diy solar

diy solar

Batteries aren't charging too full, what I'm I doing wrong?

Dacflyer

New Member
Joined
Jun 5, 2020
Messages
349
I been learning a lot on here and I'm fairly good at what I'm doing, but I'm having an issue on a new setup.. My batteries are not getting a full charge.. I have 9x 295w solar panels 3s3p, about 105vdc going into Epever AN6415, this goes to 14s 3p battery bank, each string has its own BMS ( JK SMART-BMS)
I've gone all thru my settings, i do not know what setting is not right preventing me from getting a full charge.. I'll post screen shots here and see if anyone can give me any advice to try
please ask me any and all questions.. I can't see what the deal can be,,, SOC on the Epever shows nearly full charge, but the BMS's are all nearly half as much..
 

Attachments

  • 20211014_163940.jpg
    20211014_163940.jpg
    97.6 KB · Views: 22
  • Screenshot_20211014-163625.jpg
    Screenshot_20211014-163625.jpg
    217 KB · Views: 24
  • Screenshot_20211014-163555.jpg
    Screenshot_20211014-163555.jpg
    216.8 KB · Views: 25
  • Screenshot_20211014-163700.jpg
    Screenshot_20211014-163700.jpg
    217.8 KB · Views: 20
  • 20211014_163928.jpg
    20211014_163928.jpg
    111.8 KB · Views: 17
  • 20211014_163918.jpg
    20211014_163918.jpg
    115 KB · Views: 10
  • 20211014_163953.jpg
    20211014_163953.jpg
    102.3 KB · Views: 11
I would assume with a 14S pack that the chemistry is NMC not LFP. I that is the case 4 volts per cell or 56 volts is a conservative voltage that might be 90%. How did you reach the conclusion that your battery is not getting a full charge?
 
I have not seen that wire resitance reading before. Also, I don't like the differences in your wire readings.
 
I have a similar setup with a JK BMS on a 14S Li NMC bank. But mine is built from Chevy Bolt Cells, and I am using a Schneider XW-pro to charge the cells, so the charging is a little different, but the BMS should act the same.

I have been trying to help Dacflyer with settings, and his system seems to be working well, but the BMS units just don't seem to want to balance out and show full state of charge when all of the cells stop charging at 4.1 volts.

I was originally charging my cells to 4.15 volts per cell, and my JK BMS was set to 4.2 volts for the cell over voltage protection. It would report 99% or even 100% most of the time when the charge cycle ended at 4.15 volts. I then lowered my full charge voltage to 4.1 volts per cell. I expected the JK BMS to report a lower state of charge, and for a few cycles, it was only topping out at 88 to 92% charged, but when I checked on it 2 weeks later, it does now report 99% with the cells at 4.1 volts. The software in my BMS seems to see the termination of the absorb charge as full, even if it is not the peak voltage set. Dacflyer's three JK BMS units appear to be stuck and are not resetting the full charge state.

Is there anyone else here using the JK BMS that is not getting a proper state of charge reading in the BMS? Did you figure out what is causing it? Here is my JK-BMS shortly before the charge ended.
IMG_0026.PNG
You can clearly see the cells are still under 4.1 volts, and it is showing 99% full.
 
I would assume with a 14S pack that the chemistry is NMC not LFP. I that is the case 4 volts per cell or 56 volts is a conservative voltage that might be 90%. How did you reach the conclusion that your battery is not getting a full charge? The remaining % is staying around 50 to 70ish % of full.... You can see in the screen shots..
 
I have a similar setup with a JK BMS on a 14S Li NMC bank. But mine is built from Chevy Bolt Cells, and I am using a Schneider XW-pro to charge the cells, so the charging is a little different, but the BMS should act the same.

I have been trying to help Dacflyer with settings, and his system seems to be working well, but the BMS units just don't seem to want to balance out and show full state of charge when all of the cells stop charging at 4.1 volts.

I was originally charging my cells to 4.15 volts per cell, and my JK BMS was set to 4.2 volts for the cell over voltage protection. It would report 99% or even 100% most of the time when the charge cycle ended at 4.15 volts. I then lowered my full charge voltage to 4.1 volts per cell. I expected the JK BMS to report a lower state of charge, and for a few cycles, it was only topping out at 88 to 92% charged, but when I checked on it 2 weeks later, it does now report 99% with the cells at 4.1 volts. The software in my BMS seems to see the termination of the absorb charge as full, even if it is not the peak voltage set. Dacflyer's three JK BMS units appear to be stuck and are not resetting the full charge state.

Is there anyone else here using the JK BMS that is not getting a proper state of charge reading in the BMS? Did you figure out what is causing it? Here is my JK-BMS shortly before the charge ended.
View attachment 68937
You can clearly see the cells are still under 4.1 volts, and it is showing 99% full.
Did you see how high the Cycle capacity is on mine ?
I wonder if i need to or should reboot the BMS's ??
Is a bit of a hassle to do since the BMW batteries have fully enclosed battery connectors..
Usually i have to get a skinny wire to stick down in the connector to start up the BMS's
 
Nom. voltage per cell is 3.7v and 4.2vpc max.
here is one of the pages from BHU. They are only selling single batteries now, But i had bought a complete car module last year.
Trunk , tunnel and hood module.
 
Dacflyer's batteries are essentially the same chemistry as mine. Both are repurposed EV cells. His are in rigid prismatic cases, mine are pouches, but the voltage curves should be the same.

In my JK BMS, I have the cell over voltage protect set at 4.2 volts, and I have my Schneider XW-Pro set to charge to 56.7 volts. 56.7 / 14 = only 4.05 volts per cell now. When I first lowered my charge voltage, I did not make any changes to the settings in the JK BMS, and expected it to not show a full charge. And for a few days, I saw it was topping out about 85% which is what I expected. But then I noticed it accepted this new lower charge voltage as full, and it now reports 99% again when it get's to 4.05 volts. On mine it seemed like it took about a week to learn that when the charge stopped, it would not call that full.

For some reason, Dacflyer's BMS does not seem to be finding the full charge point and all three of his are at different levels, even though they are charging in series and hitting the same voltage. The main question here is if anyone else is seeing this behavior on the JK BMS. I can't get mine to do what his is doing. My guess is they are running different firmware in the BMS.

What mine is doing now is actually a little bit of a problem. I know I am only charging to 85%, but since it thinks the battery is full, it reports that I have the full entered amp hour rating available. This is not true. When charged to just 4.05 per cell, my 360 AH pack is really only holding about 300 amp hours. So when I do discharge down to 50%, I am actually at only 35% remaining, since I started at 15% down.

So I see it both ways. It would be nice to have them auto cal and all read the same at the same voltage. Maybe doing one cycle up to the full 4.2 volts per cell to get the BMS to see true 100%To do this, the cells would need to be tightly balanced to not have any hit over4.2 volts.
 
ok,, i been eagle eyeing my SCC software today. I see where it charges up to a point, then drops off..
I think this goes from bulk to float ? maybe this gives any more clues ?
 

Attachments

  • SCC capture 10-17-21.JPG
    SCC capture 10-17-21.JPG
    112.5 KB · Views: 10
see where it charges up to a point, then drops off..
Maybe I am missing something but you want it to stop charging when it’s full, right? The graph shows the charge current dropping, not the battery voltage or SOC.
This is like gas a pump shutting off when your gas tank is full, a good thing.
 
Looking at that graph, it appears each vertical major division line is at 2.5 minutes apart. So this shows a total charge time of only 4 minutes. That is very short. I assume your batteries are just full. You need to run them down a bit and let them do a full charge cycle that lasts over an hour. That plateau looks like 32 to 33 amps. Is hat the total current that is feeding all three batteries? If so, each one is only seeing about 11 amps with 3 units. Your stated capacity is 75 AH amps per unit, for a total of 225 am hours.

My system is now 720 amp hours, so I will try to work the math to see where this should fall. I will use my charge data from yesterday. Here is my graph
FullCharge10-16-21.PNG
My charge current runs about 29.15 amps into the 720 amp hour battery bank. This is a slow charge rate of just 0.04C rate. Your 32.5 amps into 225 amp hours is 0.14C rate, so about 3.5 times faster. That is fine, I will use that as a factor. That rate is certainly safe for the cells. My charge cycle started at 8:55 am, but to make this easier, I will use the time from the pack being at 53.2 volts (3.8 volts per cell) to see what your should expect. My charge then ran from 9:43 am to 3:22 pm. That is 399 minutes, or 6.65 hours. At an average of 29.15 amps we get 193.85 amp hours were put into the calls. Total capacity of 720 amp hours means I put in 27% of the capacity in that time. Assuming, I am topping out at 85% charged, that means I stated at 85 - 27 = 58% charged. So 53.2 volts on a 14S pack should be pretty close to 58% state of charge. That is actually pretty close to what I see on the BMS as well. It shows a little high, like 64% because it thinks 85% is actually full.

So, if you can run the batteries down to say 3.8 volts per cell (53.2 volts total) that should be around 58% charged on you batteries also. How long should it take to bring them back up to 85% charged with 32.5 amps of charge current? 225 amp hours x 27% = 60.75 amp hours. 60.75 / 32.5 = 1.87 hours to top it back up, assuming your solar charge controller can crank that 32-33 amps for the almost 2 hour time frame.

The voltage of our NMC cells does move a lot more than LFP cells, so to get watts, we need to look at the voltage change as well. Looking at my graph, you can see the voltage in this range is a very linear rise. This makes the math pretty easy. We can just use the average of the top and bottom voltages. 53.2 volts + 56.7 volts = 109.9 / 2 = 54.95 volts. In your case, 54.95 volts x 60.75 amp hours = 3,338 watt hours. Do you think you will be able to cycle that much out and back into your batteries each day for a few days? Then we should be able to see if the BMS will start accepting that as a full charge. Using the same voltage range, that would be 54.95 volts x 193.85 amp hours = 10,652 watt hours. I actually move a little more since I start at 52.5 volts.
 
earlier in the day it was a constant charge, then once full it was taking breaks. ( And yes, the total current is going into all 3 batteries. )
I'm sure the batteries are getting full charge,, it's the SOC on the BMS's that isn't reading accurately..
also.. am i seeing right.. it seems strange that all 3 batteries reach full charge all at the exact time Except for the SOC reading...
to me it seems like the SCC is turning off at the set voltage before the BMS's
this is also puzzling me.. Maybe i should disconnect 1 string and charge it manually with my bench charger and see what the BMS does..

I finally got 3 more BMS's on order for the other 3 strings of batteries...they might be here before christmas...lol
 
to me it seems like the SCC is turning off at the set voltage before the BMS's
this is also puzzling me..
I would hope your SCC normally stops charging (?and not “turning off”?) before the BMS hits high voltage disconnect. The BMS is a last chance failsafe, NOT intended for use as a charge controller.
Your BMS high and low voltage disconnects should be set at the battery safety limits:2.5V lvd, 3.65V hvd. Conservative values might be 2.8V and 3.6V…your level of comfort may vary.
 
yes.. my desired settings are low is 3.1vpc 2.8vpc is supposed to be the lowest,
And high 4.1vpc.. max is 4.2v, but i keep it slightly lower..
Yes, i meant to say stops charging.
 
Back
Top