diy solar

diy solar

Battery/BMS Issues

P

PaulAtreides

New Member
Joined
Mar 8, 2021
Messages
37
Good morning everyone......I'm having some issues with my battery/BMS and was hoping to get some feedback from you to help me troubleshoot. My battery is a 280Ah DIY made from Eve cells (I think I assembled it the last week of March) and the BMS is the Overkill 4S 120A. I immediately did a top balance by putting the cells in parallel at a constant 3.65V. After a week of charging, the cells were STILL drawing amps, but only about .5A. At that point, I decided they were probably done balancing, so I removed the bench charger and let them sit in parallel for a day.

After that, I put them in series and connected the BMS. The first attached pic shows the battery after that first assembly. Everything looks good, right? So I thinking cool, I've got a battery. I put it in my travel trailer about a week later after doing some re-wiring to relocate the battery from the tongue and everything's good.

Yesterday morning, after some use, I check the BMS and everything's still good (battery dropped to 99%). Then, towards the end of the day, it says my battery is down to 54% but there's no way it could have dropped this low from a few LED lights and a stereo. I then connect the trailer to shore power (my house) and can see that the battery is charging, but it's only drawing about 10amps. I let it go and slowly over the course of an hour or so, it drops down to .5 amps. Doesn't make sense, right? I look through the BMS app and see Capacity Reset and decided to press it. The SOC instantly jumps to 73% (up from about 55%). The battery charges a little longer, but stays at 73%, so I hit Capacity Reset again and it jumps to about 93% (see second attached pic). FYI, I used the generic LiFePO4 profile in the BMS settings.

Now, it shows 93% and has stopped charging. I'm totally scratching my head here. Any ideas where I should start looking to troubleshoot?
 

Attachments

  • Initial, After Balancing.jpg
    Initial, After Balancing.jpg
    64.6 KB · Views: 38
  • Current.jpg
    Current.jpg
    60.6 KB · Views: 39
PaulAtreides said:
Good morning everyone......I'm having some issues with my battery/BMS and was hoping to get some feedback from you to help me troubleshoot. My battery is a 280Ah DIY made from Eve cells (I think I assembled it the last week of March) and the BMS is the Overkill 4S 120A. I immediately did a top balance by putting the cells in parallel at a constant 3.65V. After a week of charging, the cells were STILL drawing amps, but only about .5A. At that point, I decided they were probably done balancing, so I removed the bench charger and let them sit in parallel for a day.

After that, I put them in series and connected the BMS. The first attached pic shows the battery after that first assembly. Everything looks good, right? So I thinking cool, I've got a battery. I put it in my travel trailer about a week later after doing some re-wiring to relocate the battery from the tongue and everything's good.

Yesterday morning, after some use, I check the BMS and everything's still good (battery dropped to 99%). Then, towards the end of the day, it says my battery is down to 54% but there's no way it could have dropped this low from a few LED lights and a stereo. I then connect the trailer to shore power (my house) and can see that the battery is charging, but it's only drawing about 10amps. I let it go and slowly over the course of an hour or so, it drops down to .5 amps. Doesn't make sense, right? I look through the BMS app and see Capacity Reset and decided to press it. The SOC instantly jumps to 73% (up from about 55%). The battery charges a little longer, but stays at 73%, so I hit Capacity Reset again and it jumps to about 93% (see second attached pic). FYI, I used the generic LiFePO4 profile in the BMS settings.

Now, it shows 93% and has stopped charging. I'm totally scratching my head here. Any ideas where I should start looking to troubleshoot?
Click to expand...
Did you capacity test and draw until a cell hits 2.5v? I would charge until you get a cell to 3.65v, hit capacity reset, then discharge until one hits 2.5v
 
The generic LFP profile is probably expecting 3.65 volts to be 100% charged. Your average voltage is about 3.45 which is fine, and is actually about 95% charged, so it is close. I don't see this as a problem. It is reporting that you have 245 AH left. That is likely pretty accurate. After a few cycles, you may see the BMS adapt a little to the charge cycle. My JK BMS is a little different, so I am not sure if they will do the same thing, but after the third cycle or so, when the absorb stops, mine resets to 100% charged, even though I am at 57.2 volts, instead of the full 58.8 volts. I only run down to 50-60% and it reports that pretty accurate, so I have not bothered to reduce my stated capacity from 360 to the 324 that is the 90% I am only charging up to.
 
Zwy said:
Did you capacity test and draw until a cell hits 2.5v? I would charge until you get a cell to 3.65v, hit capacity reset, then discharge until one hits 2.5v
Click to expand...
I didn't do a capacity test. I don't have an inverter, so I can't do one efficiently and quickly.
 
GXMnow said:
The generic LFP profile is probably expecting 3.65 volts to be 100% charged. Your average voltage is about 3.45 which is fine, and is actually about 95% charged, so it is close. I don't see this as a problem. It is reporting that you have 245 AH left. That is likely pretty accurate. After a few cycles, you may see the BMS adapt a little to the charge cycle. My JK BMS is a little different, so I am not sure if they will do the same thing, but after the third cycle or so, when the absorb stops, mine resets to 100% charged, even though I am at 57.2 volts, instead of the full 58.8 volts. I only run down to 50-60% and it reports that pretty accurate, so I have not bothered to reduce my stated capacity from 360 to the 324 that is the 90% I am only charging up to.
Click to expand...
I'm not sure I trust the numbers I'm seeing. What got me spooked was the sudden drop from 99% to 55% with very little usage. I literally turned on my LED lights in the morning and shut them off around 6pm. The stereo was on for a few hours. Definitely not enough to drop it that far. I don't have my panels hooked up yet, but I'm interested to see if they'll take the batteries up to 100%.
 
Turn on some load and watch the current draw and make sur it agrees with the load you are running. How many watts does all of your LED lighting draw? If it totals 48 watts, that would be 4 amps on 12 volts. Does the BMS show 4 amps, or whatever your LED's should be taking? Watch the amp hours "Remaining" window, far left above the charging switch on the app. If it is pulling 4 amps, that should drop one amp every 15 minutes. If you can pull 10 amps, then it should drop one every 6 minutes. With it set to 280 AH as the total, it should also drop 1% after each 2.8 amp hours used.

If your LED lights draw 4 amps, and you had them on for 12 hours, that would be 48 Amp Hours, that alone is 17% of the battery. If it is 10 amps of LED's, then 12 hours is 120 Amp Hours, or 43%, that would get you down to near 55% remaining. The stereo could also pull a fair amount of power. My son's small Kenwood head unit pulls about 2 or 3 amps average, and cranked up loud I saw it peak at 7-8 amps. The subwoofer amp in his car pulls up to 16 amps with loud music, it is fused at 40 amps. I don't think your energy usage is out of line. Check for loose connections and watch the battery voltage and current and see if the numbers make sense. 280 AH is a lot of battery. That would be like having 4 normal 100 ah deep cycle lead acid batteries, but at 12 volts, it is still just 3.36 KWH of storage.
 
PaulAtreides said:
The SOC instantly jumps to 73% (up from about 55%). The battery charges a little longer, but stays at 73%, so I hit Capacity Reset again and it jumps to about 93% (see second attached pic). FYI, I used the generic LiFePO4 profile in the BMS settings.
Click to expand...
Looking at your OP, the pic shows the cell "graphs" as about 50% full at 3.5v. This should be 100% full (i actually use the voltage where the cells settle within an hour as MY 100% value).

So for me, 3.35v is my settle voltage (13.4v) is what i call 100%. Anything else is mostly surface charge and will fluctuate significantly with minor charge or load.

You did not mention what charger you have or what your charge settings are (other than default for LiFePO4).
 
GXMnow said:
Turn on some load and watch the current draw and make sur it agrees with the load you are running. How many watts does all of your LED lighting draw? If it totals 48 watts, that would be 4 amps on 12 volts. Does the BMS show 4 amps, or whatever your LED's should be taking? Watch the amp hours "Remaining" window, far left above the charging switch on the app. If it is pulling 4 amps, that should drop one amp every 15 minutes. If you can pull 10 amps, then it should drop one every 6 minutes. With it set to 280 AH as the total, it should also drop 1% after each 2.8 amp hours used.

If your LED lights draw 4 amps, and you had them on for 12 hours, that would be 48 Amp Hours, that alone is 17% of the battery. If it is 10 amps of LED's, then 12 hours is 120 Amp Hours, or 43%, that would get you down to near 55% remaining. The stereo could also pull a fair amount of power. My son's small Kenwood head unit pulls about 2 or 3 amps average, and cranked up loud I saw it peak at 7-8 amps. The subwoofer amp in his car pulls up to 16 amps with loud music, it is fused at 40 amps. I don't think your energy usage is out of line. Check for loose connections and watch the battery voltage and current and see if the numbers make sense. 280 AH is a lot of battery. That would be like having 4 normal 100 ah deep cycle lead acid batteries, but at 12 volts, it is still just 3.36 KWH of storage.
Click to expand...

280 is definitely a lot of battery, that why the drop to 55% wasn't making sense as the usage was minimal.

The LEDs draw 2A. Not sure about the stereo. I'll turn on both tonight and see what the BMS app says and check it with my clamp meter as well to confirm. Thanks!
 
MisterSandals said:
Looking at your OP, the pic shows the cell "graphs" as about 50% full at 3.5v. This should be 100% full (i actually use the voltage where the cells settle within an hour as MY 100% value).

So for me, 3.35v is my settle voltage (13.4v) is what i call 100%. Anything else is mostly surface charge and will fluctuate significantly with minor charge or load.

You did not mention what charger you have or what your charge settings are (other than default for LiFePO4).
Click to expand...
I noticed that as well about the graphs. Not too concerned about them though. I guess I could change the settings if it really bothered me.

The charger is the WFCO 8900 series that came with my trailer. Not sure which model yet, but it's a 35A min.
 
The 45% drop in charge appears to be merely the surface charge and does not reflect the actual capacity of your battery(s).

If you look at these SoC charts, you will see that it differs significantly from what your BMS is telling you.


The top chart shows 99.5% full at 3.45v per cell. Your BMS considers everything above that as 45% of your total capacity.

And here is a similar thread:
diysolarforum.com

SOC Help

I have a question I hoped everyone could help with about SOC. I found a chart on the forum that displayed Lifepo4 SOC, but it contradicts what my BMS says so I hoped for clarity. We had no sun today in Colorado and my batteries voltage is 13.1 to 13.09 when my heater kicks on. The SOC chart on...
diysolarforum.com diysolarforum.com
 
Most decent "Smart" BMS units that show a battery capacity percentage use a process called coulomb counting. It actually measures the current flowing in and out of the battery and tris to count how many amp hours have gone in and came back out. But if you start with a charged battery before connecting it, the BMS has no idea how many amp hours actually went in. In the case of my JK BMS, it watches the charge cycle. When the charge current tapers off to under about 5 amps it assumes the battery has reached full charge. It then resets the AH remaining to the value I set for the battery capacity. Then when current starts being drawn from the battery, it will start subtracting. Pull 10 amps for 6 minutes, and it will drop one amp hour off of the remaining. Pull 60 amps for one minute, and it drops another amp hour. Until it sees a few cycles, it might not reset to full properly. Mine took about 3 charge cycles before it started setting to 100% at my lower voltage setting. I am actually only charging to 88% or so, and on the first charge cycle it did show about 90%. But after it saw the charge current fall off and stop at only 57 volts, it now sees that as 100% full. Each BMS has it's own way of determining what 100% full actually is. You said yours has a "capacity reset". Does the manual explain what it does and when to activate it? If you hit it when the battery is at the wrong state of charge, it might throw off the coulomb counting.

Are there any adjustments on the WFCO 8900 series 35 amp charger? Is it designed for a 4S LFP pack, or was it intended for a Deep Cycle lead acid? What was the voltage at the cells when the current was falling below 10 amps? and when it was down to just 0.5 amps? That sounds like the charger went into a CV (constant voltage) mode where it reduces current to hold the cells at a fixed absorb voltage. While it was charging down to just 0.5 amps, what did the BMS status screen show for current, voltage, AH remaining, and SoC? if you measure each cell with a meter, does it read pretty close to the cell voltages shown in the BMS app?
 
GXMnow said:
You said yours has a "capacity reset". Does the manual explain what it does and when to activate it? If you hit it when the battery is at the wrong state of charge, it might throw off the coulomb counting.
Click to expand...

Nothing in the manual, but found this on the forum: FUNCTION SETTING -> RESET CAPACITY is useful if remaining capacity reported has drifted to the point of being inaccurate. This can happen with very low current being missed, measuring accuracy cumulative errors, calibration errors, repeated partial state of charge cycling etc.

Two things - Maybe I just haven't cycled it enough? and, Maybe the capacity reset fixed the issue with the huge SOC drop since it came back up to "normal" after I did the reset?

Per WFCO, the 8900 series is compatible with lithium. There's no switch. But they do make converters with the LiFePO4/lead acid switch.

1618259435669.png

What I showed in the second pic (93% SOC) is basically the same info as when it was charging at 0.5A. Nothing changed, the amps just went down to 0A. I need to double check voltages tonight.
 
All you need to do is fully discharge the battery to when the lowest cell displays close to 2.5 volts, then fully charge it until the highest cell is close to 3.65 volts. If you want to make sure the LVD and HVD of the BMS are working then go all the way.

You might have to do this a couple of times to get accurate readings. The BMS is not smart enough to know the true capacity of your battery until you have done this. Once you have done, this the BMS readings should be fairly accurate.
 
PaulAtreides said:
I didn't do a capacity test. I don't have an inverter, so I can't do one efficiently and quickly.
Click to expand...
Who said you need an inverter? Just draw it down using whatever you have. I used a blower motor for mine, sure it took some time.
 
Update: looks like I've got a parasitic draw. When I thought it went down to 55% suddenly, I actually think it did so over time. It just happened so quickly it appeared to be sudden. I charged the battery last night via the trailer converter and unplugged it before I went to bed. When I woke up, it was down to 58% and continuing to drop. Something is using up the power. I put my clamp meter on the positive cable and it's at .17A. So I'm thinking it would have to be on the battery itself, or on the negative side right? Would a bad chassis ground cause a draw like this?

Edit: also, the BMS is not registering any outgoing amps.
 
Last edited:
280ah / .17a = 1647h to completely discharge at this rate.

.17a will scrub the surface charge off from 14.4v down towards nominal/settle voltage (~13.5v) overnight.
 
Dropping 42% overnight is huge. How many hours would you say? Let's use about 8 for this calculation.
That would be losing 5% per hour. 280 x .05 = 14 amps of draw. 0.17 amps would take 600 hours to burn off 42% of 280 amp hours.
There has to be something else going on here. A bad ground can't cause current draw like that.
What were the cell voltages when you turned off the charger? What were the cell voltages when it said 58% remaining? Did the BMS ever show any current draw? What does the BMS show for "time left" and "remaining"?

Most BMS units break the negative side. So the positive side should go to your fuse and cutoff switch then to all of your loads. The negative side of the battery should go to the BMS B- lead, and noting else (besides the BMS balance lead). The output of the BMS on a common port would be the P- lead which then goes to your ground and - buss bar for all of the loads' negative sides. For all practical purposes, the BMS is part of the battery. I have seen one install where the user also ran a heavy negative cable from the battery negative to the frame ground. This basically eliminated any chance for the BMS to measure the current or interrupt the load or charge current. He thought the system was working fine until he left something on overnight and it killed the battery since the BMS had no way of shutting it down.

What I think you need to do is isolate the battery system from the trailer and test to make sure the battery is performing correctly on it's own. If it is not, we need to trace that first. If it is working correctly, then we have to trace what is going on in the trailer. I would start with just the battery cells and BMS and nothing else connected. Connect just the charger and bring the battery to your full charge voltage. For a 4S LFP, I would suggest bringing it to 4 x 3.4 = 13.6 volts if you can adjust the absorb voltage on your charger. Once it is charge, what does the BMS read? If it did not go to full, is it at least 85% or above? If not, do the capacity reset. This should be your new full charge setting. Disconnect the charger, so it is just the battery and the BMS. The BMS will pull a small amount of current, but it should not be enough to pull the battery down for many days. Before my inverter/charger arrived, my battery bank sat a month with the BMS powered, and it only dropped about 1-2% out of my 360 amp hours. So let that sit overnight and see if the voltages drop. LFP cells will settle down a little, so you may see some drop, but the capacity remaining should not change much. If this is looking good, apply a small load. What do you have available? Would it be hard to just run a wire to the LED lights that you said pull 2 amps? That would be a good start. Enough to show on the BMS current meter, but not so much to kill it quick. Running the 2 amp of LEDs, the capacity meter should drop 1 amp hour every 30 minutes. In 84 minutes, it would drop 1%. Yes, 1.4 hours. At 2 amps, it would take 70 hours to use 50% of a 280 AH battery bank. That is 2.9 days of your 2 amp LED lighting. And that is just 50% of the capacity. So you should be able to just check it every few hours with the LEDs on and see if the BMS is even close to these numbers. Just check it every 3 hours or so and see if it is losing about 3%. If it is going faster, you have a problem.
 
MisterSandals said:
280ah / .17a = 1647h to completely discharge at this rate.

.17a will scrub the surface charge off from 14.4v down towards nominal/settle voltage (~13.5v) overnight.
Click to expand...
It never got above 13.7V with the full charge. When I woke up, it was 13.3V. That .17A is probably just the maintenance draw for the stereo, fridge, etc. It's got to be something else :unsure:
 
You must log in or register to reply here.

Similar threads

K
Uneven discharge causing issues
2
Replies
35
Views
2K
karamazov
K
S
EEL Battery Box Balancing Issues
Replies
7
Views
449
Surferdudemi
S
G
Jk bms balancing issues
2
Replies
34
Views
2K
hogback
H
H
Latest Mason 280L issues
2
Replies
29
Views
1K
Neil H
N
G
Communication protocol issues JK BMS and Solax inverter
Replies
4
Views
715
giuseppe211
G
Back
Top