v and 8p for 20 amp controller. How many you got 13s. A good back up bike and it should last.
Do you have a spot welder ? If not leave the tabs on.

 

[Apex]

The 48v controller states 40 amp continuous 100 max.
So far Ivery harvested about 60% of the laptop packs I have, this has yielded almost 80 cells holding above 2.8v
They are mostly samsung, lg, and Sony cells, Google tells me they are all 2600mah and what data I can find shows a C rate of 1.5.
I plan on buying a spot welder (709a?), currently researching which Chinese welder is somewhat reliable.
And yes, I will definitely slow charge the cells the first time before the leak down and capacity tests.

 

[Overclocker]

wow ! Nice score on those packs.
Do you know what cells, make and model, capacity etc are in them ?
17s ?.... The packs appear to have printed on them "10.8 volts, 2100mA, ??

Amperex (ATL) INR18650 2200mah. got them via amazon last year. about $1 per cell. not bad i suppose...