Amount of cells? many or few

[MK2R]

Flippy, have you not read the thread at all? I am testing them all before i am using them...

It's a little hard to follow because the info is spread throughout your replies.

So you have:
100x Samsung 26J rated for 2600mAh 5.2A that test at 2400-2500mAh?
100x Panasonic ??? rated for ??? that test at 2500-2600mAh?
500x Chinese FST "rated" for 2000mAh that average at 2016mAh?

yeah, i am sorry. yes it is pretty spread out mainly you and me typing

I have:
100x Samsung 26J 5.2A
100x Panasonic NCR18650 PF 10A
700x FST 2Ah under testing

I am not so sure about the tested cells i got thou. i did read somewhere that u should fully charge your cells, let them wait for some time (1hr) and then capacity test them. i tried it on my FST and doing it instantly you rated 100-150mah lower than if u let them stay rest at fully charged for 1 hour. I tested my samsung and panasonic this way too, the samsung was at 2611 and the panasonic almost 2800. i am not sure what i should do with this knowledge. i went and did the nor test once again after charging them and the samsung did turn out at 2507 (tested earlier at 2478) and the panaconic read at 2598 (don't remember earlier tested but something like that).

I mean, as long as they are paralel strings are balanced it's whatever i guess. if all cells are tested equally scewed in one direction or the other. but what i shall do with the calcualtion of the capacity, idk. do i manually need to retest them all again :?

 

[MK2R]

Maybe. i think i might re-evaluate some things. maybe i should start out with a smaller pack, continue testing the FST cells (those 700cells cost less than the samsung and panasonic together). at 700 cells 1A output is going to be 2.6kw nominal.

Just going to mix the samsung and panasonic, at 13s i can manage 195 cells, average 2.5A. would be a 13S15P. at nominal 5.2A (samsung will be the weaker link) i could take 5.2A per cell, at 3.7V that's about 20w per cell, it is a total of >3.7kw nominal output.

So how does that sound instead?
13S15P nominal output 5.2A
Capacity 1,8kwh
ouput 2kw would be discharge at 1.1C

and it would weight in at just "just" 8,8kg batteries and be way easier to move around. a 700 cell pack is kinda heavy when u think about it

 

[fatty]

i did read somewhere that u should fully charge your cells, let them wait for some time (1hr) and then capacity test them. i tried it on my FST and doing it instantly you rated 100-150mah lower than if u let them stay rest at fully charged for 1 hour. I tested my samsung and panasonic this way too, the samsung was at 2611 and the panasonic almost 2800. i am not sure what i should do with this knowledge. i went and did the nor test once again after charging them and the samsung did turn out at 2507 (tested earlier at 2478) and the panaconic read at 2598 (don't remember earlier tested but something like that).

I mean, as long as they are paralel strings are balanced it's whatever i guess. if all cells are tested equally scewed in one direction or the other. but what i shall do with the calcualtion of the capacity, idk. do i manually need to retest them all again :?

Yes, the capacity testing spec is to wait an hour.

I think you are overcomplicating this though. This is not a performance-critical application -- there is no reason to build so close to your required capacity that the difference/variance in capacity is relevant.

 

[fatty]

Maybe. i think i might re-evaluate some things. maybe i should start out with a smaller pack, continue testing the FST cells (those 700cells cost less than the samsung and panasonic together). at 700 cells 1A output is going to be 2.6kw nominal.

Just going to mix the samsung and panasonic, at 13s i can manage 195 cells, average 2.5A. would be a 13S15P. at nominal 5.2A (samsung will be the weaker link) i could take 5.2A per cell, at 3.7V that's about 20w per cell, it is a total of >3.7kw nominal output.

So how does that sound instead?
13S15P nominal output 5.2A
Capacity 1,8kwh
ouput 2kw would be discharge at 1.1C

and it would weight in at just "just" 8,8kg batteries and be way easier to move around. a 700 cell pack is kinda heavy when u think about it

-I'm too poor to buy cells and not use them.
-I'm too busy to test 900 cells and not use them.
-I'm too busy to worry about babying such a pack, and rebuilding it.

If you're already testing every cell, then the money and time cost is already sunk, regardless of how you big build the pack, so I'd build it as big as possible. Build it into a wheeled cart, a lift-able pallet, or just build it in place.
-7-8 Samsung
-7-8 Panasonic
-As many of the Chinese cells as test good
=all together would give 7kWh
Charge to 4.0 or 4.1V, use only 1.8kWh (25%) at 2kW (<0.3C) and the pack could last 10 years. Never worry about it again unless you move.

 

[flippy]

fully charge the cheap cells and leave them at 4.2V for at least 48 hours and then check voltage. self discharge is a huge issue for old and crappy cells and can kill your capacity and battery pack balance. a bms with high active balancing current (resistive will NOT work in your case!) is mandatory in your case however you make the battery.

 

[MK2R]

Okay, i shall fully charge them and think some more and test.

 

[MK2R]

Okay i am confused... when i charged the batteries and tested them in the lii-500 all batteries are "warm", i mean they are not room temperature but they get warmer (at max settings) even the samsung got warm

i took 3 of the worse FST cells in a 3S config, i am powering a 10W (12V) led lamp with them right now, and they don't get warm at all. correct me if i am wrong but the 3s pack is operating at 800mA right now, correct? (0.8*12 = 10w). why are they not getting warm but they get warm in the lii-500?

I would estimate that they are about room temperature. been running it for 1 hour now, fully charged the 3 batteries b4 i spotwelded them, they should contain ~1600*3 (17.7wh). should be able to power the led for 1hour 45 minutes if they are drawn to the absolute bottom voltage. i will stop the test soon thou.

i spun off a bit there...
but why is there a heat difference?

 

[flippy]

but why is there a heat difference?

there isnt. the difference comes from the higher charging current and the heat the charger produces.

 

[MK2R]

but why is there a heat difference?

there isnt. the difference comes from the higher charging current and the heat the charger produces.

Okay! so it is just the charger itself (combined with actual heating in the batteries) that makes the batteries feel warm in the charger?

 

[eMark]

a bms with high active balancing current (resistive will NOT work in your case!) is mandatory in your case however you make the battery.

SO TRUE! An ebike pack whether new cells, mixed cells of similar rating or salvaged cells is discharged to a cut-off voltage usually within a couple hours or less. Then again recharged usually within 24-48 hours to at least 3.97v before again being discharged within a few hours from being charged (somewhere between 3.97v to 4.2v depending on usage and maintenance application).

Even if a few cells have noticeable self-discharge that's not as much of a concern because the self-discharge is not that noticeable over 24-48 hrs. However, a power wall serves a different application ... thus the rationale for power wall parallel groups having more cells than a typical ebike battery (10S5P, 12S6P, 14S7P).

Thus the rationale for a power wall BMS with high active balancing ... same thinking as why cars have a high active BMS ... namely that of an alternator

 

[flippy]

Okay! so it is just the charger itself (combined with actual heating in the batteries) that makes the batteries feel warm in the charger?

basically yes.

same thinking as why cars have a high active BMS ... namely an alternator

a normal battery in a car does not have any form of BMS and an alternator has nothing to do with protecting the battery so i dont follow your analogy in the slightest.

 

[eMark]

a normal battery in a car does not have any form of BMS and an alternator has nothing to do with protecting the battery so i dont follow your analogy in the slightest.

I can't believe that you actually believe "a car does not have any form of a BMS" (Battery Maintenance System) OR that "an alternator has nothing to do with protecting the [car] battery".

To say a car battery does not have an active Battery Maintenance System (BMS) is not like you. Here I was posting a reply (in layman terms) to SUPPORT your previous post for why it's a good idea for MKR2's power wall to have provision for an active charge BMS instead of a more passive resistance discharge BMS (like that of most DIY ebike battery builds).

Your input in this ES Battery Forum has always been most valuable and evidence that you are VERY knowledgeable, but your two replies in parenthesis are contrary. Why you actually believe "a car does not have any form of a BMS" (Battery Maintenance System) OR that "an alternator has nothing to do with protecting the battery" is perplexing. It's not like you.

It would've been more helpful to MK2R if you might have followed up my supportive post by recommending one possible "active charge BMS" DIY build or commercially available for his power wall instead of ripping on my post that wasn't helpful or even necessary.

 

[flippy]

instead of making it somehow a personal attack or whatever i simply said i dont get your analogy and why i didnt. thats it. if i am missing something in your analogy then please explain it.

and the only reason he needs an active balancing setup is because of the poor quality of the cells and you need several amps of current to keep the battery balanced. resistive would not work as the currents are too low or simply too wasteful.

 

[eMark]

... the only reason he needs an active balancing setup is because of the poor quality of the cells and you need several amps of current to keep the battery balanced. resistive would not work as the currents are too low or simply too wasteful.

Whether or not the cells are "poor quality" is possibly more subjective than objective, but for the sake of this discussion let's assume they're "poor quality", but nonetheless still quite usable/viable for the OP's application.

There's probably more than a few DIYers that would have more than a passing interest in what "active balancing setup" is available commercially (or DIY) for such a power wall having "poor quality" cells"(e.g. mixed cells of questionable quality) that are still usable for some power wall applications.

It would be helpful if you would suggest a couple options for an "active balancing setup" for MK2R and other similar power wall applications using "poor quality" cells. It would be well worth the effort to provide some ES links and other links as far as "active balancing" for power wall applications whether with salvaged cells of "poor quality" or possibly even mixed cells of OK quality.

Getting back to OP thread ... AMOUNT OF CELLS? MANY OR FEW ... IMO, whether having more cells than fewer cells of "poor quality" in a reasonable power wall application is not as critical as having the RIGHT KIND of "active balancing setup" for a power wall with "poor quality" cells. Being there is possibly more than one option, depending on the power wall application, then such an "active balancing setup" could be expensive (subjective), especially considering the cells are "poor quality".

What did you have in mind (reasonable cost) for an "active balancing setup" for his power wall of "poor quality" cells assuming they are usable/viable for the OP's power wall application ?

 

[MK2R]

guys don't fight. i appreciate all imputs <3

many of the cells i've tested have been able to hold the charge of 4.2V for about 1 week, is that alright?

I might be able to exclude the "volt leaking" cells and still have a lot left. is that a resonable option or should i get a bms with active balance and use them all?

What to say about my led "experiment".

V start: 12.68V, plug in led drops to 12.58V. i get a 0.1V drop at the load 10W (~800mA).

If i do my calculation correct:

i know the voltage drop, 0.1V
i know the current which is almost 800mA (0.795A at 12,58V)

0.1/0.795 = 0,1257. this is the total IR of all 3 cells correct? so an averge 0.042. or less or more 42mili ohms per cell.

If i add 3 identical cells in a 3S2P configuration, would it be R/2 then? 0.021, 21mili ohms per paralell group
Or well the in basic it would be said in cells in paralell resistance would be (delta R / n cells).

So the more cells you add in paralell, the less heat will be produced in each cell at a set current, as long as the current stay fixed.

Am i getting this concept correctly ?

And as a direct consequence of this, the voltage drop will be also lower at a fixed current?

 

[eMark]

So the more cells you add in paralell, the less heat will be produced in each cell at a set current, as long as the current stay fixed.

Am i getting this concept correctly ?

Even "less heat" with more cells in parallel is likely why flippy calls them "poor quality" cells and possibly even more so with a power wall that isn't associated with "raw performance" ... say like that of a performance mountain ebike battery. Here's one youtube for starters explaining an "Active Cell Balancer" ...

https://www.youtube.com/watch?v=fEVQ_8DUpm4

Just so you know that the BattGo BG-8S is a resistance discharge balancer; whereas your power wall needs an active charge balancer/equalizer. Maybe someone knows of a commercially made or DIYer active charge balancer (if flippy doesn't) that would be a suitable fit for the "poor quality" cells in your power wall and its application.

 

[fatty]

So the more cells you add in paralell, the less heat will be produced in each cell at a set current, as long as the current stay fixed.

Am i getting this concept correctly ?

Yes

And as a direct consequence of this, the voltage drop will be also lower at a fixed current?

Correct

 

[fatty]

and the only reason he needs an active balancing setup is because of the poor quality of the cells and you need several amps of current to keep the battery balanced. resistive would not work as the currents are too low or simply too wasteful.

I think there's two ways to look at this:
1) Poor quality cells will tend to come out of balance, which can be compensated for by active balancing (cost and complexity)
or
2) Poor quality cells are cheap and the OP has a surplus. I wouldn't waste time and money on an active balancing BMS, but rather compensate with more cells in parallel, which will statistically decrease parallel string variance and reduce imbalance. Likewise, more capacity means less charge/discharge, which means less imbalance.

Two means to the same end. Except more capacity also gives.. more capacity...

 

[fatty]

... same thinking as why cars have a high active BMS ... namely that of an alternator

I don't understand this either. ICE cars don't have BMS, alternator or otherwise. EV cars do have BMS, but it's not the alternator.

 

[nicobie]

Don't forget about the safety factor when using iffy cells.

I'd for sure cull the bad ones. Set parameters. Don't risk your house!

 

[MK2R]

understood. i am leaning more towards using the good cells for my powerbank and the FST cells for something less risky, unless i manage to get a lot more (to reduce load per cell).

If i make cell blocks of 6 cells wide, would it make sense to get the cells with the lowest IR at the core and the ones with the highest IR more lateral in the pack? this is just my way of thinking, so i keep the ones that will heat the most (little in general but at least in theory it sounds good) lateral for more contact with surounding air (more cooling).

Is this in theory the right way to at least look at this problem when cells vary a little in IR (let's say 25-35, 25 at the core and 35 at the outermost part).

Or is it at so little value that in practical applications it wouldn't matter?




Speaking of poorer quality cells with higher chanse of creating imbalance over time. when does the imbalance mostly build up? is it during consumption of the stored energy and standby leackage, or when charging up the battery - or both?

*Let's say in theory you connect the pack to some low power solarpanel setup with a mttp module and charge it a little every day (let's say 300-1000w per day depending on weather on a 5kwh pack), and then you manually use a bit then and then to not get the pack to 100% SoC, sometimes days use a lot more so it would take a few days at that low amount of solar power to get it to full (only reason is to not waste solar energy since it can't be charged more than 100%). - maybe not the most well thoughtout idea but roughly something like that at least, hope you get the general idea

 

[flippy]

Even "less heat" with more cells in parallel is likely why flippy calls them "poor quality" cells and possibly even more so with a power wall that isn't associated with "raw performance" ... say like that of a performance mountain ebike battery. Here's one youtube for starters explaining an "Active Cell Balancer" ...
https://www.youtube.com/watch?v=fEVQ_8DUpm4
Just so you know that the BattGo BG-8S is a resistance discharge balancer; whereas your power wall needs an active charge balancer/equalizer. Maybe someone knows of a commercially made or DIYer active charge balancer (if flippy doesn't) that would be a suitable fit for the "poor quality" cells in your power wall and its application.

i believe i already linked to this bms before but here is the link again: https://www.aliexpress.com/item/4000529723243.html

get the either of the 2 big boy versions as it has the current capacity and the high 2A active balancing current needed for this battery to stay in shape.

Speaking of poorer quality cells with higher chanse of creating imbalance over time. when does the imbalance mostly build up? is it during consumption of the stored energy and standby leackage, or when charging up the battery - or both?

the answer is yes. to all. there will be a big imbalance between the cells and lots of different levels of self discharge, IR and capacity. so every block you make will suffer from different values of each metric so a constant 2A balance current will be needed to keep the pack balanced. as it will take time for each high block to transfer energy to the lower block. dont forget that 2A is about 8W of balancing power. that is about 1 cell worth of energy per hour it can move.

 

[eMark]

... same thinking as why cars have a high active BMS ... namely that of an alternator

I don't understand this either. ICE cars don't have BMS, alternator or otherwise. EV cars do have BMS, but it's not the alternator.

When flippy said active charge vs active discharge i assumed he was referring to more than what is the following stand alone active charge equalizer.

i believe i already linked to this bms before but here is the link again: https://www.aliexpress.com/item/4000529723243.html

get the either of the 2 big boy versions as it has the current capacity and the high 2A active balancing current needed for this battery to stay in shape.

To say it is an active "charge" stand alone balancer is a little misleading. When flippy says the cells are "poor quality" ... thus my car alternator analogy was that MK2R also needs an alternate low power source such as a small DIY wind turbine or solar panel power supply to feed his powerbank.

Let's say in theory you connect the pack to some low power solarpanel setup with a mttp module and charge it a little every day (let's say 300-1000w per day depending on weather on a 5kwh pack), and then you manually use a bit then and then to not get the pack to 100% SoC, sometimes days use a lot more so it would take a few days at that low amount of solar power to get it to full (only reason is to not waste solar energy since it can't be charged more than 100%). - maybe not the most well thought out idea but roughly something like that at least, hope you get the general idea.

This was the same logic i had i mind with my car alternator analogy to get the most longevity out of MK2Rs powerbank with "poor quality" cells.

In other words the https://www.aliexpress.com/item/4000529723243.html active balancer could use some additional help :thumb:

 

[flippy]

To say it is an active "charge" stand alone balancer is a little misleading.

i never claimed it was a stand alone active balancer. those are dumb imho and serve no puropse in 99% of cases. stand alone units are only there as a fix for a mistake usually made earlier in the pack build by choosing the wrong bms for example.

 

[eMark]

To say it is an active "charge" stand alone balancer is a little misleading.

i never claimed it was a stand alone active balancer. those are dumb imho and serve no puropse in 99% of cases. stand alone units are only there as a fix for a mistake usually made earlier in the pack build by choosing the wrong bms for example.

My definition of "stand alone" is that ... https://pt.aliexpress.com/item/4000529723243.html ... doesn't provide true "active charge" like that of a dc balance charger or analogy of a traditional 12v dc voltage regulator, alternator, etc (car BMS). To say the aliexpress is an "active charge" BMS is a little misleading as it's more of a voltage equalizer ... more so than an "active Charger" between the p-groups. Similar comparison is the following equalizer analogy ...

An analogy would be several small inter-connected fish pools in need of maintenance that slowly leak at different rates. The pools that leak the least will provide water to the pools leaking more thus hopefully maintaining all the pools at the same level ... wouldn't consider that an "active charge" compared to the pools being periodically "charged" by a spring to prevent them from all going empty.

... https://pt.aliexpress.com/item/4000529723243.html ... is certainly a welcome improvement over traditional discharge BMSs that for the most part are pathetic being more of a BPS at best than a BMS. That Ali BMS is an active leveling equalizer. By active i take that to mean that this Ali Smart BMS is constantly active :thumb: meaning it doesn't wait until the charged battery pack is close to FULL before balancing the p-groups via resistance discharge drain.

Thank you for providing that Ali Smart BMS link. Eventually so-called "active charge" BMSs will replace the cheaper BMS that in some cases don't even provide a reliable BPS (Battery Protection System).