question

vovanenok avatar image
vovanenok asked

REC Q BMS vs Active Balancer

Hi all, it's been around 2 months since I installed my new system which includes:

  • 2 x Quattro 5000/70 120V inverters
  • 2 x Victron 250/100 SCCs
  • 2 x Victron 250/60 SCCs
  • Cerbox GX
  • REC Q BMS
  • 2p16s battery bank (3.2V 420A)

The last 2 days we really sunny and I found the REC BMS was not good in balancing my battery pack at higher voltages

1644706717468.png

As you see it's almost 0.35V difference between the MIN and MAX cells.

Is my bank too big? 2p16s battery bank (3.2V 420A) ~ 40kwh (though I don't think those aliexpress batteries can provide 100% capacity). Any BMS users, do you use a separate active balancer?


Before the last 2 days it was mostly cloudy and at lower voltages the cells looked balanced

Thanks a lot in advance!

BMS
1644706717468.png (172.8 KiB)
2 |3000
Toggle Comment visibility. Current Visibility: Viewable by all users

Up to 8 attachments (including images) can be used with a maximum of 190.8 MiB each and 286.6 MiB total.

7 Answers
nebulight avatar image
nebulight answered ·

You should have put all your cells into parallel, then charged to 3.6 then let them rest, then charge up to 3.65v and let them rest together for a few days. This will ensure they are all balanced at the top.

Just because they were all at 3.29v doesn't mean anything. They could have been at 95% state of charge while another cell was at 60% state of charge. This is due to lifepo4's very flat discharge curve. The voltage is almost the same 3.2v throughout the whole discharge of the cell. Only at the very end does the voltage drop. Here is a discharge graph for a lifepo4 cell and you can see the voltage is pretty much the same between 95% stage of charge and 20% state of charge. If the pic is too small, if you are on a computer right click and open in new window.


2 comments
2 |3000
Toggle Comment visibility. Current Visibility: Viewable by all users

Up to 8 attachments (including images) can be used with a maximum of 190.8 MiB each and 286.6 MiB total.

vovanenok avatar image vovanenok commented ·

thank you very much @nebulight , it is very helpful. As I understand this is what I should do ASAP.

If connected all in parallel, what should I use to charge them at 3.6V?

I live completely off grid, but I have 2 large 24v x 200Ah Lifepo4 batteries, so I could probably hook them up to my inverters while working on my main battery bank

0 Likes 0 ·
nebulight avatar image nebulight vovanenok commented ·
What I would suggest since you already had these in service would be to charge them up what the BMS sees as full. Then take them out of service. You can use a benchtop power supply to get them up to 3.6v. You can get some fairly cheap units on amazon that can go up to 10 amps so it may take a while. I would do once cell at a time, then put them in parallel and charge them up again as one large cell.
0 Likes 0 ·
Paul B avatar image
Paul B answered ·

This just indicates that the batteries need MORE time to top Balance . this may take a long time so I would try to keep the batteries at the 56 volt level until rec can balance them for you


give it time to do its job - it can ONLY be done when the batteries are at there max charge levels

7 comments
2 |3000
Toggle Comment visibility. Current Visibility: Viewable by all users

Up to 8 attachments (including images) can be used with a maximum of 190.8 MiB each and 286.6 MiB total.

nebulight avatar image nebulight commented ·

Also to add on to what @Paul B stated, did you first do a top balance of the cells connected in parallel before you installed in series and to production? SOC for each cell may have and most certainly was, different when they arrived.

0 Likes 0 ·
vovanenok avatar image vovanenok nebulight commented ·

Thanks @nebulight When the batteries arrived, they all were almost perfectly balanced.

0 Likes 0 ·
Show more comments
vovanenok avatar image vovanenok commented ·

Thanks @Paul B

Yes this is what my REC Q BMS was doing, it signaled the Cerbo GX / charge controllers to stop/float and it was giving error code 1 (Single or multiple cell voltage is too high. Wait until the BMS does its job).

I'm reading the REC Q BMS description, it says passive cell balancing up to 0.9 A per cell

It is passive and it is only 0.9 A, so that's why it was taking so long, plus it was wasting the energy. Not cool :(

What active balancer would you recommend?

0 Likes 0 ·
Paul B avatar image Paul B vovanenok commented ·
Nothing wrong with Pasive balancing the wastage is very small like 3 watts

We mainly use 123smart BMS but thats also Passive but upto 1.5amps

we have also tried TOA BMS from NZ this is active unit and so are some of the Rec units

Active I have found is a bit missleading as it may take 2amps out of the overfull cell for say one min then stops taking amps and discharges what it HAD taking into the lowest cell so a 2 amp active balancers RMS is only really 1 amp, but they do not wast the energy . (I do not no of a Active bms unit that can take power from one cell and simutanilasy deposit it into another) so a lot of them are overstating what they can do.

for myself Pasive is fine



0 Likes 0 ·
Show more comments
pierrot avatar image
pierrot answered ·

Hello @vovanenok, did you test the capacity of your cells ? because I believe that 420Ah 3.2V Aliexpress cells were proven to be scam... I hope I'm wrong.

Are those Lifepo4 cells ? Because your graph shows 3.715V Max Vcell, and as far as I know Lifepo4 cells should never be charged over 3.65V..

In your case, passive balancing should work, but not when your cells are in the "red zone", because at high voltage, internal cell resistances go through the roof...




3 comments
2 |3000
Toggle Comment visibility. Current Visibility: Viewable by all users

Up to 8 attachments (including images) can be used with a maximum of 190.8 MiB each and 286.6 MiB total.

Paul B avatar image Paul B commented ·
Good pickup I did not see that 420A/H note , I did notice that the cell resistances were all over the place - so thats not good as well
0 Likes 0 ·
klim8skeptic avatar image klim8skeptic ♦ Paul B commented ·
I would not be sure that a BMS should be relied upon to measure cell resistance, given user cell interconnections/user BMS wiring..
0 Likes 0 ·
vovanenok avatar image vovanenok commented ·

hi @Pierrot ,

1. I know about the aliexpress battery "quality" and I didn't expect to get the full capacity. My intention was to significantly improve my old lead acid 10kWh battery bank performance (capacity + high load operation).


Are those Lifepo4 cells ? Because your graph shows 3.715V Max Vcell, and as far as I know Lifepo4 cells should never be charged over 3.65V..

2. Yes those are Lifepo4. Based on the attached PDF, I should keep the batteries between 50.4 and 54V. How bad is it to overcharge a Lifepo4 battery to 3.8V once? Also curious how bad would that be to discharge it one time let's say to 2V? I'm wondering if something goes wrong with the BMS, would overcharge / overdischarge kill the batteries significantly?


3. I didn't perform a capacity test, but a few days ago it was sunny and my VRM showed 33 kWh production and 13 kWh consumption. Based on those numbers, can I assume the batteries absorbed 33-13=20 kWh?


In your case, passive balancing should work, but not when your cells are in the "red zone", because at high voltage, internal cell resistances go through the roof...

4. So basically I need to balance them at lower voltage as possible?

Thanks again for you help

0 Likes 0 ·
vovanenok avatar image
vovanenok answered ·

This is a screenshot from this morning:

1644770038888.png

And my settings:

1644770295828.png

1. Do you see anything wrong in my settings?

2. Since I don't expect 100% of those aliexpress batteries capacity, what would you recommend to enter as Capacity setting? What that setting is used for? Again my battery bank is 2p16s x 3.2v 420Ah

Thanks again for all your contribution!


1644770038888.png (174.8 KiB)
1644770295828.png (168.9 KiB)
2 comments
2 |3000
Toggle Comment visibility. Current Visibility: Viewable by all users

Up to 8 attachments (including images) can be used with a maximum of 190.8 MiB each and 286.6 MiB total.

nebulight avatar image nebulight commented ·

I'd say your balance start voltage is too low. 3.2v to start balancing, it's always going to be balancing. You may be attempting to make up for the fact that your pack wasn't top balanced before install. Since you have very large cells, if there is a large imbalance, it could take weeks/months to balance. I would remove the cells from service and perform a proper top balance and reinstall. Then set your balance start voltage to 3.4v.

As you can see from your photo, the cell voltage is pretty close, but when you are at the top, that's when the voltage imbalance will show itself.

As for your capacity, you can perform a proper top balance, then do a capacity test so you know how much capacity you have. Those fake 420ah cells were likely b grade 280ah cells so chances are they aren't even full capacity. I'd say you've got something close to 250ah per group so maybe a total of 500ah? But the only way to know for sure is a capacity test.

0 Likes 0 ·
info-6 avatar image info-6 nebulight commented ·

What is against starting balancing at 3.2 Volts?

I prefer to start earlier, in order to avoid greater imbalance and subsequent permanent balancing at the top. However, I'm pleased to learn more... ;-)

0 Likes 0 ·
benm avatar image
benm answered ·

Hello,

I agree with others that you should always do a top balance before using the cells.

Or expect balancing takes very long.

I advice to used the REC default settings as start, later fine tune if needed.

Maybe keep end off charge 3.55V or 3.50V



Set start balancing 3.45V, and monitor on your GX the CCL level. (devices/REC/parameters)

If the first cell reached 3.45V the CCL will be lowered.

Your chargers will put out lower amps from this point.

IF CCL dont act as this, you have other settings, maybe in DVCC, cables etc not correctly.

If all cell reached 3.45V your bank is almost balanced. But this can takes long.

It could be that CCL level fluctuated a lot, for a long time.

After ALL cells reached 3.45V it will not take long cells reached 3.55V. After reaching all cells 3.55V, you should see CVL level is lowered.


In your first post, I see one cell as 3.71, this indicate a huge unbalance!!

But if your system is working correctly that cell should not go that high. AMPS would be lowered to avoid that cell going so high. I think there is something wrong with DCVV settings, cables or other settings in GX.

It almost look that your chargers are not controlled by the GX, maybe not connected with ve-bus or v-direct cables to the GX device or the REC CAN-data info dont arrive at GX at all !



If you make balance voltage lower that can created a unwanted unbalance. One example from what I have seen. Balancing was set at 3.4V. One cell that was the highest at 3.42, BMS start discharging this cell. Later after reaching 3.5 this cell was lower as the other. If BMS didnt discharge this cell at 3.42, probably it was not needed to charge this cell extra at 3.5V.

This was with cheap chinees cells, and is probably inherent with the quality, internal resistant etc.

2 |3000
Toggle Comment visibility. Current Visibility: Viewable by all users

Up to 8 attachments (including images) can be used with a maximum of 190.8 MiB each and 286.6 MiB total.

meltemicaz avatar image
meltemicaz answered ·

Has anyone used this NEEEY Active Balancer with the REC Q BMS? I have watched some YouTube reviews that seem positive.


NEEEY Active Balancer

2 |3000
Toggle Comment visibility. Current Visibility: Viewable by all users

Up to 8 attachments (including images) can be used with a maximum of 190.8 MiB each and 286.6 MiB total.

Panos avatar image
Panos answered ·

Hi meltemicaz


Yes i use it and it works well

2 |3000
Toggle Comment visibility. Current Visibility: Viewable by all users

Up to 8 attachments (including images) can be used with a maximum of 190.8 MiB each and 286.6 MiB total.