Thread Rating:
  • 0 Vote(s) - 0 Average
  • 1
  • 2
  • 3
  • 4
  • 5
What good is rePackr when Opus is so inconsistent?
#1
I've got over 2000 tested usable cells and am getting ready to start building my 14s??p packs. All were capacity tested on Opus chargers at 1A. Before committing to a build method I decided to make a small 12V (4s20p) pack just to get a handle on spot welding, fuses, connectors, and getting a battery with balanced capacity banks. So I pulled 20 cellls from the top end of my cull pile for the project. These tested initially in the Opus at between 1939 and 2141 for an average of 2131 mAh or about 40 Ah for the 4s20p pack.



So I pasted the capacities in repackr and got my pack distribution. Just for kicks, I decided to run each of the cells through the Opius again, but this time at the default 500 mA charge rate. The difference in capacities was surprising and unexpected. Most (3/4) of the cells tested significantly lower on the second test. The average difference was 68 mAh and the largest drop was over 500 mAh. High rated Samsung and LG cells that had tested poorly on the first test tended to do better on the second. Cells testing close to rated capacities usually tested worse.




So after that second test, it looks like I'll end up with some pretty unbalanced packs:









Not real happy with the consistency from these Opus chargers. What do you all think? Should I re-run the repackr with the new lower capacities, or stick with the distribution from the first run? I'm hoping all the effort to catalog keeper cells for my big packs wasn't wasted but after this I'm not so sure.
Reply
#2
That doesn't mean the opus was that far off (the 50mAh could be, tho). The one that was 500mAh off could be a bad cell and it's degrading.

Also, testing differences between 1000mA and 500mA can make a difference on the capacities. Though, they are usually higher with a slower current drain. It's possible that maybe you aren't getting a solid connection on the contacts in the charger. Perhaps there could be some residue on the cells. I recommend using rubbing alcohol on the cells to make sure there isn't any glue or thin film of some kind of oil. And yeah, I've found some cells to have a thin film on them. Spot welding burns right through the film, but a standard metal-metal contact could be kept apart by it.

Another option is that the holders aren't tight enough. Perhaps the spring could be tightened. Open the unit up and shorten the spring.

Usually the Opus tests consistently about 5% lower than the actual capacity. This is usually fine as most of us use the results as a guideline, rather than an actual hard and true value.
Proceed with caution. Knowledge is Power! Literally! Cool 
Knowledge is Power; Absolute Knowledge is Absolutely Shocking!
Certified 18650 Cell Reclamation Technician

Please come join in general chit-chat and randomness at Discord Chat (channels: general, 3d-printing, linux&coding, 18650, humor, ...)
(this chat is not directly affiliated with SecondLifeStorage; VALID email req'd)
Reply
#3
I'm nearing 7,000 cells over the last 2 years using 3 Opus(s) doing 500ma discharge tests - writing the ah on the cell. As I distribute cells into a pack I add them up till I get between 260,000 to 263,000. Then I put packs in 14s and hook them using Batrium BMS/monitor.

Out of the 42packs online so far - most have been right there at 260,000 but several (25%) tested lower like 240,000-250,000 ... so I simply added 4 or 6 or 10 cells to bring them up and wa la.

Another approach is to buy an iCharger X6 or something like that and test the pack as a whole - and adjust as needed to meet your target to match the other packs before putting it on line.

In any case, one key (for me) was to make the physical pack large enough to have extra (empty) slots so I could add up to 20cells after the base pack creation. Its difficult to create a large pack with 'exact' ah by adding up individual cells... the Opus, testing, cells can be a little fuzzy.... so I'd recommend planning a step/process to test and adjust the pack once its all put together and catch it there rather than focusing too much on 50ah here and there at the cell level.
bogptrsn likes this post
Reply
#4
I still do capacity and self discharge tests, but that's just to weed out the bad cells.
I simply assign cells to the 14 packs in a round-robin way, like when dealing cards for 14 players. This ensures (more of less) that cells from one source does not end up concentrated in one pack. Otherwise it's basically totally randomly selected cells between 1700~2400mAh.
With 100 cells in each pack, they end up with remarkably similar capacities, though of course not perfect.

To make adjustments afterwards easier, I've placed 3x 1200mAh cells in every pack at an easily accessible location. If a pack has a relatively low capacity, I replace one or more of those cells with a 2400mAh.
Modular PowerShelf using 3D printed packs.  40kWh and growing.
https://secondlifestorage.com/t-AJW22-s-...PowerShelf
Reply
#5
(10-10-2019, 03:06 AM)cowpen Wrote: I've got over 2000 tested usable cells 
That is a great accomplishment 

Quote:So I pulled 20 cells from the top end of my cull pile for the project. These tested initially in the Opus at between 1939 and 2141 for an average of 2131 mAh or about 40 Ah for the 4s20p pack. 
So those 20 cells----- what can you  tell me about them?
Do you have any "true"  IR information.

Quote:So I pasted the capacities in repackr and got my pack distribution. Just for kicks, I decided to run each of the cells through the Opius again, but this time at the default 500 mA charge rate. The difference in capacities was surprising and unexpected. Most (3/4) of the cells tested significantly lower on the second test. The average difference was 68 mAh and the largest drop was over 500 mAh.

68 mAh is nothing. Over 500mAh ok that is half and Ah and I would consider that something although not a deal breaker.
The funny thing is that all of these tests do not include IR.
Your results are perfectly predictable with a proper IR measuring device.

Quote:High rated Samsung and LG cells that had tested poorly on the first test tended to do better on the second.

High rated ?? 2600mAh, 2800mAh, 3000mAh, ICR chemistry, INR chemistry,etc. There is a need to know as a lot of the Samsungs and for that matter LG cells have different charge and discharge rates.
Oh and did I mention IR.

Quote: Cells testing close to rated capacities usually tested worse.
Tested worse???  Again information is required. 2000mAh cell testing at 1800mAh which is 90% capacity is not bad in my book but a 3000mAh cell testing at 1800mAh as far as I am concerned is a dead cell with 60% capacity.

Quote:So after that second test, it looks like I'll end up with some pretty unbalanced packs:
Not real happy with the consistency from these Opus chargers. 
Personally I am also not tickled with the OPUS brand of chargers but as far as consistency is concerned  I think they are relatively good when it come to that. 3% max deviation in 5 tests (granted all done at the same charge and discharge parameters) but nevertheless not bad.


I have tons of these tests  on my google drive.

Quote:What do you all think?
Asking that question"What do you all think" is like opening a can of worms.

I can tell you what I think!
I think you should invest in a YR1035+ IR meter and check the IR of each of your tested cells and anything over 75mΩ IR should not be used.
I would only keep cells that are at least 80% SOH. I.E. a 2800mAh rated cell with a minimum tested capacity of 2240 mAh or a 2200mAh rated cell with a minimum tested capacity of 1760mAh.

Keep your testing at the same level throughout your campaign. Test all cells on similar chargers, as in only OPUS, only Foxnovo ( may favorite for simplicity and accuracy) only Liitokala, only Zanflare for consistency. It is not so much about actual mAh results as all of these chargers have somewhat of an elevated result.
All have anywhere from a 3% to 10% inflated results except the Foxnovo (again my favorite). It is that you are using a baseline when using only 1 type of charger/tester.

Keep your Cell IR ≤75mΩ and your SOH ≥80% and you will do just fine.
That is if you are using ICR chemistry.


Wolf
If 18 X 650 = 2200+mAh then we have power! 
May all your Cells have an IR of 75mΩ or less Smile
Last count as of 8/7/2019
Total Number of Cells Recorded and processed                 6149
Total Cells required for PowIRwall                                   2856
Total Cells ≥2200mAh, ≥80%, ≥35mΩ, ≤75mΩ, ≥4.12V   2760
For Info Google Drive
Not your average Wolf       
            Cool
Reply
#6
Thats why you mix Them in 100mAh bins... There is No use for a normal powerwall to go closer.... 3++ years in on My Main powerwall and not a single issue and the oldest are many more. I have many many 1000s of cells.

But i also bin all cells that doesnt work on first attempt. I never give Them a 2nd chance.


Doing. Aproper test at lets say 1a ensure ir is Good enough for 500ma continues current.
ajw22 likes this post
The Ultimate DIY Solar and build place
YouTube / Forum system setup / My webpage  Diy Tech & Repairs

Current: 10kW Mpp Hybrid | 4kW PIP4048 | 2x PCM60x | 83kWh LiFePo4 | 10kWh 14s 18650 |  66*260W Poly
Upcoming: 14S 18650~30kWh | Automatic trip breakers, and alot more
Reply
#7
(10-10-2019, 01:02 PM)daromer Wrote: Doing. Aproper test at lets say 1a ensure ir is Good enough for 500ma continues current.

A proper test at 1A indeed will confirm that the cell has an acceptable IR and that it will perform acceptable at a 500mA draw.
In my case I check for an acceptable IR before I test the cells and eliminate the wasted testing procedure.
To each his own. In the end the results will be the same.
The correlation between a cells SOH and IR VS the cells SOH and proper results on a 1A test.
It is your choice.

I personally chose to check IR first and have very good results with SOH. > 90% success rate
Others choose to test by SOH results  (AKA mAh results) and do just fine with that.
In my case that success rate was < 60%.

Wolf
If 18 X 650 = 2200+mAh then we have power! 
May all your Cells have an IR of 75mΩ or less Smile
Last count as of 8/7/2019
Total Number of Cells Recorded and processed                 6149
Total Cells required for PowIRwall                                   2856
Total Cells ≥2200mAh, ≥80%, ≥35mΩ, ≤75mΩ, ≥4.12V   2760
For Info Google Drive
Not your average Wolf       
            Cool
Reply
#8
Deviation of 50-100 is normal, even up to 150mah is expected for me. For example between my two opuses one is off by.01-.02V so one is always 50-100mah shorter than the other. 500mah is probably other factors such as a self discharge or incomplete charging. It's most likely not the opus.

The key is consistency to get good results. I charge with a imax b6 clone with a caddy of 12 cells at around 500ma each to 4.20V. They *must* sit for a week or more so any self discharge cells are caught. Any under 4.15V is suspect.

I don't use repackr. It will drive me crazy trying to match the numbers to each pack, and again your capacities aren't accurate to begin with. So I follow the method daromer uses, put in buckets of 100mah.

The one thing that makes a good pack is also if you have the luxury, group your packs together. For example my packs are all panasonic/samsung 2600mah combos. I don't mix 2000mah with 3500mah batteries as it creates a wild imbalance making some cells work harder. Also I don't throw in 3500mah batteries that only tested at 2000mah, which means it is at it's end of the life already. i only keep 80% and above of actual capacity cells. That way a year down the road I have less of a chance of some cells dying faster than others.

With all that said, my packs can go for months and keep between 50mV without balancing.
Reply
#9
It's important to remember that batteries do not have an exact capacity. If it tests at 2933mAh the first time, I guarantee you it will not test at 2933mAh the second time. Many things affect capacity, with the biggest factor being temperature both ambient and of the cell.
Reply
#10
Thanks for the responses everyone. I went ahead with building my 4s20p test pack using rePackr with the second capacity test. Spot welded it up and mounted this little $5 "Capacity Controller". It has a built in balance function which I didn't really think would work that well, but it brought the banks within 0.009 V in about an hour after connecting it. Two discharge / charge cycles later and now they're all within 0.004 V. Banks seems to stay very close while cycling.

Feeling a lot more confident with the Opus results now.

Reply


Who read this thread?
47 User(s) read this thread:
Zennen (10-13-2019, 12:56 PM), Jonisingt35r (10-18-2019, 08:24 PM), Nemo (10-11-2019, 01:21 AM), 100kwh-hunter (Yesterday, 04:52 PM), ma3006he (10-11-2019, 07:09 AM), OffGridInTheCity (10-18-2019, 06:35 PM), nickydw (10-10-2019, 10:10 PM), kevinjaye (10-10-2019, 03:49 PM), newone972 (10-16-2019, 08:49 AM), ajw22 (Yesterday, 03:00 AM), LPG (10-10-2019, 08:24 PM), robovind (10-11-2019, 10:30 AM), poolizei (10-10-2019, 09:52 PM), HRD (10-11-2019, 06:26 PM), GeoMac (10-10-2019, 11:55 PM), CrAzYDr1veR (10-10-2019, 08:35 PM), xyoung347x (10-15-2019, 02:50 PM), GOTMLK (10-10-2019, 07:51 AM), BatteryMooch (Yesterday, 09:08 AM), iNeedMoarProtection! (10-10-2019, 06:04 AM), Sleonard2649 (10-10-2019, 04:01 AM), James-w (10-11-2019, 09:29 PM), Wolf (Yesterday, 12:01 PM), cowpen (Yesterday, 11:30 AM), chuckp (10-10-2019, 08:50 PM), Geek (10-11-2019, 12:27 AM), Mazlem (Yesterday, 11:55 AM), watts-on (10-10-2019, 03:56 PM), jdeadman (10-18-2019, 07:19 PM), PAF (10-10-2019, 05:39 AM), Korishan (10-11-2019, 12:19 AM), Majorphill (10-11-2019, 01:19 AM), mike (10-18-2019, 06:44 PM), Sean (10-10-2019, 10:28 PM), mr_hypno (10-18-2019, 03:35 PM), Oz18650 (Today, 03:52 AM), not2bme (10-10-2019, 04:36 PM), Walde (10-18-2019, 10:11 PM), Bubba (10-18-2019, 09:30 PM), Ibiza (10-18-2019, 08:34 PM), jonyjoe505 (Yesterday, 03:00 AM), Crimp Daddy (10-10-2019, 06:24 AM), ARLISS (Yesterday, 01:47 AM), Chablis_m (10-10-2019, 10:38 AM), HughF (10-10-2019, 03:47 PM), bogptrsn (10-11-2019, 02:02 AM), daromer (10-18-2019, 06:59 PM)

Forum Jump:


Users browsing this thread: 1 Guest(s)