BMS16C - does it really balance cells?

Hi Overlap.
I have detailed data on three functioning BMS16C balancers. My best bet is that your balancer is not defective. The BMS16C works a little different and I am still figuring that out. I can say that the BMS16C is clearly better than the DB8s.

Can you get me data? The actual voltage reading for all 16 cells. If you do this manually I would like to see it at rest after setting over night with the bumper plug unplugged. And also right after the completion of a full charge.

If you have CellLogs and want to automate the data collection, then shoot me a PM so that we can start Emailing the data files.

Then we can look at that data and give you some informed advice.

Thanks,
Dan

 

The BMS16C online manual on page 2:

Upgraded Version:
One single BMS manages 16 pairs of battery cells, i.e a total of 32 cells;
Suspending charging once any single cell exceeds 3.8V;
Suspending output once a single cell drops down to below 2.5V;
System sleeps if any cell drops to below 2V;
Advanced cells inter-charging (balancing) feature at 1.0-1.4 A current (coming soon).

So, is the manual outdated then ?

Pb

 

Hi Pb,
I do not know about what is meant by "coming soon". I do know that what I have now is clearly better than the DB8s. I appears to me that the balancing feature is working today. Clearly the overcharge protection is working.

My guess would be that what you found is out of date.

Thanks,
Dan

 

Just for sake of argument here, balance can only be observed on the high and low end of the battery pack (the voltage curve on LiFEPO4 cells is a straight line in the middle of the pack). If you are observing significant delta voltage in the middle of pack discharge you probably have capacity issues.

Depending on the accuracy of the BMS in question, it's reasonable to expect less than 10 millivolt differential when balanced (0.01v) after a full charge.

Andrew

 

Hi Overlap,
I think you are too symplistic in your thinking about this. Balancing is mor complicated tha just looking at the delta from high to low.

Taking the data manually is simple. Push the button, write down 4 voltages, push the button, write down 4 voltages, Push the button, write down 4 voltages, Push the button, write down 4 voltages. You do not need a data logger to do that.

If you choose to not communicate with data I think that you will not get the information that you need for improvements.

Thanks,
Dan

 

Hi Andrew,
Yes this is very true. That is why I asked for data at the end of a charge as well as at rest. These can both be read manually. Getting data at the low end is much harder to obtain. You really need a data logging system to capture the low point.

Almost flat. I see a variance from 3.1 to 3.3v in the middle. But the variation from cell to cell can be enough to confound this data even in perfectly good cells. I respectfully disagree with your statement about probably have capacity issues. I would say may have capacity issues.

Thanks,
Dan

 

Ohhhh Dan ,

I never said he did have a capacity problem, just pointing out that he need only pay attention on the high and low end of the pack for balance. I also said that if he is observing a /significant/ delta-V in the middle that is a tell-tale sign of a capacity problem and/or a balance problem.

My opinion would be that 0.2v is high for varience in LiFEPO4 batteries, especially in the middle of the pack--but hey, who cares about opinions anyway? . Back when I had a LiFEPO4 system in my car I was always able to get the cells within 0.01v (10 mV) balance of each other using top-end passive balancing--granted, it was with a different BMS...but I digress.

Load can screw up balance readings with a multimeter / cell-log (the BMS would be applying a small load if its actively balancing too). I assume the readings you are referring to were taken at "rest" w/o the BMS active?

Andrew

 

Hi Andrew,
I think you and I are in general agreement. Maybe our differences are from misunderstanding the other persons statement.

I assume you 0.2v comment comes from my commenting that the "flat" area goes from 3.1-3.3v. What I meant is with a LiFePo4 cell disconnected from everything and at rest, it will be at about 3.1v if its SOC is at the bottom of the flat area. And it will be at about 3.3v if its SOC is at the top of the flat area. In this region smaller variations confound our ability to really know real SOC. I did not mean to infer that those variations were in the range of 0.2v.

Thanks for your comments. You do help to clarify what I am thinking in many cases.

Thanks again,
Dan

 

I guess the bottom line is "how do we determine that the balancing is getting done ?" i guess if your delta V increases as the batteries get cycled/used (over many days); then it's not working ?
Pb

 

i have charger balancers on my Zenn. they have lights that flash when they are actively balancing. i take it, yours do not?

 

Hello overlap,
Thanks for the data.
So Cell One is at 3.43v after 1 hour. Based on my experience I would guess that this cell was at 3.8v when the balancer turned the charger off.

If you leave the system unplugged over night you will see that this cell# 1 will drop to about 3.35 and the others will also drop some.

Can you leave the system unplugged overnight and then collect the data in the morning before doing anything with the system.

Thanks,
Dan

 

I think something is either wrong with some of my cells or my BMS16C balancer. After using the system for a while I observe on my Scan Gauge that there is no longer any current being provided by the Enginer system. When I check the cell levels on the BMS, I find cells 1-4 all below 3 volts. All the other cells are over 3.2 volts. This has happened about 4 times now. While I am pleased that the BMS is shutting down the system to protect the batteries, I am concerned that the same 4 cells are always the low ones. Shouldn't the balancer keep the cells relatively uniform? Prior to installing the BMS I was using the system and monitoring the cells with CellLogs. I never saw this type of situation, all the cells remained fairly close in terms of voltage. I also find it annoying that the red low battery light does not come on when the BMS shuts down the system. If I didn't have the Scan Gauge I wouldn't know the system was not operating.

 

Hi Jeffrey,
You can put the CellLogs back on for one Charge/discharge cycle. Then record the data and download the data in LogVeiw and Save the table to XLS file. Then if you Email me those two xls files I can analyze them and give you a good Idea of what is happening.

I have had the BMS16C on for several weeks now. In my case it seems to slowly bring cells one at a time up to match the highest cell. In the middle of this process it does not look right but in the end it worked very well.

You can speed this process if you want. To have me help you further, shoot me a PM with your regular Email address.

Thanks,
Dan

 

I didn't disconnect the BMS yet but I wrote down some numbers. The first set is the morning after an overnight charge:
01- 3.33 02- 3.34 03- 3.34 04- 3.32
05- 3.33 06- 3.35 07- 3.42 08- 3.42
09- 3.33 10- 3.35 11- 3.38 12- 3.36
13- 3.33 14- 3.40 15- 3.47 16- 3.45

The second is over 24 hours after the BMS shutdown the system:
01- 2.92 02- 2.77 03- 2.64 04- 1.60
05- 3.20 06- 3.22 07- 3.26 08- 3.26
09- 3.24 10- 3.26 11- 3.26 12- 3.20
13- 3.22 14- 3.27 15- 3.27 16- 3.27

The system is currently charging.

 

Bank #1 is not good, esp. cell #4. It's in danger of going bad if it hasn't already.