cheap active cell balancers

[john61ct]

Yes LTO is fantastic, but very low energy density.

The li-ion and lipo cells only reason to consider is the very high energy density. Very fire risky and short lifespan.

LFP density is in between wrt density, very fire safe, 10x the lifespan all things being equal.

 

[TrotterBob]

If you need to utilise more capacity

(really? why? poor planning?)

If all you're going to do is make condescending remarks we have nothing to discuss.

 

[TrotterBob]

Am I looking for the basic same layout for the same circuitry design?

Not sure what you're asking here. The 5A balancers look different. How is your pack assembled? With solder or spot welds or hardware? And the thing to keep in mind that longer wires will add a bit of extra resistance vs the shorter ones on a balance lead. I've noticed that adding just a bit more torque to the bolts on these 15a LFP cells decreased the internal resistance showing on my charger.

 

[eMark]

https://forum.esk8.news/t/testing-active-balancer-results/4580

if im going down to 2.5v on the lowest group , the highest might be 2.8v. which gives 0.3v difference the bigger the difference the more amp are used to balance, i think its 1amp on these boards.

you wouldnt be able to balance a pack at 3.5v where theres a big capacity bubble,there would be no accuracy
even 4.1v has more capacity than 2.7v

The only difference in your JW 1.5A Active Balancing Board (ABB) from his in the following test chart is his JW 1.5A ABB is rated at 2.8v and yours at 2.4v. With his 2.8v ABB even a p-group with an imbalance of -0.30v (11s-300mV) can be balanced within 50mV. Apparently an imbalance of 60mV (7s) is his allowable benchmark for what he considers an **acceptable** ABB variance instead of 20-30mv variance being the [striving] norm of a packs' BMS feasible {but rarely achievable] **acceptability**.

So since when does an Active Balancer Board redefine what's been an acceptable p-group variance (25-30mV) by increasing p-group variance to 60mV as acceptable? Is that just a "CON" to make one believe that a "cheap" ABB is a worthwhile investment? Or was a BMS rating of 25-30mV an unrealistic goal that was rarely, if ever, achievable?

There seem to be other inconsistencies in the following chart that call into question whether a "cheap" ABB is really all that helpful when it comes to any regeneration of a packs' p-groups. The same p-groups (5s & 6S) are still the weakest link being the achilles heel with a "cheap" Active Balancing Board.

The point being whether there's a significant delay [before pack failure] when 5s & 6S charged p-group imbalance variance is still 100mV (down from 160mV) and discharged p-group imbalance variance is still 390mV (down from 540mV) ???

The standard BMS norm of variance is 25 to 30mV; whereas according to this chart a variance of 50-60mV using a "cheap" Active Balancer Board is now considered **acceptable*. At least with a 3s1p or 6s1p Lipo pack or a 10s3p or 17s4p Lion pack can we agree that an ABB shouldn't murder a pack, even if it's usefulness is debatable. Both a BMS and an ABB serve a useful purpose as long as we know the limitations.

So, is 50 to 60mV the new norm for **acceptable** variance between 18650 p-groups being that a BMS variance of 25 to 30mV is rarely if ever achievable ??? The bottomline being that at least one or more cells in p-group 5s & 6s need to be replaced sooner than later ... Right ???

 

[john61ct]

Absolutely, no practical reason to narrow the delta any further.

But a better device is adjustable, leaves that decision up to you.

These are only needed with poorer quality cells, or past EoL spec secondhand

where achieving narrow deltas is much harder

using resistance burning, low balancing rates, can literally take days.

And if done at the top of the curve, actually causing accelerated wear, reducing lifespan, causing greater imbalances later.

With good quality healthy cells being cycled sensibly, rebalancing should take at most an hour or two and only be needed a few times a year.

Again, balancing cells is really not important in the first place, unless your protective circuits are functioning off pack-level sensing, either LVC or HVC.

Certainly nothing to do with "regenerating" anything.

 

[eMark]

Bottom balancing so that p-groups are within 10mV of each other is worthwhile as a pack ages with increased c/d cycles. A BMS will never be able to top balance an ageing 18650 pack -- whether bulk charged to 90% or FULL.

A "cheap" Active Balancing Board still won't cure or regenerate a 12s pack with problematic cells like p-groups 5s & 6s in the above chart. At least one or more unhealthy cells in p-groups 5S & 6S need to be replaced sooner than later.

 

[JackFlorey]

A BMS will never be able to top balance an ageing 18650 pack -- whether bulk charged to 90% or FULL.

You can always balance any pack with enough balancing power. But you end up doing a lot of work (=inefficiency) once you get major imbalances.

 

[goatman]

time to do a test and post results of mah/0.1v to see where its most accurate to balance a pack

 

[eMark]

You can always balance any pack with enough balancing power.

But that's no indication that 56 individual cells in say a 14s4p pack are within 50mV of each other even if the parallel groups appear healthy (within 50mV of each other).

The downside of a spot-welded pack is that the parallel group voltages do not represent the actual health of each cell in the parallel groups. In the following chart the actual health of the 2s3p cell at 3.43v would be hidden in a spot-welded pack as only p-group voltage can be easily checked after relaxing voltage rest. Individual cell health can't be easily checked in a spot-welded pack.

To determine which cell in a suspect p-group is bad you need to check the resting voltage of each cell in the p-group instead of waiting until the hidden bad cell (spot-welded pack scenario) is so bad that the BMS shuts down the entire pack. I don't use a BMS as i bottom balance the p-groups within 10mV of each other before bulk charging at 1.5A rate.

The following individual cell voltages are after resting the pack for 14 days (1/1/21 to 1/15/21). Nine cells were replaced with new 30Q6 KH1T cells. The 141 cell at 3.73v wasn't replaced. After 125 more cycles in 2021 (250 total), 4 more 141 cells need to be replaced (see other 141/K chart below this one) including the 3.73v cell) (141 in bold red) ... ...

But you end up doing a lot of work (=inefficiency) once you get major imbalances.

The worst cell in the above 30Q 10s3p pack is 2s3p at only 3.43v needing to be replaced before it demotes the entire p-group and finally triggering the BMS to shut-down the pack.

So far after 125 cycles all of the 9 new 30Q6 KH1T were OK after being tested this month. However, 4 more of the 141 cells (after another 125 cycles in 2021) need to be replaced (red) ...

141--141--141--141--141.......141--141--141--141--141
.K.....K.....K....K....141........K.....K.....K.....K.....K
141--141--141--141--141.......141--141--141--141--141

Thus the downside of a spotwelded pack; whereas APL's Pressure Contact Battery box ... https://endless-sphere.com/forums/viewtopic.php?f=14&t=113512 ... as well as other non-spotwelded designs permit individual cell testing after so many c/d cycles thus preventing a worse case scenario. This feature testing individual cells is crucial with Grade B cells, salvaged cells and high energy dense cells with 30Q apparently more prone to self-discharge.

Unlike goatman i've had little success in atempting to **regenerate** the nine cells replaced in January of 2021 or 4 more 141 cells that will need to be replaced before March.

Bottomline: Using my AstroFlight Blinky Balancer or my "cheap" 5s Active Balancer Board (IMO) doesn't **regenerate** an unhealthy cell it just prolongs its inevitable demise.

 

[Chalo]

You can always balance any pack with enough balancing power.

But that's no indication that 56 individual cells in say a 14s4p pack are within 50mV of each other even if the parallel groups appear healthy (within 50mV of each other).

Cells connected in parallel are forcibly held to the same voltage. There could be a problem with a single cell's self-discharge being passed along to its parallel neighbors, but they will remain in balance with each other (just not with the other cell groups in the pack).

 

[JackFlorey]

But that's no indication that 56 individual cells in say a 14s4p pack are within 50mV of each other even if the parallel groups appear healthy (within 50mV of each other).

In any competently wired pack, the cells are within 50mv of each other in any paralleled group. The wiring forces them to be so.

If your argument is that a weak (low capacity) cell can be hiding within a parallel group, then I definitely agree there. In that case the weak cell will be supported by the stronger cells near it, and avoid overcharge or overdischarge. In a large battery (say, 10s 10p) the hope is that the weak cells will be randomly distributed such that the average capacity of each parallel group is close to each other, thus requiring less balancing.

To determine which cell in a suspect p-group is bad you need to check the resting voltage of each cell in the p-group

If you are talking about before assembly - definitely. Getting known-good and well matched cells will give you the most reliable and easily balanced pack possible.

instead of waiting until the hidden bad cell (spot-welded pack scenario) is so bad that the BMS shuts down the entire pack.

That doesn't happen. One of those 10P cells might be weak, but it will be exactly the same terminal voltage as all the others, since they are connected together with very low resistance connections.

If a lot of those cells are weak, of course, that cell group will go low-voltage and shut down the entire battery. But if in our hypothetical 10s10p battery, one cell in each parallel group is weak, the entire battery will still see good balance.

 

[john61ct]

Problem is, if poor SoH cells exist in a high-cell-count pack, they will **not** be evenly distributed.

Overall pack life with suboptimal cells will be a fraction of the pack made from new, top-notch Grade A cells, and performance suffering all the way to EoL.

Yes, weldless assembly methods nirvana would greatly help atomizing / rebuilding to keep on using scrapped cells of varying SoH

but personally I think not worth the far higher costs and uncertainty, especially if you count your time & trouble.

Buy top-notch cells, spot-weld into packs, all uniform quality (not 30Q) and treat them properly in order to get 700-1000 cycles (also depending on use case, C-rates)

then at 70% SoH sell it to some sucker that wants to try to fool around with your scrapped cells, fantasies of "regeneration".

 

[goatman]

all you have to do is look at test data John to see its not a "fantasy" even better yet why dont you get some equipment and do some of your own testing to share with the rest of us to back up your words :wink:

 

[eMark]

Cells connected in parallel are forcibly held to the same voltage. There could be a problem with a single cell's self-discharge being passed along to its parallel neighbors, but they will remain in balance with each other (just not with the other cell groups in the pack).

Yes and No.

The rationale among spot-welders and their packs seems to be that the supposed health of the p-groups is a good enuf indication of the packs overall health, but is it?. For example goatman's 17s4p pack of salvaged 30Q cells were all paralleled and at rest from May to December when we agreed to run some comparitive charge and discharge tests on his 17s4p plasti-dipped pack and my 10s3p Vruzend pack.

After discharging the 17s4p to 51.0v (?avg cell voltage approx 3.00v?) the pack was rested for 2 hrs and then disassembled to record individual cell voltages. It's possible that all 17 of the parallel groups were within 0.005v of each other before a 2 hr rest using his **regenerative** technique on his salvaged 30Q cells with votages all over the place when first received several months ago.

Are we to just assume that the voltages of each of the 4 cells in each p-group are within 50mV to 75mV of each other because the p-groups are within 50mV to 75 \mV of each other whether a pack with at least 100 c/d cycles on new cells or new build using salvaged cells that's been **regenerated**.

Let's compare the 4 cells in 9s having cell voltages of: 2.908v, 2.908v, 2.838v, 2.842v (variance of 70mV) OR the 4 cells in 13s having cell voltages of: 2.908v, 2.863v, 2.813v, 2.856v (variance of 85mV) OR the 4 cells in 17s having cell voltages of: 2.660v, 2.937v, 2.944v, 2.928v (variance of 284mV). These 3 p-group average voltages are: 11.496v, 11.440v, 11.469v for variance of 56mV with 17s (11.469v) only having a variance of 27mV among these 3 p-groups, yet witing its p-group there's a variance among the 4 cells of 284mV.

I could be wrong (it's happened before), but is it not possible that DIYers that only build spot-welded packs are under the opinion that with 17 p-groups within say 57mV of each other that a cell within one of the 17 p-groups that has a variance of 276mV from the average cell voltage of the other three cells being 2.9363333v (2.936v - 2.660v = 276mv variance).

These 3 p-group average voltages are: 9s at 2.874v, 13s at 2.860v, and 17s at 2.867v. The p-group variance is only 0.0140v (140mV); whereas the variance among the 12 cells in these 3 p-groups is 0.0284v (284mV ... 2.944v to 2.660v). Who really thinks an Active Balance Board is going to revitalize a pack that is already over the hill? Isn't tht part of the problem in that a builder waits until the aging pack is unbalanced and then decides attaching an Active Balance Board will be the answer.

 

[eMark]

That doesn't happen. One of those 10P cells might be weak, but it will be exactly the same terminal voltage as all the others, since they are connected together with very low resistance connections.

That's a false narrative due to spot-welding pack rationale. A mentality that says no need to check individual cell voltages. It's as if you'd have us believe that one never needs to check individual cell voltages as a pack ages. Well that may be true with spot-welders and their packs, but only because it's NO easy matter to measure individual cell voltages. In fact it's impossibe without first removing some of the spot-welds to find a bad cell. Anyone that's had to do this knows it's anything but fun.

Your rationale is based on spot-welded packs that make it practically impossible to measure individual cell voltages. That's the probem and reason why NESE, APL's Pressure Contact box ... https://endless-sphere.com/forums/viewtopic.php?f=14&t=113512 ... and other innovative designs make more sense than a spot-welded pack. IF we could be assured that all the cells in a spot-welded packs were Grade A (e.g. high energy dense cells) than maybe all that's needed is an Active Balance Board and no need to ever tear into a spot-welded pack.

Until we can be assured that all the cells we receive are Grade A ... then a design like NESE, APL's and others (even Vruzend for experimental use) have merit over a spot-welded pack. The exception being that the builder first PreTests all the new cells to see how many are really Grade A and replace bad (Grade B) with Grade A before spot-welding.

 

[JackFlorey]

That's a false narrative due to spot-welding pack rationale. A mentality that says no need to check individual cell voltages.

That's a true fact. Not because I say so, but because Ohm's Law says so.

Like I said, before pack assembly by all means check cell voltages (and ESR's and capacities) so you start out with the best balanced and matched pack you can. Once it's assembled, all the cells in a parallel group will have the same voltage. Again, this is from Georg Ohm, not me. The cells CANNOT change their voltages.

It's as if you'd have us believe that one never needs to check individual cell voltages as a pack ages. Well that may be true with spot-welders and their packs, but only because it's NO easy matter to measure individual cell voltages. In fact it's impossibe without first removing some of the spot-welds to find a bad cell. Anyone that's had to do this knows it's anything but fun.

Yes, you can absolutely disassemble a pack, check voltages and reassemble it. However, it will not be illuminating because all the cells will have been forced to the same voltage by Mr. Ohm's thing. If you then cycle them and re-test them you will learn a lot about each cell and can then rebuild a pack keeping the new measurements in mind. Seems like a lot of work, but if you want to do it, go for it.

Until we can be assured that all the cells we receive are Grade A ... then a design like NESE, APL's and others (even Vruzend for experimental use) have merit over a spot-welded pack.

Provided your mechanical setup supports it, again, go for it. Spot-welds don't need a side load to maintain good connectivity, so for many applications, they work a lot better. (i.e. simpler, lighter, more rugged.) But if your application can stand the extra weight of a compression member, or isn't going to be exposed to shock, then great - go for one of those options.

 

[eMark]

Yes, you can absolutely disassemble a pack, check voltages and reassemble it.

Maybe you've found a way to easily disassemble a spot-welded pack to check individual cell voltages and then just as easily reassemble it. You may be the only one to have ever done so :wink:

However, it will not be illuminating because all the cells will have been forced to the same voltage by Mr. Ohm's thing.

Well, daa if all you're ever going to do is only measure the p-group voltages as if that's sufficient. You give the impression that Mr. Ohm says it isn't necessary to separate the cells for reading their individual cell voltages (after a rest to stablize inherent character) and only then after it's obvious there's a bad cell, but not sure where its location in the p-group. Thus the fun of removing some of the spot-welded bus bars.

Yes they appear to all be forced to the same p-group voltage until they are separated to check individual cell voltages afer a rest to determine which are good and which could be bad. You know very well that all cells in a p-group don't have the same inherent voltage when separated to check individual voltages. The inherent health of the cells in a p-group is only determined after the cells in a p-group are separated, rested and individually tested.

Your post points out what apparently so many spot-welders believe or want to believe is that if a 4p group is at say 3.75v then all cells are still at 3.75v after being separated, rested and tested again as their own inherent voltage is stablized. This faulty logic is misused if you think it support spot-welding as a better alternative than a pack design that is easy to disassemble and reassemble. If all new cells that are sold to DIY builders were EQUAL being Grade A then less need for a pack design that is easy to disassemble and reassemble if there are no bad cells, thus Pre-Testing unneccesasry.

If you then cycle them and re-test them you will learn a lot about each cell and can then rebuild a pack keeping the new measurements in mind. Seems like a lot of work, but if you want to do it, go for it.

Exactly what i've been saying. Thus the importance of PreTesting new cells that are going to be spot-welded. If all new cells were EQUALLY Grade A then no need to Pre-Test before building a spot-welded pack.

 

[JackFlorey]

Maybe you've found a way to easily disassemble a spot-welded pack to check individual cell voltages and then just as easily reassemble it. You may be the only one to have ever done so

Oh, I've done so, back when I didn't have enough money to do anything right. There is no way the process could be described as "easy."

Yes they appear to all be forced to the same p-group voltage until they are separated to check individual cell voltages afer a rest to determine which are good and which could be bad. You know very well that all cells in a p-group don't have the same inherent voltage when separated to check individual voltages.

Yes, they do - because they have been forced to that voltage.

When you separate them you don't see their "inherent voltage" whatever that is. You see their self discharge rate. In a well matched pack this will be almost identical cell to cell. In a poorly matched pack - especially with cells of different ages - you will see different self discharge rates. Li-ions self discharge VERY slowly so it will take a while for you to see any real difference, though.

Exactly what i've been saying. Thus the importance of PreTesting new cells that are going to be spot-welded. If all new cells were EQUALLY Grade A then no need to Pre-Test before building a spot-welded pack.

On that we agree.

 

[john61ct]

Voltage and the balancing issues we started with here really has nothing to do with the topic now under discussion.

The main indicator (measurable symptom) of cell health is capacity. This is never in reality (for these small cylindricals anyway) the maker's nameplate rating, ideally when new within 90% at standard max/starting and min/stopping setpoints, and at a low C-rate, say 0.1C

The secondary attribute is ESIR, but that requires well standardized test conditions, specialised knowledge and gear to get repeatably accurate results.

Grade A cells bought new from a trustworthy supplier (ideally from the same production run at the factory) all should be pretty closely matched in both attributes.

If you are able and willing to use a no-weld system, then you can periodically atomise the pack and re-run all your standardised benchmarks on each individual cell

in order to see how evenly they are wearing out in their journey toward EoL.

With the option of replacing lemons with new cells as they are discovered.

If your pack is spot-welded, you are pretty much locked in to treating that pack as a single monolithic unit, and replacing it as a whole when it reaches EoL.

In neither case are benchmarks going to be useful at the cell group level, only checking individual cells.

Mucking around with scrapped cells already approaching EoL in the first place gives welded packs a much shorter lifespan, and

maybe putting so little value on your time, also tilts in the direction of seeking a weldless solution that works for your use case.

You cannot tell anything useful about SoH from the voltage of individual cells, and although uneven self-discharge may be a factor in sorting/binning a cell collection into go/no-go

it really is a completely separate variable from SoH testing and the wearing-out over time process.

 

[eMark]

I wouldnt recommend that model for any pack larger than 4,000mah. Also do not snooze or leave an active balancer unattended while its connected to a pack.

Such a scenario wouldn't apply to either a "true" Active Balancing Board or a "cheap" balancing board (unless defective causing a short).

The reason being any transfer of energy is from a p-group with a higher SoC to a p-group with a lower SoC. JMO, "Cheap" Active balancer boards ... balance by resistance discharging like a cheap BMS only differene is they function throughout the charge (or discharge) not just at the top of the charge. At least that's my take on the difference between a "cheap" active balancer board and a "true" active balancer board. The "true" definition of energy "equalizing" of the p-groups is only if there's an increase in the SoC of the lowest/lower p-groups and not just a decrease in the energy of a p-group with the higher SoC.

Cells connected in parallel are forcibly held to the same voltage. There could be a problem with a single cell's self-discharge being passed along to its parallel neighbors, but they will remain in balance with each other (just not with the other cell groups in the pack).

That concept is rather misleading because the bad cell (suffering from high self-discharge) is slowly draining the good cells. Kind of like the saying, "One bad apple can spoil the good apples next to it". Unfortunately two good cells in a 3p group can't regenerate one bad cell ... NOT POSSIBLE ... can't turn a grade B cell into a grade A cell.

I have a similar active balancer, but the 4s version. I stuck it on a 3s lipo that was out of balance and it was balancing it, no charger connection needed. This would have been at more like 3.5V/cell. It did the job, but I think it only balances to within 0.1V.

It basically functions the same as the Lipo Astro "Blinky" by resistance discharge to somewhat balance cells, but not completely "equalize". Your Lipo 3s advantage is that it is only 1p. The 6s "Blinky" ... https://www.astroflight.com/106-astro-blinky-battery-balancer-details.html ... at $33.65 costs more than that "cheap" 17s 1.2A active balancer board.

Does anyone really believe a "cheap" 17s 1.2A Balancer Board for $27.35 (Amazon) or $26.50 (Aliexpress) is a true energy transfer of higher SoC cell voltage from a p-group to increase the cell voltage of a lower SoC p-group so as to "equalize" cell voltages ... https://www.aliexpress.com/item/33002903687.html?src=google&memo1=freelisting&src=google&albch=shopping&acnt=708-803-3821&slnk=&plac=&mtctp=&albbt=Google_7_shopping&albagn=888888&isSmbAutoCall=false&needSmbHouyi=false&albcp=9317063908&albag=94962804715&trgt=1482034423097&crea=en33002903687&netw=u&device=c&albpg=1482034423097&albpd=en33002903687&gclid=Cj0KCQiA2sqOBhCGARIsAPuPK0gpi609ojRiiv0i1XafCTsXS5PQKjfFGHh7ZAzRfzcPYCUGqS3n-PkaAg2WEALw_wcB&gclsrc=aw.ds&aff_fcid=91b559dc4c694751a6ad0511ae52429a-1641231982552-00608-UneMJZVf&aff_fsk=UneMJZVf&aff_platform=aaf&sk=UneMJZVf&aff_trace_key=91b559dc4c694751a6ad0511ae52429a-1641231982552-00608-UneMJZVf&terminal_id=3131f10fdae84e7bb972df5102b44d3f

PROOF is needed that a "cheap" Active Balancing Board can transfer energy from a p-group with a higher SoC to a p-group with the lower SoC ... thereby increasing it's previous SoC by approx the same transfer of energy (mV) from the p-group with the higher SoC. goatman needs to run a test with just a 2s portion of his Active Balancer Board or buy just a 2s balancer board.

Then check the voltage "equalizing" between two 18650 cells being tested with an initial SOC variance of at least 0.50v. Say one is 4.10v and the other is 3.60v. Periodically check the cells each hour to see if the 3.60v cell actually increases in SoC energy (mV) by the same amount the 4.10v cell decreases in SoC (mV) energy.

Both cells first must be stabilized so their voltages stay the same (no decay or bounce back) for at least 4 hours before goatman runs his experiment to see if there is actually any increase (transfer) in the SoC of the 3.60v cell.

 

[TrotterBob]

The reason being any transfer of energy is from a p-group with a higher SoC to a p-group with a lower SoC. JMO, "Cheap" Active balancer boards ... balance by resistance discharging like a cheap BMS only differene is they function throughout the charge (or discharge) not just at the top of the charge. At least that's my take on the difference between a "cheap" active balancer board and a "true" active balancer board. The "true" definition of energy "equalizing" of the p-groups is only if there's an increase in the SoC of the lowest/lower p-groups and not just a decrease in the energy of a p-group with the higher SoC.

Both of the active balance boards I have take current from the highest voltage cell and they do charge the cells with the lowest voltage. I have verified this twice by taking a small 6s 1300mah pack that was fully charged, then discharged 2 cells down to 3.8v and connecting the 1.2A balancer.

Resistance balancers should not be labeled as "active cell balancers" as its misleading. Best to read the fine print and watch product reviews.

 

[eMark]

Both of the active balance boards I have take current from the highest voltage cell and they do charge the cells with the lowest voltage. I have verified this twice by taking a small 6s 1300mah pack that was fully charged, then discharged 2 cells down to 3.8v and connecting the 1.2A balancer.

Resistance balancers should not be labeled as "active cell balancers" as its misleading. Best to read the fine print and watch product reviews.

When i arrived home last night there was a small Amazon package by the front door. Was the same brand Active Balancer Board as the OPs 17s 1.2A JW Active Balancer Board. It was only 5s (experimental use) and Wonder of Wonders it does what those marketing gurus say it does ... transfers energy from p-groups with higher SoC to p-groups with a lower SoC. The downside is that there's still an imbalance of 0.100v (100mV) after balancing.

Just finished some pre-testing of the 5s 1.2A JW Active Balancer Board. One test was with three 30Q cells that i replaced because of unacceptable self-discharge. A 3s test using just the black, white, blue and green wire (not the yellow and red wires for 5s). Was interesting watching the SoC increasing to 3.370v and 3.378v from 3.302v and 3.320v; while the 3.634v with the higher SoC decreased to 3.472v (the photo of the upper cell should read 3.634v NOT 3.364v). An imbalance change from 302mV to an imbalance of 102mV. If only it would balance p-groups within 30mV of each other instead of 100mV.

Doubt I will find an applicaton (other than further testing) being i bottom balance my p-groups within 10mV of each other before bulk charging (1.5A rate to 90% capacity). Also replace any weak(bad) cells annually. Now if that JW 1.2A Active Balancer Board could balance p-groups within 30mV of each other and rejuvenate weak(bad) cells of a 30Q battery pack for triple (750) performance c/d cycles it would be a bargain even if triple it's current price.

 

[SafeDiscDancing]

Those are the same active balancers I owned. (they burned out upon reconnection after working as described)

They do work, but I came to the realization they really are only useful for people doing a "Middle Balance" strategy.

Think out the scenarios:

+ "Top Balance" - You want the pack to align at the top. But you cannot LEAVE the pack there or it destroys the cells so you must store the pack at some middle voltage. Now the inequalities between cells become apparent and the active balancer now screws up your top balance. Okay, then you recharge or discharge and the errors cannot be fixed as fast as you charge because these things run at maybe one amp. You have now become a Battery Murderiing System.

+ "Bottom Balance" - Same problem. You get everything perfectly aligned at bottom then start to charge the pack. The inequalities magnify but the active balancer is trying to "correct" what is unequal. It's the same problem. Now when you rapidly drain the pack and get to where the bottom is supposed to be perfectly aligned it is not and you hammer the weak cell with reverse voltage. Another fail.

+ Only in the "Middle Balance" case could this help. But in the middle the cells tend to show almost no differences. So even here it's very suspicious.

Then people who have suffered this nightmare before decide "I will build a massive pack and just live fearfully."

Which is the victory of feelings over logic. This was why I argued to just "know" your pack and care about each cell as an individual.

 

[Chalo]

Which is the victory of feelings over logic. This was why I argued to just "know" your pack and care about each cell as an individual.

That's not a solution for people unless they're in search of another hobby. All of us already have a hobby.

As for me, I like to use active balance boards and charge to 85%. I don't discharge all the way down either. I'm confident this works well enough, and it doesn't require me to make friends with all my cells.

 

[calab]

Use basic active balancers for day to day and check manually occasionally
check manually occasionally without active balancers and balance when the delta is off, if you discharge completely check more often using a few 6s rc voltage checkers and charge the node when it gets off individually then waiting for a long balance charging process the rc chargers do. A little bit of time setting up your battery for what you want doesnt have to take up tons of time if 4s is low put a 1s charge on the 4s location do the same with discharging.