Active energy transfer balancing

[Inwo]

There are many new active balance boards showing up from China. I have tested most of them, if anyone needs an opinion before ordering.

It was requested that I do a more thorough test of the potted 4s board by a boat owner using 4s 60ah LiFePo4 for starting battery.

Getting to the end of test. I won't bother to list all the data points. Most important is how it finishes.

Test started by discharging #2 till dead <2v and charging #3 till near full.
.15 shunts to measure average current in and out.

Less than 2ma drain with switch off. Does not appear to be zero when using shunts.
The most current measured is less that 1.5 amps. However 5 amps has been seen by a user with an overcharged cell. Not me!

This is where it's at after about 3 days:
Cells 1-4

3.278 -142ma
3.239 +124ma
3.323 -68ma
3.292 -1.3ma

Differences between + & – may be test errors or losses. Ie. discharging more than charging.
Also expect high cell #3 to be discharging more than #1.
Who knows. Going the right direction in any case!

I have the same ones in 16s and 20s if someone wants to test and not wait for China.



Shows current and waveform on balance leads only.

 

[Grantmac]

Are they available for Li-ion? Programmable? Can the 20s run a lower cell count?

Thanks,
Grant

 

[Inwo]

Yes, as they are energy transfer only, they are not chemistry dependant.
Works with lto up to 4.2v cells or supercaps.
Many versions available, such as 8s and 12s.
All work with fewer cells and can be overlapped for more. I'm building a harness for 36s. 3 volt modules in series connected 4p.
Using 2 pcs 20s balancers.
First thought was that volt leads would not handle the high current. Test shows <2amps average current..

 

[Grantmac]

That's a very interesting and hooefully robust way to balance cells.
Could they be used during a charge cycle or only after? Link?

 

[Inwo]

Yes, they work 24-7. Anytime there is over 30mv difference. Unless you turn it off.

 

[amberwolf]

So if there was a leaky cell (group) that kept dropping in voltage, this balancer would drain the rest of the cells attempting to keep that group alive?

I suppose that's good that it would "save" the bad group (really, the good cells in that group) from being pulled below "death" level, but bad that it could drain the pack to whatever LVC the balancer has, and then the leaky group would still die.... and potentially the power drain from the BMS as a whole on the rest of the pack could kill the rest of the now-drained cells. :/

I'd guess it would take a long time, but on a small pack, left unattended after a ride already low in voltage....

Something to make sure people monitor their packs, I guess.

 

[fechter]

That's pretty impressive. Thanks for posting the test results. Those will be ideal for those "Frankenpacks" made from mismatched recycled cells. Also cool that it's chemistry independent.

So when all the cells are within range and no longer actively balancing the current was about 2mA?

 

[fourbanger]

Inwo, thank you for offering your expertise on this matter.

Before this thread grows too old, I'd like to ask your opinion of the super cheap active balance boards such as this.



Are they any good at all or not even worth the fiberglass they're printed on? They claim to only start balancing when one cell is out by 0.1V which is.. not very accurate. Still, I'm curious as to how they perform If they do at all.

Many thanks!

 

[Punx0r]

If they are chemistry agnostic does that mean there is now LVC?

 

[fourbanger]

If by now you mean no then I think yes.

Wow. That was a weird sentence.

 

[Punx0r]

Yep, typo, was suposed to say "no LVC".

That's a bit of a shame as it does mean it's fully capable of killing your battery rather than protecting it. DIP switches to select LVC in 0.5V increments would be a good addition.

It's great to finally see active balancing hitting the hobbyist end of the market.

 

[Alan B]

One problem with these balancers is if the cells are at unequal temperatures the balancer will unbalance the pack according to the cell voltages which are a function of the cell group temperature. So if a pack has sun shining on one side it may make the balance worse. State of charge isn't represented precisely by voltage alone. Also in mixed packs the variations in SOC vs voltage may cause the same effect. So energy gets shuffled around (and some of it lost) for no useful purpose. For these to work well the cells need to be well matched and at the same temperature.

 

[spinningmagnets]

Bump, to mark this thread. I like this idea. Tesla, Chevy Volt, and Nissan Leaf charge to 4.05V per cell. Doing that helps the cells last longer, and also provides a safety margin.

That leaves a need for some way to balance the cells. Equalizing seems like a reliable way that has no failure mode where the pack dies, like a janky BMS...

 

[eikido]

john61ct gave me some nice tips about those active balancers and i'm very eager to get one.
Thank you john61ct!!


I've been investigating a little bit and found out that those cheaper ones are not really good since they transfer energy between each cells, so cell 1 cannot transfer to cell 6. Yes after a while, all will be balanced. Also, you cannot control the "start to balance" threshold which is at 0.1v:
https://www.aliexpress.com/item/33002903687.html?spm=2114.12010612.8148356.1.6b381313pKtjV2

As i understand it, i should get something like this for better control.
https://www.aliexpress.com/item/32945565819.html?spm=2114.12010612.8148356.7.6b381313pKtjV2
A bluetooth active balancer with possibility to monitor each cell and set different settings such as thresholds.

Anyone else have other tips and things i should think about?

How are these capacitor active balancers?
https://www.aliexpress.com/item/4000241881122.html?spm=2114.12010612.8148356.11.6b381313pKtjV2
Are they better than those cheaper active balancers that only transfers energy between cells?

 

[john61ct]

I think the HA02 looks like a promising option.

And Chargery's have several users that are members here.

And yes in general I think this category / method of balancing is the way to go, great flexibility

if you don't design using sub- packs to facilitate using the hobby charger approach.

Use with multiple / any config packs.

Switch to bottom- or middle- balancing if you prefer.

Only balance when needed, less often with a faster rate

facilitates a no-BMS approach if desired, no cell-level complexity / fragility while in use.

 

[John in CR]

So if there was a leaky cell (group) that kept dropping in voltage, this balancer would drain the rest of the cells attempting to keep that group alive?...

Yep, IOW a Battery Murdering System that can kill a pack without the BMS actually having a failure.

 

[vreppeto]

Anyone else have other tips and things i should think about?

How are these capacitor active balancers?
https://www.aliexpress.com/item/4000241881122.html?spm=2114.12010612.8148356.11.6b381313pKtjV2
Are they better than those cheaper active balancers that only transfers energy between cells?

I have a 16s version of this balancer. I had it working nicely. It balanced a 16s lifepo 10ah pack in just a few hours. The eng jumper on the board is for a sleep switch in the open position. The instructions seem to indicate this should be open when initially connecting the battery pack. The feedback remarks indicate actually balancing current that reflects honest advertising.


I just received an answer to a question; ICGOGOGO on aliexpress says this board can safely be connected simultaneously with the JBD smart BMS. I tried it but when I turned it on caps 8 and 9 blew up. I will check my wiring but I believe it is just really cheap capacitors. I read about this happening and I am currently searching for the post. I suspect the caps are just crap. I tried to find replacement caps that 10x13mm and I the closest I could find to 2200uf is 1500uf. I think the 2200uf rating is an exaggeration.

edit: I remember reading that caps 8 and 9 can blow if the unit is connected to a battery that is already fully balanced. That is exactly what I did and they did blow.

 

[JackFlorey]

Yep, IOW a Battery Murdering System that can kill a pack without the BMS actually having a failure.

This isn't a BMS, just a balancer.

 

[amberwolf]

I just received an answer to a question; ICGOGOGO on aliexpress says this board can safely be connected simultaneously with the JBD smart BMS. I tried it but when I turned it on caps 8 and 9 blew up. I will check my wiring but I believe it is just really cheap capacitors. I read about this happening and I am currently searching for the post. I suspect the caps are just crap. I tried to find replacement caps that 10x13mm and I the closest I could find to 2200uf is 1500uf. I think the 2200uf rating is an exaggeration.

edit: I remember reading that caps 8 and 9 can blow if the unit is connected to a battery that is already fully balanced. That is exactly what I did and they did blow.

Sounds like a crappy design to me, but I haven't ever seen a schematic of it. They'd have to have it setup such that a very large current would have to flow thru those caps, to cause that. (high current flow, especially in reverse, outgasses electrolyte rapidly, forcing the can to swell and either blow it's seal or explode.)


FWIW, caps can blow upon connection to a circuit if the caps are installed backwards, current flow does the same thing there. But this would happen regardless of state of battery; so is probably not the case.

 

[john61ct]

This isn't a BMS, just a balancer.

The term BMS is used to cover a variety of collections of functionalities.

"Non-protective BMS" is one of the terms used for these devices.

Can be paired with "non-balancing" or "protective-only" BMS.

 

[fourbanger]

Wow. That's just plain poor. Thanks for the heads up on THAT one.

 

[JackFlorey]

"Non-protective BMS" is one of the terms used for these devices.

To me, battery protection is the #1 (and defining) job of a BMS.

To put it another way - a Trimetric monitor is not a BMS. Neither is a balancer. But the Orion Jr is.

 

[john61ct]

Of course, we can apply whatever definitions we like.

If someone says "I'm running my bank with no BMS"

but then talks about their charging setup with redundant HVC protection

600A fuses

A homemade Arduino displaying per-cell voltages while in use

and both pack level and per-cell LVC cutoffs.

I'm going to disagree, they have "a BMS", just not a single device with that label.

Some have their "non-protective BMS" (balancing device) disconnected, except when doing periodic maintenance, balance only when needed.

Some have no gear to automate that function, just do it manually with their lab style PSU.

So many ways to skin the various cats. . .

 

[amberwolf]

I just use a Biological Monitoring System (mark one eyeballs, brain, and hands).

 

[JackFlorey]

If someone says "I'm running my bank with no BMS" [and they have] pack level and per-cell LVC cutoffs

Then they have a BMS.

Agreed that you can call it whatever you like if you want. But there's some benefit in having terms mean things.