52 Volt Battery Issue (14S9P)

Joined
Jun 8, 2024
Messages
10
Location
Tampa, Fl
Hope this lands in the correct sub thread...

A little background. I've balanced several E-scooter batteries with success, but this latest issue is puzzling me.

This battery comes from an Apollo Phantom V2 E-scooter. I recently purchased two as a package knowing the batteries weren't great.
After removing the plug to the BMS, and metering out each cell pack, I found a low of 3.66 all the way up to 4.2. Just as i suspected.
I also use just basic equipment. A desktop charger, and a Fluke meter. I successfully balanced all packs to reach 58.7 volts.

So the problem..

Going for a quick test ride, about 6-8 miles, the battery quickly dropped to the cutoff and left me stranded.
I actually did this twice, using the chassis of each of the identical scooters to rule out hardware outside of the battery.
I was half way expecting that a cell pack or two dropped off and triggered the BMS to shut off, all cell packs are within .02 volts.

The obvious answer was that multiple cells within a pack weren't provided the capacity as they should but voltage is fine.
If this were the case, should i be surprised that the voltage across all 14 packs are identical?

So the question..

What could be the culprit, and what tools should i purchase to isolate the problem? And could it be a bad BMS? I'd doubt that.

Thanks!

Mike
 
The obvious answer was that multiple cells within a pack weren't provided the capacity as they should but voltage is fine.
If this were the case, should i be surprised that the voltage across all 14 packs are identical?
I think your battery has just bit the dust totally, that's all. Or at least enough cells throughout the pack have reached end of life. You say you manually balanced, so yeah, they're identical. But if they're all identically depleted, the voltages will still be same or similar to each other, while still being a bad pack, if the capacity is diminished.

Only way to know for sure is capacity test each parallel group individually, and compare it to specs when new.
 
I think your battery has just bit the dust totally, that's all. Or at least enough cells throughout the pack have reached end of life. You say you manually balanced, so yeah, they're identical. But if they're all identically depleted, the voltages will still be same or similar to each other, while still being a bad pack, if the capacity is diminished.

Only way to know for sure is capacity test each parallel group individually, and compare it to specs when new.
I was leaning towards that as well. And I'd definitely lean on that answer if i had balanced them, came up with equal voltages, then did the test ride to only see the entire pack deplete after a short run. What stumped me however, is that i took readings on each of the 14 packs to find the voltages were all within .02 volts. Then when returning the pack to the charger, it went right back to the topped charge. After testing a second time, the results were identical. Just seemed unlikely that of the 126 cells, they found a way to go bad perfectly across 14 individual cell groups.

Short of pulling all of the nickel strips off to isolate, is there a trick to do it?

Endless-Sphere recommends a load tester, but the link goes to an out of stock tester.
 
Just seemed unlikely that of the 126 cells, they found a way to go bad perfectly across 14 individual cell groups.
Doesn't seem unlikely at all, actually. Seems like even degredation of a well-cared-for, aged pack. If you assume it was well made in the first place, with equally matched cells, then I would expect them to age out at roughly the same time. Anyway, you'd only need 1 particularly poorly aged cell per parallel group to bring down the whole group.


Short of pulling all of the nickel strips off to isolate, is there a trick to do it?
If you can access the nickel strips directly, you can use alligator clips to carefully test the parallel groups without pulling off the nickel.
 
Doesn't seem unlikely at all, actually. Seems like even degredation of a well-cared-for, aged pack. If you assume it was well made in the first place, with equally matched cells, then I would expect them to age out at roughly the same time. Anyway, you'd only need 1 particularly poorly aged cell per parallel group to bring down the whole group.



If you can access the nickel strips directly, you can use alligator clips to carefully test the parallel groups without pulling off the nickel.
Great feedback! I really do appreciate it. The pack actually has great access to all of the nickel strips. I've been testing and balancing the packs through the BMS wires for now. I know that to test the load capacity, I'd need to clip directly to the strips on the parallel pack. Is there a tester you recommend to check the capacity, without removing the strips?
 
You could use a balance charger. You could use a Liitokala or Opus and rig something to run test leads off the charger leads. If they're fully charged, you could use a load tester. I don't have a recommendation for a specific one of those. Whatever you're comfortable with.
 
Sounds like just voltage sag. If you pull a lot of amps out of a battery, the voltage will dip, then come back up once you remove the load. That dip in voltage then triggers the low voltage cutoff on the BMS just as if the battery were actually depleted. Voltage sag is greater for old batteries, mistreated batteries, and cells being used outside their specifications (e.g. cells with a 5A max continuous discharge instead of 30A).

Only fix is A) drawing less amps, e.g program your controller to use 20A max instead of 40A, B) hacking the throttle to not go up to max if you can't do that, C) bypassing your BMS and discharging directly from the pack which is a fire risk, or D) getting a new battery or replacing the cells or adding more cells to each p-group or adding another pack in parallel without a battery blender.
 
Update...

So after a bunch of testing, I still struggle to find the best solution to repair this battery.
Recap, battery stays perfectly balanced, but range is reduced to 1/5 of it's range
4-5 volt sag under load, but doesn't recover well

I've attached a file with my capacity test results.

The battery is a 52 volt, 14S9P, with unknown 2500 mAh batteries
The test was at 15 CC across the 9 parallel battery packs/sections
The test was run with a 3 volt cutoff
Each pack/section started at roughly 4.15-4.17 volts
Each test ran for approx 1:25 and averaged 2405 mAh (divided by 9 to get the individual cell level)
In the attached spreadsheet, I recorded the instant voltage drop at the beginning of the test (only 2.7A per cell)
I also captured the resting voltage at the completion of the test

If the spreadsheet is not readable, I can try to post results in another way.

Thanks for any help that can be provided!
 

Attachments

  • Phantom Battery Capacity Test.xlsx
    11.1 KB · Views: 7
> unknown 2500 mAh

I don't think anything unexpected is happening here. Generic cells that aren't one of the main brands like Samsung, LG, or Panasonic all fail pretty quick. They are just there to get past the 30 day return period after initial sale, certainly not meant to last into the hands of a second hand customer. Don't feel bad, I bought plenty of knockoff Dyson batteries and the like before I learned too.
 
Update...

So after a bunch of testing, I still struggle to find the best solution to repair this battery.
Recap, battery stays perfectly balanced, but range is reduced to 1/5 of it's range
4-5 volt sag under load, but doesn't recover well

I've attached a file with my capacity test results.

The battery is a 52 volt, 14S9P, with unknown 2500 mAh batteries
The test was at 15 CC across the 9 parallel battery packs/sections
The test was run with a 3 volt cutoff
Each pack/section started at roughly 4.15-4.17 volts
Each test ran for approx 1:25 and averaged 2405 mAh (divided by 9 to get the individual cell level)
In the attached spreadsheet, I recorded the instant voltage drop at the beginning of the test (only 2.7A per cell)
I also captured the resting voltage at the completion of the test

If the spreadsheet is not readable, I can try to post results in another way.

Thanks for any help that can be provided!


So the initial internal resistance would work out to be (4.15V-3.83V)/21A=0.32V/21A ~0.015Ohm for 9 cells in parallel, or 9*0.015Ohm = 0.135 Ohm for each cell, if I'm getting my battery math right (which I might not). That's not great. And the cells are down to 1/10th of their nominal capacity. Maybe that's just what it is.
 
Definitely not looking great.

I do plan to purchase and replace the 126 cells with Samsung 25R's to get the reliability back into one of the two second hand scooters. As tested, the range for this Phantom V2 was 40 miles in real world riding, by many sources. My current Apollo Ghost gets about 32 miles. The group I ride with gets an average of 30 miles as well. A new battery with the Samsung cells should give me a good buffer when riding with these other riders.

Thoughts are to dissemble the two batteries completely and individually test each cell. Of the 252 cells, if i end up with half that can be put back together to make a complete 52 volt battery, that would be ideal for the backup Phantom. I'd use it as a spare for friends that don't typically ride. I'd dial down the power and probably still manage a 10-15 mile range scooter.

I have testers, a spot welder, shrink wrap, and nickel strips. Would just add a reliable BMS to make this backup battery. At this point, nothing is a waste of time, I actually enjoy the whole process. Or I'm a glutton for punishment. Not sure which.

I'm seeing some reliable, US based, battery suppliers online. Where do you recommend purchasing the 25R's?
And any suggestions on a decent 14S seperate port BMS?

Thanks as always!
 
Further down the rabbit hole...

So I broke down both 14S9P batteries packs, now in the process of sorting the 252 individual cells. I started by doing an internal resistance test with an RC3563 AC tester. I sorted all batteries under 22 mOhms, I have approx the 126 needed to build a new pack. Average IR reading is 21.63 mOhms. With the cells charged, I ran a capacity test as well. The cell are 2500 mAh cells. The average so far is 2522 mAh.

The plan is to us the pack builder tool to sort them into their 9 packs/sections based on both capacity and internal resistance.

Am I on the right track for this gamble with used cells? I have found a few cells that had very high IR and one the had no voltage whatsoever. Hopping I can get a decent pack out of this effort.

I do have all of the supplies to assemble safely, although I'm still looking for an adequate BT BMS. The scooter runs 2 1200W motors.
 
I would try get some new cells, been here done it bought the t shirt wasted the money i can guarantee that all the effort of making that pack and cost of nickel etc will be broke again in under 6 months as those cells are well aged as they neighbours have failed.

I know it seems cost saving but its counter productive and ends up costing more and possible taint your love of the hobby.

Problem we got is all us folk need a battery thats good but the econmics of the price will shtink ainy really passing on to the bottom feeders like us who buy cells in low numbers, we are bent over a barrel pants down when i can buy and fill a shitty plastic chinnese gas tank for the ptice of a few cells no wonder the ev transition is slow.

Negativity aside, try find a deal on some loose cells that fit yout needs and start assembling packs in time you could even make a little money from it not lose.

I think many of us on hesr started i yours shoes reclaiming cells from packs to make new one and i speak from experience ive had hundreds of cells been tested for IR and load sag organising them into groups and making a pack from them just to fail on my 3rd ride climbing a hill about half of the original rating.

Wish u well on your journey best of luck with your ride.
 
I ran into the same problem with the same scooter Phantom V2.
I had the same results that you had previously posted
I did everything you did, Still no positive answers
I want to know what happened.

I broke down to 126 cells, tested each individually
Capacity - 2600 average
IR 26 m - average
Temp discharging at 300 ma 100 F average at the anode
Volts capable 4.1 v average
I am about to do a simple load test
2.6 amp load test for ? minutes to drop to 3.0 volts this should tell the story of the cell.
compare this value to different known new batteries
knowing what I know now
I could have, done this test by using 26 amps for the group of 9 but I don't
recommend that because if you have a couple of bad cells you might have a remote chance of creating
a thermal runaway!

I will let you know what the result is on my post
"BMS - Battery Management System (Technical) 14s8p 52 volt 25amp"

But you could help me out right now if you have time,
I would like to know what the voltage is across the small red and black wires
of the BMS while it is charging, and what the voltage is across the 7 blue and white
wires while it is charging.

Then I can pot even more results in my findings.
Thank you!
 
Where do you recommend purchasing the 25R's?
And any suggestions on a decent 14S seperate port BMS?

I buy new cells at the 18650 Battery Store.

BMSs are every where including Amazon and Walmart (who'd of thought?). You need to decide if you want expensive or cheap. Do you want Bluetooth with it so you can check cell voltages live? You probably need to do some homework to come up with a BMS you are comfortable with.
 
I've attached a file with my capacity test results.

The battery is a 52 volt, 14S9P, with unknown 2500 mAh batteries
The test was at 15 CC across the 9 parallel battery packs/sections
The test was run with a 3 volt cutoff
Each pack/section started at roughly 4.15-4.17 volts
Each test ran for approx 1:25 and averaged 2405 mAh (divided by 9 to get the individual cell level)
In the attached spreadsheet, I recorded the instant voltage drop at the beginning of the test (only 2.7A per cell)
I also captured the resting voltage at the completion of the test

If the spreadsheet is not readable, I can try to post results in another way.

Thanks for any help that can be provided!

Nice, you attached the Makerhawk across each group and ran 14 separate tests at 15A per group, What stands out to me is the drop from 4.14 to 3.83 upon the start of test, That''s a 4V sag on a 9P battery. I think it should be 2V, That's what we get with old cells.




Starting VoltageSag at Beginning of TestTime (Hour)Total AmpsmAh per CellEnding Voltage
14.153.831:3121.69424103.45
24.153.831:2320.92923253.45
34.153.841:2822.02524473.40
44.163.831:2621.58323983.43
54.163.781:2721.76624183.43
64.173.871:2521.37923753.41
74.153.781:2621.71624133.43
84.163.851:2922.27524753.37
94.183.881:2722.10424563.31
104.173.81:2621.74624163.44
114.193.711:1819.69221883.41
124.183.891:2822.16224623.32
134.183.841:2721.85824293.39
144.183.841:2822.10424563.37
 
Am I on the right track for this gamble with used cells? I have found a few cells that had very high IR and one the had no voltage whatsoever. Hopping I can get a decent pack out of this effort.

I do have all of the supplies to assemble safely, although I'm still looking for an adequate BT BMS. The scooter runs 2 1200W motors.
Out of 252 cells, three or four have failed then. That's pretty bad stats. Having 9P sure covers a lot of fails. If these were 25R's, there would be no fails.

You can go ahead and build a battery, but you already know the existing cells are saggy and tired, They've shown a 1% failure rate which will only go higher. I imagine you've already built it though,

Sure, our time is free and it's fun to experiment if you're so inclined. I worry more about the sinister aspects of batteries though, Cheap cells beget cell failures. Some cell failures are leakers Some leakers become heaters, Heaters eventually become flamers.
 
Back
Top