diy solar

diy solar

Strange EVE 280Ah cell behaviour during charging

Jan Gils

New Member
Joined
Oct 18, 2020
Messages
24
I have 24 EVE cells up and running for the last 8 months in 3 parallel banks of 8 cells in series (24V) with a Batrium BMS. Because I've seen some BMS shutoffs, I'm investigating the cells to check for problems. Yesterday I started charging the battery at around 24% SOC to 80% and noticed that cell 1 had a much higher voltage then the other cells:

Here is a screenshot at 47% SOC. I'm charging with 142A and cell 1 has 3.69V while the other cells are at around 3.30V.

SOC 47perc.png

Then at around 70% SOC, the voltage of cell 1 drops to 3.46, wich is still higher then the others, but not abnormal. Charging is still at 143A.

SOC 70perc.png

I increased the charging rate to around 200A, but nothing really changed. After reaching around 80% SOC I stopped charging and all cells went to the same voltage 3.33V

SOC 82perc.png

Why does cell 1 behaves like this? Is it a bad cell I need to replace, or could something else be going on (bad connection, bad voltage measurement)? I'm trying to understand whats going on as it trips the BMS. The cell voltage of cell 1 went up to 3.82V and then started dropping to 3.46V. While the cell was a high voltage, the BMS was balancing the cell with 0.82A as you can see in the first image.

The bank was fully top balanced to start with.
 
A few things going on here and could be a combo of things.
Your Config Review: (terms - be careful ! )
3 EVE 24V/280AH Standard 8S Battery Packs in Parallel within a single Battery Bank. What Cell Grade and are they Matched & Batched ?*
*Matched & Batched are more $ and come with a cycle report of the 3 pass test, These are NOT just matched for IR at standing storage voltage as most vendors do. Please clarify the Grade of Cells and if B or Bulk or Properly Matched & Batched. No need to chase ghosts.

Pack Config for 8S, Cell 1 is typically NEG and Cell 8 is POS

1) When charging Cell 1 will always be a little higher which is not unusual and this should level down as the battery pack charges.
This is disconcerting at first blush and can be problematic in certain conditions.

2) Loose connections would also present Heat Issues (Big Clue) and if the terminals etc are hot on Cell 1 then ensure the terminals are perfectly clean, that there are no burrs or ridges on busbar or lugs (pending on your assembly) and make sure it's all tight. You'd be surprised that oil/wax and ridges & burrs CAN cause all sorts of nastiness that's a PITA to chase down (everyone checks it last).

3) High Rate of Charge (Current/Amps) when cells are reaching above 80%+. LFP does NOT like that ! CC High Amp to 80-85% and then CV charging to finish them off. Once at 80% cell IR starts to increase and the cells will start reducing the Amps they take. A "RUNNER" cell will certainly standout at this point (usually 3.400 Volts +/- a bit) at which point it can run high. Lowering Amps sent to pack can mitigate this a "bit" and allow the others to continue charging to level up (over simplistic).

REMEMBER: There is a Difference between "Working Voltage Range" (3.000-3.400 Vpc) and "Allowable Voltage Range" (2.500-3.650 Vpc) The Amp Hours the cells are rated for comes from the "Working Range" hence why Grade A Cells will test out a MINIMUM of 10% Over label rating when discharged from 3.650 to 2.500 to ensure the Correct AH is delivered from the working range. Bulk / B-Grade cells will rarely ever be able to deliver full AH from the working range, typically 10% less than stated rating.

Now this can get into a very long set of instructions so instead I will suggest you do the following so we can shorten this up and "deal with it" rather than covering all possibilities.

- Check the connections to Cell-1 to ensure they are good. correct as needed.
- Allow packs to discharge a bit as normal to around 60% or so, so you can do a respectable charge run.
- Set the Charge Voltage to 27.8V (3.475Vpc)
-- Your EndAmps/TailCurrent is 14A for 280AH packs (280AH X 0.05 = 14A) always based on Highest AH pack within a battery bank.
-- You can increase that "a little" up to about 18A. and can add 2A per pack in bank. This triggers the transition from CC to CV (float)
- Set your FLOAT Voltage 27.4V (3.425Vpc) (Float is CV, constant current, variable voltage)
- Charge the Bank as normal and observe (Log if possible) once cells go OVER 3.300Vpc (*1)
- Observe the AMPS being taken by the Packs within the bank, as packs reach target voltage @ Full Amp Charge resistance will start to reduce Amps taken, take note of the amps @ voltage (you should see "little" difference between packs as they all reach the same voltage)
-- You MAY even see a pack disconnect due to a runner, that's ok for now. Let the whole cycle go through (might be tough to watch) till charging is done and system enters into Float Mode... Allow to float at least 2 Hours and stop charge. Batteries WILL Settle (normal) and cells within will balance up within the pack in an hour or so (No heavy or crazy loads during this please, better to have NONE actually)
! TAKE NOTE OF Highest & Lowest Cells at End of CC Charging & at End of Float
--- After of hour of Idle Settling, all packs should be very very close together at "Terminal Voltage" and all of the cells within all of the packs should also be pretty darn close to each other. Again take note of Cell Voltages,

*!* These High/Low cells are essentially the limiters... They will trigger the Hi & Lo Volt cutoffs and so they create a "boundary".
(*1) A runner cell will likely be higher than that while the rest may be lower... This is due to the IR differential, this is made worse by higher amperage being delivered to cells. Higher Amps = Faster Running.

IMPORTANT:
DO Account for Voltage/Amperage drop from charge source to Battery Pack Terminals and correct for it. You have to focus on what the voltage is at the Battery Terminals and NOT the charger ! Verify with a good reliable DVOM. NOTE, there is ALWAYS a little loss going through Battery Terminal Lugs (they are most often Brass and it's not the greatest, worse the more recycled it is). Brass is HIGHLY recycled, there is likely WW2 Casing Brass in the brass item you bought yesterday at the store ! Each recycle makes it a little whiter, the more "orange" it is, the more pure it is.

CHARGE PROFILES = PITA !
Every Charger be it an SCC or Inverter/Charger or dedicated charger, have different profiles, some great and some well not so much. LFP requires 2-STAGE CC Constant Current & CV Constant Voltage Variable Current. The Transitioning from CC to CV is handled differently with various brands of types of equipment. It can never be as simple as we would like eh ! Makes for lots of Blue Air LOL. So this of course makes it fiddly and difficut to just say "use this setting" because every single setup is different. Even worse is some companies use terminology that is far less than clear when describing how their profiles work and how they are staged.

The above info will help to further Diagnose what is happening within the packs and the bank overall.
Please be very clear as with what Grade & Level of cells you purchased and do understand that a video with a Voltage & IR Testers such as a YR1035 Hand Held battery tester is NOT MATCHED ! That is ONLY Voltage & IR Matched at standing storage voltage ! The VENDORS PLAY ON THE WORDS / TRANSLATION factor to excuse this obscene behaviour.
PS: If you are Active Balancing and can set parameters for that, start Active Balancing at 3.300Vpc "During Charge" and allow for 20mv differential.

Hope this helps, Good Luck.
Report back, I'm tagging this thread to watch it.
 
Sorry, did read the book.....But I call bad connection and at 3.65 volts why didn't the BMS shutdown the charging?
 
I would get a voltmeter and see if the readings are correct. Looks like probably blockmons. Blockmons can go bad. Maybe swap that one with one on another cell.
 
I would get a voltmeter and see if the readings are correct. Looks like probably blockmons. Blockmons can go bad. Maybe swap that one with one on another cell.
what are blockmons??
 
Thanks for all the information. I will do some more testing and report back. Things progress slow as the battery bank is built into a sailboat, on which I'm living with my wife and kids while we're sailing around the world. Currently we're in Martinique and have been underway for 7 months.
  • I have an Flir infrared on board and will look at the connections.
  • The cells have M8 studs and bars are custom made out of 40x8mm copper profile. The copper quality is rated and approved for bus bars. There should be no burs or grease as everything was cleaned thouroughly.
  • Charger is a Victron Quattro 24V 120A and a Skylla-i 100A 24V. Charge profile is set at 27,8V for bulk.
  • I will change the blockmon with a spare and see if that's an issue
  • I also have a good fluke and manually measure + measure the currents between the 3 banks. All cabling is same length to avoid unbalances, but will check again with the Fluke amp measurement.
  • As for the 3.65V, I changed the BMS shutoff to 3,85 to still be able to keep charging and see how far voltage increased.
  • I have voltage sense cables to the chargers, so voltage drop is taken into account. Cables sizes are 120mm2
  • Cells are A grade, but I don't think they were batched and matched.

Whend doing some tests today, I noticed that at 100A charge, the voltage remains pretty stable, but increasing to 150-200A the voltage immediatly increases to 3,75. Lowering the charge rate to around 100A it drops to 3,57V.

I'm starting to think the blockmon is damaged. I have a hard time understanding that the voltage of the cell dropped at higher SOC.
 
Next time you are charging and you see cell1 spike in voltage- test that cell with your Fluke DMM.

Remember something is fibbing to you - the hard job is figuring out what it is - my guess the blockmon#1. Swap it with a different one and see what happens.
 
One thing to measure would using a volt meter between the stud top and the bus bar on the same cell.
I have found a few problem that was as there was a decent size (50 mv) there and should be only 4 or 5.
 
Some feedback. I measured everything, and checked contact issues or voltage drops, but everything was OK. So I changed the cellmon and the issue is resolved.

I noticed some small voltage differences between the Fluke and the cellmons, so I'll contact Batrium to go through the recalibration steps or define some voltage offsets. At the same time, I also reduced the balancing amps to 1A and put a temp restriction of 55°C in place. This should preserve the cellmons better for the future.
 
Back
Top