Self-Discharge UPDATE On 30Q 141 Cells

eMark

100 kW
Joined
Nov 2, 2019
Messages
1,163
Location
Minne-apple, USA
A couple reasons i went with 30Q 15A even with my "conservative" etriking ...

1) With it's low IR wanted to see if i could get 500 cycles from my 10S3P 30Q 141 pack over 3-4 years by charging to only 4.1V and discharging to only 3.5V. Hopefully, maintaining enough capacity for a 5-6 mile round trip with minimal pedal assist through 500 cycles.
(2) Also hoping my TLC "conservative" (non-raw performance) would negate any eventual high self-discharge tendency that has been reported by others using the 30Q "136" cells.

My 10S3P is wired with two 5S balance leads for cell logger alarms. When split into two 5S3P sub-packs for bottom balancing before straight charging the two 5S3P packs paralleled together for 1.5A (0.5C) straight charging. Then reconnecting again as one 10S3P 30Q 141 pack.

This was the pack's first winter rest having accumulated 170 charge/discharge cycles from mid-March thru mid-November of 2020. So, I'd hoped that my TLC would've yielded favorable good news when recently testing the 30Q 141 cells for any self-discharge.

On January 2nd I charged (balanced) all 30 cells with each cell at 3.75V. Sadly my "conservative" etriking with the newer (2019) Samsung 30Q "141" cells still resulted in some of the 30Q "141" cells suffering from self-discharge. Five of the cells suffered from what i'd consider high self-discharge. This self-discharge was first evident after three days on 1/5/21 then more so on 1/10/21, and even more so on 1/15/21. Of the thirty 10S3P Samsung 30Q 141 cells twenty-one still remained at 3.74-3.75V on 1/15/21. However, nine of the thirty cells suffered from self-discharge ranging from 3.72V to as low as 3.43V on 1/15/21. Following are the self-discharge voltages of nine cells, five of which i'd consider high self-discharge.

5S1 Sub-Pack
1s1p/3.66V; 2s3p/3.43V; 4s2p/3.70V
5S2 Sub-Pack
1s1p/3.72V; 1s2p/3.68V; 4s2p/3.72V; 5s2p/3.70V; 3s3p/3.62V; 5s3p/3.57V

On average i recharged my 10S3P 30Q 141 pack every 1½ days so any self-discharge wasn't apparent or problematic for my etriking enjoyment.
 
You don't have any pictures. How do you parallel individual cells using multiple "packs" of serially connected cells? Or maybe I have this wrong?

The observed phenomenon of high self discharge may be true for many cells, and further may go largely unreported. Especially if the charge process has an effective balancing routine to rebalance the pack.

I've seen this in marine LFP installations. The original philosophy (circa 2012) was balancing wasn't required. And it wasn't with "new" properly behaving cells. But that self discharge, in paralleled groups of cells, with no routine balancing eventually shows up once any self discharge sets in. That cell or cells become "low" and are constrained by the high cells calling for end of charge. This self discharge isn't by itself in most cases isn't a reason to trash the pack...or even the cell. But does require some degree of balancing.
 
BlueSeas said:
This self discharge isn't by itself in most cases isn't a reason to trash the pack...or even the cell. But does require some degree of balancing.

I had to trash batteries where i used 30Q cells because of self discharge and know about others who had similar issues with that cell.

One pack i was trying to manually balance by charging up all groups which had lower voltage, but after 2-3 month it was quasi unusable again because the BMS did cut off whereas controller was still showing 30-50% SOC, which means that one group went below 2,5V where the highest still had above 3,6V or so.
I also found out that self dicharge depends on the actual voltage of the cell and it could be that lets say at 4,1V it will discharge very quick to 3,8V for instance and then almost stops.

It were always those packs built with 30Q where this happend and since i noticed that i don't use this cell any longer.

I am sure that even the largest BMS with highest balancing current available would not have been able to keep this pack usable since it would mean that you would have to charge such battery every few days, beacause if you let it sit for some weeks or month over the winter for example, it would run so bad out of balance that even a BMS with very high balancing current would not be able to equal it out in acceptable time.

Self discharge means energy will be turned into heat. The more quick that happens (the "bigger" the internal short), the more heat will be produced in those cells and the higher the risk of battery fire, so better don't take the risk and replace the pack.
 
BlueSeas said:
You don't have any pictures. How do you parallel individual cells using multiple "packs" of serially connected cells? Or maybe I have this wrong?
My two 5s3p sub-packs are paralleled for bottom balancing and straight charging with my 6S balance charger. The parallel cells in any p-group are always attempting to balance one another ... even as the cells in a p-group go astray with age. The 2nd photo with all the bus bars removed is for testing each individual cell for any self-discharge. In the 3rd photo i put a towel over the pack as extra protection during my etrike outing. Will report back at end of month with on-going self-discharge cell voltages of the two 5S3P packs now that they are in their winter's rest.

BlueSeas said:
The observed phenomenon of high self discharge may be true for many cells, and further may go largely unreported.
Have any of us ever heard ANY complaints about the VTC5 or VTC6 packs having high self-discharge? Maybe that's why those cells cost $2-$3 more than 30Q cells. The 30Q is one of Samsung's highest production cells having other usage besides DIY ebike pack builds. Even though Samsung is aware of the self-discharge problem they are making a ton of money so as one might say "leave well enough alone". They still do OK even if the 30Q cells are $2-$3 less than the VTC5 or VtC6. One supplier currently is selling 30Q 136 cells for $3.99

BlueSeas said:
This self discharge isn't by itself in most cases isn't a reason to trash the pack...or even the cell. But does require some degree of balancing.
That's one reason why i decided to be a stickler about conservative etriking discharge, bottom balancing, charging at 0.5C, 4.1V charge to 3.5-3.4V discharge and forego the use of a BMS. Most DIYers may not even be aware of the self-discharge as most packs are recharged the same day or within a couple days. Even with my worst cell its "high" self-discharge is mild compared to even a conservative ebiking 1 hour battery discharge outing.

Even so i will be replacing as many as ten cells in my 10S3P pack in February, but that's another story as the ten 30Q 141 cells i ordered turned out to have a "6 KH1T" code on the pink tube label instead of a "141" code ... even though the supplier's webpage showed a "141" code on the label. The "6 KH1T" is not the same as the code as on the metal can. I've both emailed and called this reputable supplier. Customer service won't even tell me why they sent me 30Q "6 KHIT" cells instead of 30Q "141" cells.
 

Attachments

  • Testing Voltage Of Individual Cells.jpg
    Testing Voltage Of Individual Cells.jpg
    129.3 KB · Views: 2,719
I think I understand the pack build, charging and the testing you accomplished now.

I think the voltage here was a typo: "Of the thirty 10S3P Samsung 30Q 141 cells twenty-one still remained at 4.74-4.75V on 1/15/21". You meant 3.74-3.75.

Did you take a reading 18-24 hours after removing the parallel busbars? This would be a better baseline for comparison. If the RC charger supports it, probably even better to discharge to target voltage. Either charge or discharge, it should be a very low rate to eliminate charge differences in the parallel strings. 100mA or less. And I would have probably used more like 4.05V as a starting position. My 4P20S NCR18650GA pack will sit at 4.05V for a month (my normal charge). It only discharges evenly and very slowly because of the BMS load.

Do you have any idea what the current draw while riding is? Or maximum number of minutes at normal speed from full charge to full discharge? I'm guessing you are close to a 1C rate, or 1 hour. Regardless of all the "peak" ratings, that's not a light load for a small cell. My 72V pack is good for about 3 hours, sure I hit peaks (40 mph), but on average for 15-20 mph bike path riding. My pack does have more than 2X the energy and I'm only on 2 wheels. Thinking you might want more paralleled cells or more voltage to maximize battery life.

I'm not able to comment on specific cells like the 30Q vs. others. But I have seen this issue with large scale LFP banks at much lower discharge rates.
 
COOL looking bike you have ... what's your top speed ?

madin88 said:
It were always those packs built with 30Q where this happend and since i noticed that i don't use this cell any longer.
Apparently QC didn't correct the high self-discharge problem with the newer 30Q "141" 30Q released to suppliers in 2019.

madin88 said:
I am sure that even the largest BMS with highest balancing current available would not have been able to keep this pack usable since it would mean that you would have to charge such battery every few days, because if you let it sit for some weeks or month over the winter for example, it would run so bad out of balance that even a BMS with very high balancing current would not be able to equal it out in acceptable time.
BMSs are almost worthless when it comes to balancing as the process of resistance bleeding is VERY SLOW.

It's my understanding that you need to leave the charger plugged in after the green light comes on in order for the BMS to even do a half-decent job of balancing the p-group cells within 25-30mO of each other. Supposedly this should be done for 12-16 hours once a month and later every two weeks (with frequent battery use) as the pack ages. WHAT'S your take on the balancing usefulness of a BMS ?
 
BlueSeas said:
I think the voltage here was a typo: "Of the thirty 10S3P Samsung 30Q 141 cells twenty-one still remained at 4.74-4.75V on 1/15/21". You meant 3.74-3.75.
YES! Went back an corrected typo ... Thanks!
BlueSeas said:
Did you take a reading 18-24 hours after removing the parallel busbars? This would be a better baseline for comparison.
AGREE! Waited 1 day for p-group cells to settle on their own voltage. Even so digital dc meter only reads in hundreds of a volt so some of the cells could have been close to 3.747 volts and others close to 3.753 volts. So for the purpose of the 30Q self-discharge whether some of the 21 cells still read 3.75V and others 3.74V after two weeks was not relevant to my "high" self-discharge testing of the 30Q 141 cells.
BlueSeas said:
My 4P20S NCR18650GA pack will sit at 4.05V for a month (my normal charge). It only discharges evenly and very slowly because of the BMS load.
That's because your battery pack's cells aren't 30Qs :wink: ... Hoped the newer 30Q "141" cells would be an improvement :(
BlueSeas said:
Do you have any idea what the current draw while riding is?
Averages between 5-8 amps most of the time. Controller will pull up to 19 amps when going up a steep hill without pedal assist, but usually closer to 10-15 amps for only 15-30 seconds with pedal assist (checked with Watt Meter). So no one can say that my 10S3P 30Q VRUZEND battery was abused. Never been bumped, dropped, mishandled, or otherwise mistreated.

Actually, my 10S3P 30Q "141' is perfect for my conservative etriking with charging at 0.5C to 4.1V and discharging to 3.5V-3.3V (resting) with most 30Q outings only 5-6 miles round trip. Also have two UPP (UnitPackPower) batteries (12.5A & 10A). Take one or both along for outings up to 20+ miles, but only 4-5 times a year when checking on my EABL Trail (14 bluebird nest boxes). I've been known on the spur of the moment to just take off and roam new ground (many bike pathways in and around the metro). If necessary will stop an recharge one of the UPPs at a McDs, area library and other place while enjoying a relaxing meal with my chromebook.

Three more photos ...

First Photo: My customized etirke with several modifications including the Plano Container with pad-locked lid. Place to slide in one of the UPP batteries below Plano and room to carry the other UPP in Plano besides other things (e.g. snack, Home Depot harware or Menards, EABL nest box, tools, camera, etc). Also can fasten my Bogen tripod to the Plano lid. Carry a snack or book in the rear attached tote bag.

Second Photo: Pink tube label "6 KH1T" of recent 30Q purchase. Was suppose to receive "141" 30Q cells as shown on webpage ... https://www.imrbatteries.com/samsung-30q-18650-3000mah-15a-battery/ ... no explanation given by customer service as to why the change ??

Third Photo: Same codes on all the ten metal 30Q cans wrapped with the "6 KH1T" pink tube labels. According to page 15 of the 30Q datasheet "K" stands for 2020 production as "E" is 2014.

Hopefully will report back in July (after 100 more cycles on my DIY 10S3P VRUZEND V2.1 with 30Q "141 / 6 KH1T" cells. Hopefully none of the 9-10 "6 KH1T" cells that will be replacing the current self-discharging "141" cells won't suffer from any self-discharge.

Happy New Year and Happy Trails to All :thumb:
 

Attachments

  • Samsung 30Q Pink Tube Label.jpg
    Samsung 30Q Pink Tube Label.jpg
    88.7 KB · Views: 2,634
eMark said:
It's my understanding that you need to leave the charger plugged in after the green light comes on in order for the BMS to even do a half-decent job of balancing the p-group cells within 25-30mO of each other. Supposedly this should be done for 12-16 hours once a month and later every two weeks (with frequent battery use) as the pack ages. WHAT'S your take on the balancing usefulness of a BMS ?
AFAIK most chargers cut off at 1/10 of the rated current and will not start charging again if you do not reconnect. You would need a Labor PSU set to very low current (similar to balancing current of BMS) and leave it connected.

I would say it depends on the battery. For those packs with 30Q it was totally useless because self discharge current was higher as the balancing current of BMS.
It might keep your battery usable if the drift is really small, but otherwise i see it only as a gimmic because in the very most cases if there appeard a problem with selfdiscahrge it was to much of an issue.
 
madin88 said:
AFAIK most chargers cut off at 1/10 of the rated current and will not start charging again if you do not reconnect. You would need a Labor PSU set to very low current (similar to balancing current of BMS) and leave it connected.

I would say it depends on the battery. For those packs with 30Q it was totally useless because self discharge current was higher as the balancing current of BMS.

It might keep your battery usable if the drift is really small, but otherwise i see it only as a gimmic because in the very most cases if there appeard a problem with selfdiscahrge it was to much of an issue.
AGREE! Was referring to normal packs left plugged in 12-16 hrs (once a month) after green charge light comes on. This company spokesman further suggested that as any pack ages increase this procedure to bi-monthly when battery is used daily. Does it make sense to do this extra top balancing to achieve p-groups within 25-30mΩ of each other? Is it true that a BMS will only continue to top balance (resistance discharge bleeding) when the charger is left plugged in after the green light comes on and not when when charger is unplugged ?

FWIW, I partially disassembled the one 5S3P sub-pack to remove the worst cell on 1/16/21. Then charged it at 0.5C to FULL, discharged it at 0.4 amps to 3.3V and charged to 3.80V (resting voltage). Since then self-discharge has again been about 0.15V daily this time. Will be replacing ten of the "141" cells suffering from self-discharge with ten "6 KH1T" 30Q cells in each of the ten p-groups. I do have a few extra 30Q "141" cells. One of the ten p-groups has two cells in its p-group suffering from self-discharge so will replace the two cells in that p-group with one of the "6 KH1T" 30Q cells and another "141" cell.

That's the advantage of a reasonably priced VRUZEND V2.1 kit for ease of assembly and disassembly. That's why I chose the V2.1 kit for my experimental testing of my 30Q "141" 10S3P DIY pack knowing that the "141" cells may still have a problem with self-discharge :(

Will UPDATE on-going results again in July (after 100 more cycles) assuming everything works out OK. Will be interesting to find out if more of the other 20 "141" cells starts showing signs of self-discharge and whether any of the "6 KH1T" cells suffer from self-discharge after 100 cycles (on average charge every 1 1/2 days.
 
have you tried "desulfating" the 30Q to see if thats all thats required to "fix" the self discharging?

it wouldnt be a big deal to desulfate a pack every 100 cycles or 75 cycles to restore it

40T shares the same characteristics as the 30Q and es member rg12 mentioned he had a 40T cell self discharge

desulfating link
https://endless-sphere.com/forums/viewtopic.php?f=14&t=109601
 
Sad I just got my 20s 6p Samsung 40t cells it has a BT BMS not turned in yet. I thought they were the cat's meow.
And order a NESE 14s, 8p kit and was planning on using the Samsung 30 q cell. I guess I should have a change of mind
 
3qty 30Q packs here

2 used cold freezing weather

both dead less than year

most cells exhibit self discharge


1 pack mild weather

3 yrs going strong

no measurable self discharge


coincidence? possibly

but now refrain from

below 50F charge
 
madin88 said:
30Q selfdischarge (test from docware):

https://endless-sphere.com/forums/v...e2259830a0147f423c4c907f9f&start=500#p1613728

from 100% soc to 50% (3,7V) within 32h :lowbatt: :lowbatt:

goatman said:
40T shares the same characteristics as the 30Q and es member rg12 mentioned he had a 40T cell self discharge

Would be bummer if 40T has similar issue. Can you share more infos? maybe i missed it but could not find anything in the topic you linked.

rg12 and 40T
https://endless-sphere.com/forums/viewtopic.php?f=14&t=107895

samsung 48x will go 3000 cycles with 1c(5amp) charge/discharge in "unwindy conditions" to 70% original capacity (4.15vto3.0v)

above 50f was mentioned as too cold
i run my tests at 100f

be nice if emark could desulfate one of those bad cells he has to see if it solves the self discharge
 
goatman said:
have you tried "desulfating" the 30Q to see if thats all thats required to "fix" the self discharging?
Being that there's no evidence (testing) that that's all that's required to heal/repair the "high" self-discharge problem my plan is to replace one cell in each p-group with a new 30Q cell. IMR sent me ten 30Q "6 KH1T" cells instead of "141" 30Q cells. Will be interesting to see if any of these new "6 KH1T" cells suffer from self-discharge after 100 more cycles (by July) as well as any of the other 20 cells that currently don't show any signs of self-discharge. Don't use my etrike from mid-November until nice weather in March/April. Only used my VRUZEND 10S3P 30Q (non-BMS) experimental pack in temps between 50-45 degrees F three times last year.

Have always bottom balanced my VRUZEND 10S3P before straight charging with little or no top balancing required with p-groups within 10-15mΩ of each other (don't use BMS). Usual discharge is from 4.1V to 3.5-3.4V. Only discharged to controller cut-off voltage of 3.2V three times since DIY build to check bounce-back resting pack voltages and recharge mAh capacity. Was hoping my non-abusive (TLC) would eliminate any 30Q high self-discharge. All "high drain" cells are probably more prone to a higher than normal self-discharge as the battery ages with 30Q apparently the most prone.

Will post all of the thirty cell voltages in February beginning with the resting charged voltages at 3.74-3.75V on 1/2, then the degree of self-discharge on 1/15 and on-going self-discharge on 1/31. This self-discharge varied from moderate to high self-discharge. Have tried what may be "desulfating" with the one "high" self-discharge cell, but once a cell suffers from "high" self discharge its of little use (IMO) ... unless there's research to prove desulfating is more than just a short-term fix.
goatman said:
be nice if emark could desulfate one of those bad cells he has to see if it solves the self discharge
Am more than willing to give it a thorough test with the one cell that suffers the most from the "high" self-discharge.
Please provide detailed instructions for your recommended procedure (KIS), and will begin ASAP assuming my testing apparatus is sufficient for the task :thumb:
 
i just do a 0.6amp discharge to 2.5v then charge it back to 4.2v then do the 0.6amp discharge to 2.5v and charge it back up and its done

heres the link
https://endless-sphere.com/forums/viewtopic.php?f=14&t=109601
 
goatman said:
i just do a 0.6amp discharge to 2.5v then charge it back to 4.2v then do the 0.6amp discharge to 2.5v and charge it back up and its done

heres the link
https://endless-sphere.com/forums/viewtopic.php?f=14&t=109601
OK, will finish your above (KIS) procedure today (Monday, 1/25 by 8:00pm CDT) and track results through 1/31. Will begin with resting cell voltage at 10:00pm on 1/25 (after your "desulfating" KIS above procedure). Then record amount of any self-discharge beginning on 1/26 at 8:00pm, 1/27 at 8:00pm, 1/28 at 8:00pm, 1/29 at 8:00pm, 1/30 at 8:00pm, and 1/31 at 8:00pm. Will report back with resulting voltages on 2/1/21 (around 2:00pm CDT).

Will also record mAh capacity. Did a similar test a few days ago, but only discharged to 3.35V with the capacity near 2,400 mAh, but not sure off accuracy of testing equipment ... www.amazon.com/gp/product/B07H32WVTF/ref=ppx_yo_dt_b_asin_title_o02_s00?ie=UTF8&psc=1 ... Only discharges at 0.4 amps (can only use bay #4 for discharging at 0.40). It's mAh charging capacity is very close to same mAh rating as both of my 2S-6S balance chargers.
 
eMark said:
goatman said:
i just do a 0.6amp discharge to 2.5v then charge it back to 4.2v then do the 0.6amp discharge to 2.5v and charge it back up and its done
heres the link
https://endless-sphere.com/forums/viewtopic.php?f=14&t=109601
I pulled that one cell suffering the most from high self-discharge (Sub Pack 1 - 2S3P) on 1/16 and tried a similar low amp discharge followed by low amp charge without any noticeable improvement. Have about finished with your double down [desulfating] procedure discharging at 0.40 amps (MiBOXER discharge rate) and charging to FULL at 0.60 amps (time consuming procedure). Will post results of 2nd double down attempt Friday afternoon and attach two other charts (1/10 and 1/15) showing on-going self-discharge of nine cells. Triple checking to make sure the charts (1/10, 1/15 and then 1/31 don't have any typos). Was going to post the 1/10 chart today, but just noticed a typo so will post it Friday with 48 hr self-discharge minimizing of second double down [?desulfating] procedure.

Needed to use additional technique for further discharging so bounce back resting voltage was at your 2.50v suggestion as my MiBOXER lowest discharge resulted in bounce back of 3.3v resting (after 1 hour). Discharged with homemade light bulb device (at 1 amp) continuing from MiBOXER 3.3v resting until light bulbs were off. Then after an hour the bounce back voltage was 2.51v (your recommended discharge rate). Then charged at 0.60 amps until FULL. Then discharged 2nd time at 0.40 amps until 3.3 volts with MiBOXER. Then continued with homemade light bulb discharger resulting in bounce back of 2.50v resting after one hour. Charged again at 0.60v until FULL resulting in 3,394mAh.

Charging to FULL (5:33 hrs.) from 2.50v with result 3,394mAh the second time was encouraging. IR was 35mΩ at beginning of charge, down to 29mΩ at 95% full, and 26mΩ from 98% to FULL. Will track any further self-discharge over 24 - 48 hrs and post result Friday. Attached is the first 1/5 chart showing varying degree of nine cells self-discharge since 1/2. The 10S3P p-groups of the two 5S3P sub-packs were balanced between 3.74v and 3.75V on 1/2/21. Will attach the corrected 1/10 chart and 1/15 chart Friday showing the ongoing varying degree of self-discharge of the Sub-Packs 1&2. Was going to post the 1/10 chart today, but noticed a typo.

We will know by late Friday afternoon if that [desulfating] procedure lessened the previous degree of "high" self-discharge. Welcome any further suggestions that could minimize (heal/repair) current "high" self-discharge of sub-pack 1 - 2S3P cell.
____________________________
Typo edit: sub pack 1 - 2S3P was 3.68v on 1/5 ... was 3.56v on 1/10
 
i dont know if you can run your pack down to 2.5v

it wouldnt be hard for me because i have a phaserunner controller

your 3p so if i put 1.8amps of electronic freewheeling and the lvc to 2.5v (10s would be 25v)

all i would have to do is block up the rear wheel(motor) touch my throttle and the wheel would just spin at 1.8 amps until the the lvc was reached.

dont know if you could do the same with PAS?

i just discharged to 2.5v and charged right away, i think the battery i tested would bounce back to 2.85v within 2 minutes, i never tried to get right to 2.5v like you did

i very much appreciate your efforts, its nice to get answers whether it works or not :bigthumb:
 
goatman said:
i dont know if you can run your pack down to 2.5v

i very much appreciate your efforts, its nice to get answers whether it works or not :bigthumb:
The cut-off voltage of my 20 amp Controller is 3.2v. Have only pulled 10S3P cells down to 3.2v three times since March 2020 ... average discharge from 4.1v to 3.4v over 170 charge/discharge cycles.

Are you possibly wondering if LION high drain energy cells (for. e-power tools) are intentionally designed to be drained to 2.5v and charged to FULL to achieve the best performance including cycle life ? Have been wondering when someone (serious or tongue-in-cheek) might post a reply that with my conservative etriking i should've used M36 or MJ1 cells with my DIY VRUZEND kit (probably true).

My rationale for selecting the 30Q "141" cells was for experimental testing ... to see if conservative etriking v. raw performance ebiking would hopefully minimize high self-discharge.

The VRUZEND V2.1 is NOT party to high self-discharge of a high energy cell (e.g. 30Q, VTC6, etc.). Will have some final comments about VRUZEND V2.1 (inherent measured mΩ resistances), experimental testing and incorporating the ten 30Q "6 KH1T" cells in the two 5S3P sub packs.

(see typo edit at bottom of my previous post)
 
madin88 said:
I had to trash batteries where i used 30Q cells because of self discharge and know about others who had similar issues with that cell.

Self discharge means energy will be turned into heat. The more quick that happens (the "bigger" the internal short), the more heat will be produced in those cells and the higher the risk of battery fire, so better don't take the risk and replace the pack.
That kind of SEVERE discharge (internal short) is not what most associate with high self-discharge of a high energy LI-ion cell. The 30Q self-discharge you reference ("the higher the risk of battery fire") IMO would more likely result from a combination of a cheap/faulty BMS, pack abuse and possibly human error (pack "abuse" is relative to user interpretation/application). A 30Q internal short is not the kind of high self-discharge one associates with 30Q high self-discharge.

My previous chart (1/2-1/5) and the two attached charts (1/10 and 1/15) represent the degree of "high" self-discharge with 30Q as i take was the following post by BlueSeas ... where he's coming from when he says ...
BlueSeas said:
The observed phenomenon of high self discharge may be true for many cells, and further may go largely unreported. Especially if the charge process has an effective balancing routine to rebalance the pack.
IMO, You are so correct, but it depends on what is considered "high" in comparison to some "high" benchmark ... https://batteryuniversity.com/learn/article/elevating_self_discharge
goatman said:
have you tried "desulfating" the 30Q to see if thats all thats required to "fix" the self discharging?

it wouldnt be a big deal to desulfate a pack every 100 cycles or 75 cycles to restore it
That's assuming you have evidence that your procedure will "fix" the kind of high self-discharge we associate with a high energy cell; especially like that of 30Q.

Actually performed your suggested "desulfating" procedure three times. The third time discharged at 0.40 amps and charged at 0.40 amps. Decided to stop at 3.90v for closer comparison with a self-discharge from 3.74v to 3.68 volts over the first three days (see previous 1/2 to 1/5 chart). So far after 36 hours (1 1/2 days) the self-discharge is now 3.87v (0.03v self-discharge), so no noticeable improvement. See previous 1/5/21 chart with self-discharge from 3.74v to 3.68v (0.06v discharge). See typo edit chart correction in previous post (3.68v not 3.56v)
 
so it was the cell from sub pack 1, 2s3p that you desulfated? and its sitting at 3.87v right now, be interesting to see how much it drops in a week or 2 week period compared to the jan.1 to jan.15 loss
 
goatman said:
i just do a 0.6amp discharge to 2.5v then charge it back to 4.2v then do the 0.6amp discharge to 2.5v and charge it back up and its done

heres the link
https://endless-sphere.com/forums/viewtopic.php?f=14&t=109601

Pretty interesting what you found out :thumb:

If this desulfating would help to heal 30Q from selfdischarge, would be great, but i have my doubts since most BMS will cut off at 2,5V and then such groups should quasi heal themselves each cycle which seems to not be the case.
 
999zip999 said:
If fully charged and balance does the pack keep full capacity in use.
Only charge to 4.1v (bottom balance before straight charging at 0.5C (1.5 amps) 95% of the time with 170 charge/discharge cycles since mid-March thru mid-November 2020.

Decided to discharge that cell with the highest self-discharge below 2.5V (last week) so resting voltage (after 1 hr) was at 2.5v. Was surprised how low i had to discharge that cell with it still able to bounce back to 2.51v after 1 hr. Here are the mAh capacities at the following voltages charging at 0.60 amps (0.2C) beginning at resting voltage of 2.51v to FULL ...

2,400 mAh at 3.99v (71%)
2,644 mAh at 4.08v (85%)
3,000 mAh at 4.15v (92%)
3,340 mAh at 4.20v (98%)
3,394 mAh at 4.20v (FULL)

During charging mΩ IR was 35mΩ at 75%, 32mΩ at 92%, and 26mΩ from 98% to FULL. I compared the IR of this, the highest self-discharge cell, to six other "141" cells i've had in storage for 13 months and to the "6 KH1T" 30Q cells i recently received. There wasn't more than a few mΩs difference between the high self-discharge cell and these other stored "141" and new "6 KH1T" 30Q cells.

Have removed the ten cells with self-discharge from the two 5S3P sub packs. Made up a 5S2P pack out of these ten self-discharge cells for further experimenting this month. Will continue experimenting to see if (?desultating?) self-discharge can be minimized ... welcome any more suggestions. That cell with the highest self-discharge was discharged to 3.44 resting volts before assembling it with other nine cells of the 5S2P pack with voltages on 1/31/21 as shown on 1/31 chart posted tomorrow afternoon.

This is the pairing of the ten cells in the 5S2P experimental pack ...

1S2P... 2S2P.. 3S2P.. 4S2P.. 5S2P
3.65v, 3.62v, 3.53v, 3.53v, 3.56v
3.44v, 3.73v, 3.70v, 3,70v, 3.70v

Will post the 1/31/21 chart tomorrow showing the placement of the ten cells in 10S3P that are listed above and in attached 5S2P photo. Decided not to post the 1/22/21 chart as little variance between the 1/22 chart 1/31 chart. Undecided on three possible options when i reassemble the two 5S3P sub packs later this month. Appreciate which option you guys would favor when i list three possible options tommorow.
 
Back
Top