cheap active cell balancers

[TrotterBob]

The downside is that there's still an imbalance of 0.100v (100mV) after balancing.

How long did you leave the balancer plugged in? The Heltec website says work continues until the voltage difference is .03v or less. Which can take a while since the current drops as the voltage delta between cells drops.

https://heltec-bms.com/project/2s-8s-1-2a-active-equalization-balancer/

Both of the Rc hobby chargers I have do a much faster job of balance charging and are accurate to .02v or better. At least the Revolectrix charger I have is. I've got a Turnigy thats only accurate when it wants to be, which is fairly typical of 4 button chargers AFAIK.

 

[SafeDiscDancing]

... and it doesn't require me to make friends with all my cells.

Point well taken. And I am not saying everyone should get as deep inside the battey as I do, but I am saying there is a benefit if people actually use their brains this way.

We live in an age where people dislike effort... so I understand that this is "my thing I like".

Also... I am very much in tune with the other things with the battery like how warm or cold they are. I find that a cold battery is weak, but an overheated battery is weak too. So there are many variables which I enjoy thinking about.

But this is a type of thinking that is not for everyone.

 

[eMark]

The downside is that there's still an imbalance of 0.100v (100mV) after balancing.

Apologize for photo typo in previous post (attached photo improvement says it all).

How long did you leave the balancer plugged in?

Those little red lights stayed on all the time during the energy transfer and then went out after balancing within 102mV from the cell with the highest SoC to the cell with the lowest SoC. Total time was 67 minutes. The equalization was done when the little red lights went out and didn't come on again after waiting a few minutes.

The Heltec website says work continues until the voltage difference is .03v or less. Which can take a while since the current drops as the voltage delta between cells drops. ... https://heltec-bms.com/project/2s-8s-1-2a-active-equalization-balancer/ ...

That's the way an ABB should work. If a BMS can top balance (resistance discharage) the p-groups within 25-35mV of each other ... so then should an Active [equalizing] Balancer Board do likewise.

I can bottom balance my p-groups within 10mV of each other in 15-30 minutes. After bulk charging to 90% SoC the p-groups are within 30-40mV of each other at the top ... so no need to top balance or balance charge.

Both of the Rc hobby chargers I have do a much faster job of balance charging and are accurate to .02v or better. At least the Revolectrix charger I have is. I've got a Turnigy thats only accurate when it wants to be, which is fairly typical of 4 button chargers AFAIK.

My HTRC Duo C240 uses the same technology as the iCharger for measuring IR ... it gives you a good enuf indication of IR differences between p-groups to know if there's a problem cell(s) in a p-group.

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.

Sounds to me from your reply that you're the kind of friend a high energy pack or any pack is happy to have its BEST friend

+ "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.

First-off if that's the case the DIYer/owner/operator was never a BEST friend to his pack. If you treat your pack with TLC (as i do and as described by Chalo) i've found my p-groups actually come closer together as they near and arrive at 90% SoC capacity (even with bulk charging), after first bottom balancing. At least that's my 30Q 10s3p Vruzend experimental pack experience. Your scenario is one in which the operator was possibly never true blue best friend to his battery pack. That said i'd agree that too many think that adding an Active Balancer Board after some aging or injury ... hoping an ABB may help cure what ails the pack.

That's possibly why the JW ABB only balances within 100mV; whereas, the Heltec ABB balances within 30mV and intended more for those Operators that are a BEST friend to their pack. They know their ABBs limitations and it's best-fit application. They don't wait until the pack is already ailing. When that's the case the best an Operator could hope for is balancing the p-groups within 100mV and even then it may be too late and a waste of money attaching an ABB.

Bottomline: It seems if you're going to use an ABB go with it from the beginning and use the Heltec (30mV instead of 100mV). Be a BEST friend (like Chalo) to your battery pack. If you treat it right it will treat you right (well maybe not if you have 30Q cells that aren't all Grade A)

 

[eMark]

The downside is that there's still an imbalance of 0.100v (100mV) after balancing.

Those JW 1.2A boards are suppose to balance cells (p-groups) within "0.03V" (30mV) of each other, so returned that JW 5s 1.2A active balancer board as it was apparently defective. Ordered the 6s 1.2A Heltec for further experimenting ... https://m.media-amazon.com/images/I/61X5+CD6wxL._AC_UY327_FMwebp_QL65_.jpg.

The Heltec website says work continues until the voltage difference is .03v or less. Which can take a while since the current drops as the voltage delta between cells drops ... https://heltec-bms.com/project/2s-8s-1-2a-active-equalization-balancer/

This tme will run some tests on six 30Q imbalanced cells (some having high IR) as well as a few old Lipos (some with high IR). Should have results to share by end of month. Testing plan is to only use the Heltec 6s 1.2A active [equalizer] balancer board when cells are at rest--not during charging or discharging ... to see how long it takes to balance resting cells. Resting cells that are stablized with no further voltage decay or bounce back.

 

[Kimchi]

I might have missed something, how are active cell balancer to be used?

I have the same balancer and monitor as this guy.
https://youtu.be/WKGk5Of4sik
But I also have a charger attached as well.



Balancer is most effective when charger is near full and the charge amp will drop towards one or below.
After it's balance or close to balance at 57.6v. I then charge it to 58.5v to get my cells to 4.17v.

Are you guys doing it separately?

 

[john61ct]

Many resistance-type top balancing devices have too slow a current balance, do not have adjustable balance start/stop setpoints, or simply suck otherwise.

An active balancer can be used INSTEAD.

An additional advantage is you can balance independently of the charge cycle at whatever SoC%/voltage point you desire, rather than stressing your cells at the top.

Using more than one device for balancing is not advised, nor is balancing at more than one SoC%/voltage point.

Obviously you still need to be able to MONITOR your per-cell/group voltages, the hobby/charger or BMS can still be used for that, even if you do not use it for actually balancing.

 

[eMark]

I might have missed something, how are active cell balancer to be used?

I have the same balancer and monitor as this guy.
https://youtu.be/WKGk5Of4sik
But I also have a charger attached as well.

The usefulness of that 5A Active Balancer (https://youtu.be/WKGk5Of4sik) and others like it are IMO more for a low drain ebike battery of questionable cells and low drain powerbanks of salvaged cells (even different brands) that have been tested and sorted with approximately the same capacity, internal resistance and MCD amp rating. So as to makeup parallel groups that are as similar as possible to each other (as in that youtube).

In that youtube the best Active 5amp Balancing of that 14s10p battery of salvaged cells after several hours was 0.047V. However, after disconnecting that 5amp Active Balancer the 14 parallel groups returned to an imbalance of 0.134V. That's typical of a battery pack made up of salvaged cells varying in brand, IR, MCD amp rating and possibly chemistry (e.g. NCA, NMC). Cells that were free or at most no more than 50￠ each.

The point being that the best p-group balancing possible for that 14s10p battery (at rest) was 134mV. That's assuming the capacites, IR, etc are as similar as possible so say out of 200 salvaged cells that were tested 140 cells were usable. 134mV is not bad assuming the application is more mild-mannered amp drain rather than power/speed amp requirement.

A p-group variance of 150mV may be no problem with a low amperage powerbank application using that $60 5A Active Balancer Board shown in the youtube you posted ...

You could purchase a better Active Balancer that will balance the p-groups within 5mV, but WHY if the best cell grouping (p-group variance) of salvaged cells is 134mV. For all practical purposes it's probably a waste of money to purchase this Active Balancer ... https://www.amazon.com/Equalizer-Balancer-Equalization-Precision-Capacitor/dp/B09ZDSDQ7C/ref=pd_lpo_2?pd_rd_i=B09ZDSDQ7C&psc=1 ... https://www.aliexpress.us/item/3256802957321281.html?_randl_currency=USD&src=google&src=google&albch=shopping&acnt=708-803-3821&slnk=&plac=&mtctp=&albbt=Google_7_shopping&albagn=888888&isSmbAutoCall=false&needSmbHouyi=false&albcp=9594035441&albag=102695258807&trgt=296904913880&crea=en3256802957321281&netw=u&device=c&albpg=296904913880&albpd=en3256802957321281&gclid=CjwKCAjwv4SaBhBPEiwA9YzZvABeMIDJrk93f2pqby6RYL3kTVmQwBbP3mV1pmkCa-Tc8mV65VNz6BoCuhEQAvD_BwE&gclsrc=aw.ds&aff_fcid=c00d4e16f2524d1bb39e401a22362f05-1665272200883-00870-UneMJZVf&aff_fsk=UneMJZVf&aff_platform=aaf&sk=UneMJZVf&aff_trace_key=c00d4e16f2524d1bb39e401a22362f05-1665272200883-00870-UneMJZVf&terminal_id=3131f10fdae84e7bb972df5102b44d3f&afSmartRedirect=y&gatewayAdapt=glo2usa&_randl_shipto=US ...

If an ebike DIY build of salvaged cells are almost identical you may want to invest in a more expensive Active Balancer with digital readout of the parallel groups that can be balanced within a few millivolts of each other whether after discharging before charging or after charging before discharging (as needed). However, if the salvaged cell variance (Brand, Capacity, IR, MCD amp rating and Cell Chemistry) is such that balancing even within 50mV is not possible with the other 5A Active Balancer than even the best Active (equalizing) Balancer will only be able to equalize the resting p-group voltages after hours and hours, if even then.

Are you guys doing it separately?

With a DIY ebike battery build (e.g. 13s10p) if possible buy overstocked unused cells 3-5 years old for $2 (same name brand, same manufacturing run, same capacity, same IR, same MCD rating, same chemistry). That was the case with my $2 cells ... https://batteryhookup.com/products/30-100-new-samsung-inr18650-33g-3150mah-18650-cells. After the first dozen c/d cycles the parallel groups are still within 5mV of each other both after discharge and after charge ... so no need to balance the parallel groups (yet) ...

https://batteryhookup.com/products/30-100-new-samsung-inr18650-33g-3150mah-18650-cells

So no need to manually bottom balance after discharging or manually top balance after charging. Even when that becomes necessary i prefer manually balancing because it is considerably faster than using an Active Balancer Board. IMO an Active Balancer Board is OK for a low drain powerbank or ebike battery of salvaged cells, but not an amp thirsty ebike battery unless it's like 10p with more thaan enuf p-group amperage to handle the Controller rated amp drain for a couple minutes.

 

[lnanek]

Just to confirm, it's OK to leave these active cell equalizers attached and in charge of cell voltage during both 1) charge and 2) discharge?

During 1) charge, doesn't the BMS get confused and see the extra voltage on the balance wire from the equalizer trying to charge a low p group from a higher one and potentially A) cut off the entire charge due to that p group looking high or B) fight against the equalizer and burn off the extra voltage the equalizer is trying to transfer with a resistor?

During 2) discharge, it will use up capacity since no transfer is 100%, but it should also help prevent the lowest p group from reaching LVC as early as it would otherwise, right? A typical BMS can just top balance while charging by burning off current with a resistor, and cut off discharge when one p group reaches LVC, but the flyback capacitor or inverter style cell equalizers should be able to add some support to the weakest p group, right?

I've got an aging 16s20p pack of Panasonic NCR18650B 3400mAh 4.9A cells. The Daly BMS has cut off charge due to one p group being at 3.4v and the rest 3.5v. I was planning to stick a 1S charger on just that group and then try to charge again, but could also put in something like this inverter cell equalizer while I'm in there:
https://www.amazon.com/dp/B0B3D5JCRC

I see Daly has cell equalizers as well. Would maintaining the same brand be less likely to cause interference since things like HVC and LVC will match? Or does the BMS need to be replaced with one that has a balance function that can be disabled?

I'm aware the battery needs to be atomized and the cells re-rated, bad ones discarded, and a new battery made with matching cells eventually. Sure would be nice to avoid that much work for a little longer, though

 

[john61ct]

Balancer is most effective when charger is near full

That is only for top balancing.

There are many other strategies that are just as, or lots more effective

and more conducive to longevity.

 

[john61ct]

Just to confirm, it's OK to leave these active cell equalizers attached and in charge of cell voltage during both 1) charge and 2) discharge?

It might be "OK" as in not creating harm, depends on the balancer and its methodology.

The optimal path is only balance when needed, as a periodic maintenance routine.

A newish quality pack of well matched cells,

treated properly avoiding the SoC/voltage shoulders

may not need balancing for years.

As it wears, maybe once every six months, or every 100 cycles, whatever.

It is certainly a waste to balance at more than one specific point in the SoC/voltage curve, pick a strategy and stick to it.

Whatever device you use, best to ensure it is easily removed swapped out without ripping the pack apart, one good pack may go through several in its useful lifespan.

If you find it "necessary" to frequently rebalance a worn pack, it's likely time to retire / replace it with a better one.

 

[eMark]

I was planning to stick a 1S charger on just that group and then try to charge again.

Try that first being that your 16s20p may be nearing EoL. Charge to FULL assumng BMS will again allow. *Then leave the charger plugged in for 8-10 hours after green light comes on. Take your bike for a spin to see if readout stays on 5 bars longer than before. If so that would be an indication that the sixteen 20p-groups are more equalized than before. Hopefully just that one p-group is the only problem causing the BMS to shut-down charging. You may have to switch from a 5amp charger to a lower amp charger if the problem persists.

You may have to occasionally bottom balance that one p-group to keep it inline with the others so BMS will allow a 90% or 95% charge. A 100% charge shouldn't pose a problem as long as you put your 16s20p to use within an hour or two after charging to 95% (or even 100% charge).

Occasionally doing the above *routine (as needed) could prolong the life of your 16s20p for another 100 cycles. You want to do whatever it takes to extend its life as long as possible ... which could also mean more casual riding and less raw performance riding.

 

[john61ct]

Balancing by definition is done to the pack as a whole.

There is no such thing as balancing one cell/group.

You either use bottom balancing and disable any device from balancing at the top, or v/v

balancing at more than one spot on the SoC%/voltage curve is not just a pointless waste of time but counter productive, harmful to longevity.

EoL is supposed to be based on declining Ah capacity, long before gross wear symptoms begin to manifest.

After that point you are literally playing with fire.

 

[lnanek]

Found a neat paper confirming OK, too:

The flying capacitor method has the advantages of simple structure, low energy loss and can be balanced in the charging and discharging process. It is a widely used balancing method at present

.https://academic.oup.com/ijlct/article/16/1/199/5880318#

Some research is even toward some allowing mixing batteries:

Different from the buck-boost converter and switched capacitor, the tapped inductor has advantages in step-free adjustment and is suitable for balancing hybrid source packages composed of batteries and aged batteries.

https://www.frontiersin.org/articles/10.3389/fenrg.2022.843453/full

Although that paper is hilarious in that it claims generating waste heat is a feature, not a bug :lol:

 

[john61ct]

great resources, thank you

 

[lnanek]

Tried out the Sunkko. It's quite slow so far. After two hours of balancing the cell that moved the most only moved 0.044V. So at that rate, will be 5 hours to fix a .2V difference.

Spreadsheet:
https://docs.google.com/spreadsheets/d/1XT4xQ6P1ahd7AsG5Jk8rGRP9Ll9OYpAuspREU4qa5nM/edit?usp=drivesdk

Photos:



That said, the manual says use the thick balance wires that come in the box and solder both ends without any connectors for best performance. I'm reusing thin tiny balance cables, adding a tiny hobby cable in-between via a connector, and then using their easy screw on connector to the board too. Heat sinks aren't hot at all, so no worries about temperature if I could get more current through it.

Also haven't tried charging at the same time, which people say help the flying capacitor cell balancers.

Wonder which type is faster? Flying capacitor or this inverter transformer thing.

 

[john61ct]

Balance CURRENT is the key factor for speed.

Speed only matters when top balancing since sitting at high SoC for long shortens lifespan.

If you have an "active" balancer you can do so at any ine voltage point so around storage voltage no harm done sitting there for days.

The other definition of active, the balance current rate stays constant evan as the delta approaches the balance-stop setpoint.

This is very desirable but also rare, usually the flow slows dramatically, so you cannot extrapolate finish time linearly.

 

[lnanek]

All very true for at home balancing, definitely. :thumb:

For balancing during discharge, though, I'd hoped fast equalization time (high equalization current), if left on during discharge, might be enough to fight voltage sag on the weakest p group (highest internal resistance p group) and keep the battery running longer before the BMS LVC cuts it off. That would allow higher utilization because you could exhaust more of the healthy cells before the faulty cells sagging too far convince the BMS to shutdown.

So far the strongest equalization current I've seen using an amp meter on a balance cable is .37A. My controller max battery amps is 30A. Divided by 16 p groups means 1.875A draw. So feels like the weakest p group could technically be 20% worse and still be propped up by the equalizer. Then possibly more if the equalization current could be improved, like thicker balance wires with less connectors.

It's true the current is affected by the voltage difference too, though, and I don't know what the different voltage is across the p groups during discharge. Their IR varies, though, so I assume they do sag differently to a degree, even if they start out top balanced to the same voltage when not under load.

I've heard the flying capacitor equalizers just cycle through every possibility, but the specs on this device say the first 4 cells power a microchip that determines the balancing action. So hopefully it would be smart enough to target its two inverter transformers (on the more expensive boards, the cheaper models have 1) at the two weakest p groups consistently.

 

[john61ct]

There is absolutely no point in balancing during discharge, waste of time & money, adding complexity reduces reliability.

Balancing never increases pack capacity itself, and certainly cannot increase power or reduce sag.

The weakest cell/group ALWAYS determines overall pack performance. This is why used / salvage cells are usually a waste, not just poor performance per cell but poor matching. Even brand new cells may not be well matched.

Ideally both HVC and LVC are based on per cell/group voltage.

If LVC is pack voltage based, then you want to do bottom balancing not top, and ensure your charge termination is based on per cell/group voltage.

If LVC is per cell/group voltage, then HVC can be pack based, but stay WELL below max spec voltage, and top balance regularly as needed.

With a pack in good shape maybe once every 20 cycles.

If you need heroic balancing frequently, you are past EoL and need to replace all your cells at once.

 

[Chalo]

There is absolutely no point in balancing during discharge, waste of time & money, adding complexity reduces reliability.

Balancing never increases pack capacity itself, and certainly cannot increase power or reduce sag.

Cell balancing is a benefit whenever there's an imbalance.

If you have a 1A balancing current to the weakest cell group over the course of a 1C discharge, then you've added 1Ah to the effective capacity of that group. If that's the group that limited the whole pack's capacity, then you have added 1Ah to the pack. But it could be even more, because balancing continues while the pack is resting and not only when it's discharging.

While it's true that a pack with such a serious imbalance is not what you want, active balancing can make it useable when otherwise it would no longer withstand regular use. E-bikes and other light EVs offer a second life for lots of resource-intensive batteries that would otherwise go to waste. Active balancing not only helps to redeem those sunk resources, but also increases reliability and safety in the process.

I haven't yet used active balancers to reconcile mismatched cells, but I think that's a great application for them. They should be a first resort for folks who repurpose laptop cells or other cells with different use histories.

 

[lnanek]

To solve the discharge mystery, guess I should test ride:

• Bottom balance charge until some individual p group HVC triggers, ride until pack LVC triggers
• Top balance charge, ride until some individual p group LVC triggers
• Top balance charge, turn on cell equalizer, ride until some individual p group LVC triggers

Then see what provides the most run time. Unfortunately it's a 60Ah battery and takes my dinky bike two and a half hours to run dry. So not a quick experiment...