Replacing a cell group with different cells?

[Prawn76]

Hi Chaps,

I've got 1 defective cell group in my 10yr old 36v E Bike battery.

The cells - Panasonic CGR18650CG 2250mah/10A, have been discontinued.

The closest cell's I can find in the UK (for a reasonable price) are Samsung 22P's 2200mah/10A. Does any one see an issue (particularly safety wise) with adding these, as a new cell group to the pack?

Many thanks in advance

 

[SafeDiscDancing]

After ten years your real capacity in those remaining cells are probably down to about 1500mah.

So anything new will be adequate.

Just be really careful about getting a good clean balance before you get too far into running the replacement cells.

If you run into troubles it will be due to gross mismatches of balance.

I would suggest you either top balance or bottom balance everything before you go asleep and let your BMS try to sort everything out.

Bottom balancing is the safest way to start... but few choose that.

Top balancing means you need to get everything up to near it's breaking point and get it leveled out.

 

[Prawn76]

Thanks SafeDiscDancing,

Any tips for bottom balancing 18650's?

All the best

 

[SafeDiscDancing]

Thanks SafeDiscDancing,

Any tips for bottom balancing 18650's?

All the best

Well... you need an access point to each parallel group.

Then you need to either add or subtract from the group.

Best to get your battery near bottom before you begin doing anything.

Then establish some number like exactly 3.00 volts and go about getting each group to that level.

To add you can use an old USB charger which puts out 5 volts and usually about 2 amps.

To subtract you need a resistor that can handle it.

I use four 0.1 ohm 100 watt rated resistors in series.

https://www.ebay.com/itm/100W-5-0-1-1000-Ohm-Watt-Shell-Power-Aluminum-Housed-Case-Wirewound-Resistor/274694078858?pageci=52a2603b-15f7-4b3b-840b-8bfc30b3abac&redirect=mobile

 

[eMark]

Any tips for bottom balancing 18650's?

The primary purpose this LENGTHY post is to give you some idea why Bottom Balancing is a time consuming pain. Because a dependable 10S Balance Charger can cost $350 a better solution for you may be new 36V 20Ah(50-700W) with a 3A Charger that's less than $300. (see Bottomline: link)

The following explains in more detail SDD's above post and why most don't bottom balance. Normally, it's not necesssary and even if helpful it's possibly not worth the effort on a pack near EOL. Anyway here's my attempt to dissuade you :wink: FWIW ...

The top photo is one way i've experimented with bottom balancing as a 10S pack using two 5S JST-XH 20awg balance leads (before bulk charging), but it's time consuming ... so only used that method a few times. Being this 10S can easily be split into two 5S3P packs i prefer using Y-cables and my 2S-6S Balance Charger as shown below. All of the 10S p-groups are within 25-30mV of each other after bulk charging. So far after 2 yrs and 250 c/d cycles no need to top balance after bulk charging at 1.5A rate. As you can see i don't use a BMS or balancer board, but to explain why would require another thread.

First measure voltage of each of the 10s parallel groups (via the two 5S JST-XH connectors). Then bottom balance the 10s parallel groups within 10-12mV of each other using Balance Charger set at 1S and 1.5A charge rate. Then when all 10 parallel groups are withing 10-12mV of each other i bulk charge the two 5S3P packs to 20.5V (41.0V when reconnected as a 10S3P battery pack) ...

Above checking individual p-group voltages. Bottom method for charging p-groups as needed (on avg. 6 out of 10 to balance all).

My reconfigured 5S JST-XH connector as 2-pin connector for charging whichever of the parallel groups until voltage variance is no more than 10-12mV from one another. On average takes from 15 minutes to 30 minutes to bottom balance the 10S p-groups before bulk charging. So far no need to top balance as 10S p-groups are within 25-30mV of each other after bulk charging at 1.5A rate.

_________________________

Some type of resistance discharge method may be more practical for bottom balancing your p-groups (within 30mV of each other) before bulk charging. I've used the following discharge technique for individual cell experiments as well as bottom balancing the p-groups before bulk charging. However, have come to prefer the above charging method as it's more reliable using Balance Charger set at 1S, charge rate of 1.5A and tracking the elapsed time. Having experimented with both methods for bottom balancing have come to prefer charging just those p-groups in need of a little more voltage until all 10S p-groups are balanced within 10-12mV of each other. Having done it so many times one gets fairly proficient at bottom balancing in as little time as possible.

IF there was a 10S Balance charger that didn't cost at least $300 then Bottom Balancing would be a more viable option than top balancing with a BMS or attaching provision for a Balancer Board(IMO). Your're getting an idea why bottom balancing is frowned upon (time consuming pain) unless for experimental [learning] purpose.

Assume you've checked the voltage of each of the other nine parallel groups to see how much variance in p-group voltages and whether or not another p-group is suspect? Being your pack is 10 yrs old it may not be worth your time to replace one defective p-group (except as a learning experience). You may find yourself having to replace other p-group cells in the not too distance future.

Bottomline: You may be further ahead to buy a new 36V battery pack for $300 instead of going to all the effort to bottom balance the p-groups before bulk charging on a 10yr old battery that is close to its end of life. A 10S Balance Charger like the iCharger 4010 Duo is $350 ... https://www.progressiverc.com/products/icharger-4010duo?gclid=Cj0KCQiA3fiPBhCCARIsAFQ8QzW_AT5CghoO-4MtyFEc0qssXp6MxH4UnysdAoyvNGUb9y_LMnU_Ir0aAls7EALw_wcB ... what about investing in a new 36V 20Ah with 3A Charger as better option than trying to repair a battery near its EOL - https://www.amazon.com/UPP-Ebike-Battery-Pack-20AH/dp/B081PN5WZ2/ref=sr_1_8?crid=22EM6TYCE2UF5&keywords=unit%2Bpack%2Bpower%2B36v&qid=1644100271&sprefix=unit%2Bpack%2B36v%2Caps%2C651&sr=8-8&th=1

Even with my bottom balancing routine with my experimental pack it's possible i won't get more than 4-5 years (125 cycles a year) out of my experimental Vruzend 10S3P (Samsung 30Q) pack. Even so i've had to replace nine "141" 30Q cells after 125 cycles in 2020, and six more "141" 30Q cells needed to be replaced after 125 more cycles in 2021 (total 250 cycles so far after just 2 yrs), but then they are high energy 30Q cells which are more prone to suffer from high self-discharge.

Normally bottom balancing ISN'T necessary with Grade A cells, a 36V pack charged to just 41.0V instead of 42.0V and proper care/use of the battery pack. Think most would agree that getting 10 yrs of use out of your pack of Panasonic CG cells it's now time to get a new pack. The on-going frustration of having to occasionally repair an older pack (even though Panasonic) probably isn't worth the effort ... but it's always a worthwhile learning experience ... if one has the patience.

 

[e-beach]

Upon WOT, the old cells will sag and the new ones won't which will put stresses on the new cells. Just don't let them get hot while they stress trying to make up for the sag of the old cells.

I recommend retiring your old pack and getting a new one.

 

[SafeDiscDancing]

There is quote from the movie version of Brave New World (1998) which starred Leonard Nimoy by the way.

"Don't mend it, End it."

The idea being that you must consume and be happy.

Seems the "peer pressure" wants you to consume... buy... spend some money... upgrade... get more technology.

Free will means you must choose.

I have an old A123 LiFePO4 pack which is ten years old and I've replaced some cells over the years because I was not bottom balancing and kept destroying cells.

But now that pack is working at 90% or better of original capacity because I finally know what I'm doing.

It really comes down to what you like doing.

Do you like gaining a mastery over a topic or do you "punt" and go Brave New World?

-------------------

The quickest way is the four 0.1 ohm resistors in series. (400 watts)

You can drop the voltage of a 10ah parallel string by half a volt in a minute or two.

All this stuff they sell is really weak and cannot drain fast enough because you need 400 watts worth of heat dissipation.

So the cheap and quick... down and dirty... is the direct approach.

But yeah you do need an access point and the typical balancing wires and connectors are really hard to work with.

 

[e-beach]

.....

I have an old A123 LiFePO4 pack which is ten years old and I've replaced some cells over the years because I was not bottom balancing and kept destroying cells.

But now that pack is working at 90% or better of original capacity because I finally know what I'm doing.
.........

yea, but the OP isn't talking A123 LiFePO4. He's talking what ever chemistry his 18650's happen to be. LiCoO2, or LiMn2O4 or who knows what. In any case they don't last like LiFePO4.

I myself have a Headway LiFePO4 pack that is about 10 years old and it runs at 85% of original capacity, but the internal resistance is much greater now then when it was new and it really sags on acceleration. Nothing at all like when new. It will be perfect for something else but not for an ebike motor any longer.

I still say the OPer should get a new pack and use his old batteries for something else.

 

[SafeDiscDancing]

I still say the OPer should get a new pack and use his old batteries for something else.

Nothing wrong with having extra packs either.

He can fix this one for next to nothing and have a new one as well.

I have two packs to pick from on my older ebike with one at 600 wh and the other at 300 wh.

Generally I prefer the 300 wh for less weight.