adjustable Cheap Chinese Active Balancers

Stefanseiner

Member
Joined
Apr 16, 2020
Messages
136
Hello everyone,



according to this thread -> Cheap Chinese Active Balancers


daromer said:
Yeah you unlock some capacity but only IF your pack nominal current draw is lower than the balancer can balance. And for most users thats less energy gained than the cost over 10 years IF you look towards that cycling Them that Hard so you need to bottom balance you wear out the cells within a year anyways :)


I think it disqualifies these balancers because the balancing-start-voltage is too high and you can't change the parameters.





But: has someone tested this chinese active balancer with bluetooth app?




image_aplcvw.jpg





image_qdicgt.jpg




-> Aliexpress







There are variants with 1A, 2A, 5A or 10A balancing current, and it comes with bluetooth and some values are adjustable via app.

But the description is very poor and I can't see which parameters can be changed in which range.

I installed the app just for testing but without having the balancer connected you only can change the language (between chinese and englisch).



So does anybody knows this balancer and can tell something about it?
 
That one is really a passive balancer (you can tell because it does not have any inductors).
It will only bypass (drain) current from cells.
Also if it is bypassing the claimed (high) currents, that heat has to go somewhere, there are no heat-sinks so likely will overheat fast.
It may be OK for lower balancing currents.

Real active balancers are more complex & transfer energy between cells to equalize voltage.

It has been pointed out elsewhere on this forum that trying to do active balancing:
a) is not cost effective &
b) tends to hide problems eg self discharging or weak cells
 
Stefanseiner said:
Thanks for reply.

Which current would be usefull a balancer should be able to handle when we talk about
- 14s100p configuration and a minimum using of 5kWh every day
- used, 18650 cells, tested but mixed brands and models

I actually bought that active balancer but haven't had the time to play with it yet. I use another active balancer right now and works great. Using an active balancer doesn't hide anything any more than a passive balancer. You can still tell which one is out of balance especially if you set a limit on the balancing threshold. For example my current active balancer only balances cells that are over 40mV difference. If one was to run-away you can tell that the one that is always on the high limit.

The key is logging the information. I bought the one with the bluetooth like the above, but they make one that also has a canbus interface. This will make it easier to log the information into my existing grafana interface. I now have to figure out how to use python to interface to my bluetooth only version. that's a pain.

You ideally shouldn't need much current transfer. 1-2A would be ample. If you find your cells move more than that then you have issues with your pack. A properly sized and balanced pack should have the same capacity. If I were to turn my balancer off my cells would be fine for months before needing another balance.
 
Stefan passive balancer for that small should be enough to have 100-200mA balancing current.
 
Ok so the 1A or 2A version shoud be Ok but you should have an eye on the system from time to time so you can see issues and they won't get hidden.

I found another interesting one:

image_nzoawy.jpg

https://aliexpress.com/item/4000529723243.html

Seems to be a combination of the above balancer and the known smart bms including protection board,
so that seems to be an all-in-one solution.

I guess I will give the 2A / 300A version a try.

300A is max current, 150A continous and this is twice of what I wil use, so there should be some power and heat reserve.
Canbus or WiFi instead of bluetooth app only would be very nice but I can't find one.
 
Redpacket said:
That one is really a passive balancer (you can tell because it does not have any inductors).

Talking about this:https://www.aliexpress.com/item/32945565819.html
It has huge capacitors andone big inductor (bigwhite box next to capacitor). So they're just using adesign withone pump and a lot of circuitry to connect it to the right cells. Vs the more often seendesign of one pump between every cell.


Redpacket said:
b) tends to hide problems eg self discharging or weak cells

If it's possible to poll the status via Bluetooth, one could log how oftencharge was pushed into / pulled out of each cell, making it extremely useful in identifying bad cells. Even distinguish between weak cell and self discharge problem.


PS. Haven't tested or seen review of the device. But anything else simply would not make sense or work at anywhere close to 1A.
 
ajw22 said:
Redpacket said:
That one is really a passive balancer (you can tell because it does not have any inductors).

Talking about this:https://www.aliexpress.com/item/32945565819.html
It has huge capacitors andone big inductor (bigwhite box next to capacitor). So they're just using adesign withone pump and a lot of circuitry to connect it to the right cells. Vs the more often seendesign of one pump between every cell.

Redpacket said:
b) tends to hide problems eg self discharging or weak cells

If it's possible to poll the status via Bluetooth, one could log how oftencharge was pushed into / pulled out of each cell, making it extremely useful in identifying bad cells. Even distinguish between weak cell and self discharge problem.


PS. Haven't tested or seen review of the device. But anything else simply would not make sense or work at anywhere close to 1A.


Ah, OK, understand now re "switched pump, thanks.

Agree this unit has more reporting & this can show up issues.
I was thinking of the "one wire" type active balancers which are "blind" & have little/no reporting.
Looks like the unit here does a lot more.
 
There are a few reviews out there. Here is a YT that leverages one per pack:
They appear to work great, there are some versions that support CAN bus.
 
The balancer you list in the OP is an active balancer, they operate by transferring charge from the highest cell, to the large onboard capacitors, once the capacitor is full they then transfer charge to the lowest cells.

The do work, they are configurable but they are VERY slow at balancing large packs, as they only manage 1 cell at a time and only the amount that the capacitor can hold.

From memory a full charge->discharge cycle of the capacitor was almost 2 minutes.

So far I have not been able to justify the money to see if a 10A version is that much better.

Side note: These balancers require 40-100v power source to operate, if your pack is less than this you require a boost converter.
 
not2bme said:
I actually bought that active balancer but haven't had the time to play with it yet.

BaronVonChickenPants said:
The do work, they are configurable but they are VERY slow at balancing large packs, as they only manage 1 cell at a time and only the amount that the capacitor can hold.


Now I got mine but I can't get it to work.




After studying the manual, the article description, checking all connections twice and several messages with the seller he send me this picture

image_xgvncu.jpg


It is nowhere else mentioned but You have to connect a DC source with 5V to the main input and output shortly to activate the device. For me I used two 18650 with 7.4volts and it worked well.

OK, first problem solved after 2 days.

But now I am stucking at the settings area which is locked by a password I don't know

image_pwhsxf.jpg


if I insert the device password = bluetooth password...

image_tljslh.jpg


...it is wrong. This is for the new password I set (12345) as for the original password too (1234)

image_ldlezq.jpg


And there is no other passwort which is mentioned at the description or manual.
I also tried "0000" and to leave the field empty, tried to restart the device (power off, waiting some time, power on) but nothing works.
So I tried the older app version but I can't get access to the settings area.

Any ideas?
 
now I am using a couple of this BMS balancers for some month

image_cijyzw.jpg

  • JK-BD6A24S-10P (old edged version) -> 0,6A balancing current / max. 24S / 100A continous power -> sold after a month because the 0,6A was too low for my setup
  • JK-B2A24S-15P (old edged version) -> 2A balancing current / max. 24S / 150A continous power
  • JK-B2A24S-15P (new version, with QR-code on top)
  • JK-BD6A20S-10P (brand new version) -> 0,6A balancing current / max. 20S / 100A continous power
Bluetooth app

image_cbwqio.jpg


here are some internal views. This is the "old version" but I think they only changed the case

image_hrbnte.jpg



image_boqyhy.jpg


the newer one

image_bvpubl.jpg



And want to sum up a little bit.

Pros:
  • BMS and active Balancer into one device -> no need to double the cables to the akkupacks
  • balancing is not working from cell 1 to 2 to 3 to 4 to 5 like some balancers do, but directly from the highest to the lowest. For about 3 seconds and then searching the next weakest cell.
  • one GUI for controlling / monitoring
  • they all are running with the same App via bluetooth, with the actual app version (v. 2.9.xx and above) the app runs stable. Older versions are buggy and won't accept any value you want to set
  • prices are around 135USD for the big version with 2A current and 60 USD for the smallest one. That's more then the cheap standard-BMS, but much less then a batrium.
  • runs stable
  • option fr RS485 and Canbus -> not tested yet but I will next month when the dongles arrived
  • many, many options to set
image_vlsdah.jpg



image_hlpvgn.jpg



image_btrfit.jpg



image_wpkank.jpg




Cons:
  • the manual is horrible and incomplete. The official manual describes the balancing part only, no BMS part, and some importand things are missing completely (such "little things" like the admin passwords and how to start the device -> you'll find this information at the next point
  • bluetooth connection is very instable and lasts something about 5 minutes, sometimes half an hour. You can reconnect directly and it's no big deal, but therefor you can't use an old tablet or so and mount it next to the powerwall for permanent monitoring (as I tried) because every time you want to have a look you just have to reconnect first
  • no button / switch to start the device
  • although it is shown on all pictures, the BMS comes without ringterminals at all, you have to do this for your one
image_chgthg.jpg



additional information:
  • sellers: there are many sellers on Aliexpress with this BMS. I tried three of them and none could help with questions or the passwords you need, and one sends out the wrong version. So seller nr. 4 is ICGOGOGO and it seems that it's the only one with some knowledge and the complete manual. No affiliate link, just for information because via the search you'll find it very hard -> ICGOGOGO Store
  • prices: for me, the version with 2A / 150A is the best deal. The ones with lower balancing current aren't cheap enough for the loss of half / third of the balancing current. Except the new "slimmed down" version with 0,6A / 100A for under 60USD may be a deal but my order is still on its way and I couldn't test it out
  • you will find the app and manuals here -> official Download-Site
  • because the official manual is incomplete I asked some seller and one gave me a modified version, you can download it here -> download Pdf (7,5MB) from Filehorst.de
  • passwords: you need two. The bluetooth password is "1234" and the passwort you need to enter the settings section is "123456". You can change both and so that I can better remember them I changed both to "12345" ;)
  • device will only start-up when battery-voltage (between the thin balancing-wires) is above 40,0V and B- / P- are connected to the batterie
  • furthermore to start-up the device you need a short impulse of 5V with positive on the "B-" and negative at the "P-" I am using two 18650 in series (7,4V) and this works fine too, in combination with a little dip-switch
image_tvqwjc.jpg



Network access
my next step will be to monitor the BMS via RS485 and network so I don't have a need for the bluetooth app anymore.

image_hwhuxw.jpg


It's not the app which is annoying me, but the fact that for bluetooth I have to stay close to the powerwall.

There is an Github project for this purpose, but with some code invented for the version without BMS function.
Infos at this thread -> JK-B1A24S / JK-B2A24S Active Balancer
 
There is a bit of a misconception about the way in which these balancers work.

"device will only start-up when battery-voltage (between the thin balancing-wires) is above 40,0V"

This is because they use a single power source from the whole battery to power a buck circuit, controlled, CC, CV which is then switched to the lowest voltage cell to charge it separately, powered by the whole pack. They only charge the lowest cell and do not directly (only indirectly via the whole battery string) draw power from the highest voltage cell.

This makes them sort of an ideal bottom/discharge balancing unit but makes them less than ideal for protection during the charge cycle. This is what could make them near on useless for charge protection, regardless of the cell balance current.
 
completelycharged said:
This is because they use a single power source from the whole battery to power a buck circuit, controlled, CC, CV which is then switched to the lowest voltage cell to charge it separately, powered by the whole pack. They only charge the lowest cell and do not directly (only indirectly via the whole battery string) draw power from the highest voltage cell.

Are you sure with this? I am not.
Because while balancing the cell with the highest voltage and the one with the lowest is marked in the app (first screenshot, red is low, blue is high), and while processing you can see voltage changes on both - the higher goes down and the lower goes up.
And after charging the powerwall to full level, the balancer equalizes the high-voltages too.
 
I fried one of the same type of design and traced out the board and circuit to figure out how it worked and why it went the magic smoke route. The high input voltage start is what also gives it away.

Does the highest cell voltage actually "decrease" as the lowest cell voltage increase with no charge to the pack overal and only those two cells change ? The highest cell (and all the others) would decrease as the power is drawn from the pack and (via an isolated buck) fed back into the lowest cell.

The board in the first post, the top right (yellow) transformer is the HF isolated buck.

The more recent boards you have I have not looked at yet in detail to confirm. The first board is 100% isolated buck.


If you can post a decent image of the internal board (top and bottom) I can confirm the layout. The overhead image makes some component identification imposible.
 
completelycharged said:
Does the highest cell voltage actually "decrease" as the lowest cell voltage increase with no charge to the pack overal and only those two cells change ?

yes, that's exactly what I can notice.
The last 14s60p systen I build was a little unbalanced as I finished the building. Before connecting it to the inverter I left it two days or three at the garage without charging, but running BMS.

Pictures:
I already tried to open the box earlier, but I wasn't able to do so. Maybe they glued the FETs to the case with some thermal glue
 
not2bme said:
completelycharged said:
Does the highest cell voltage actually "decrease" as the lowest cell voltage increase with no charge to the pack overal and only those two cells change ? The highest cell (and all the others) would decrease as the power is drawn from the pack and (via an isolated buck) fed back into the lowest cell.

See this post.
https://secondlifestorage.com/showthread.php?tid=9591&pid=65679#pid65679

Nice, Looking at the chart in the posts of the voltage levels it does indeed appear to be pulling from the highest cell(s) and charging the lowest. Will have to have a closer look at the board on this one.
 
https://de.aliexpress.com/item/4000001793955.html?spm=a2g0o.cart.0.0.5e063c00cKmEnp&mp=1

Hope this link to bls works, dig a little bit deeper and you can find one for li ion, the only thing you must calculate...if they say 110a it would be 80a max drawn an hour..
And please go for a bms and not a balance only.

Best igor


Ps balancing your cells you can do every year with disconnect all your packs and connect them for 72 hour in series, disconnect them and wait 72 hours for a v drop.
Or test the whole pack within 72 hours, calculate the soh and ir
 
Back
Top