Higher discharge load on end cells?

cwah

100 MW
Joined
Jul 24, 2011
Messages
4,256
Location
Between paris and london
Hello,

Recently I've decided to plug my cell-log to my bike, and have an extension allowing me to see each cells discharge load while I'm cycling.

Something strange occurred to me:
2015-01-18%2001.16.18.jpg

2015-01-18%2001.16.44.jpg


The first and last cell of my pack are sagging more than my other cells. All other cells are sagging on a similar way but not the first and last. Furthermore, once fully charged, all cell end up on the same voltage. But during discharge, first and last go a bit lower (due to higher sag).


So, what are the possibilities?
1. My first and last row of cell have lower capacity? (unlikely)
2. The pack construction created more load on the first and last row of cells?
3. It's a normal behaviour, so maybe I should add 1 more cell in parallel for the first and last row of cells?
 
It could be temperature related the end cells will stay a lot cooler and thus likely to sag more when its cold.

The only other thing is that the end cells probably have less compression on them than the ones sandwiched in the middle.
 
these are 18650 cells, so compression wouldn't be an issue at all. I'm wondering if it's a natural behaviour, maybe strain is naturally higher on the end?

I noticed on my other packs usually the end cell are usually less aligned to the others...
 
From my reading here, it appears after standing a while the end cells will pick up again. I'm very aware of this migratory effect people are reporting, but I have nothing solid to offer on the subject. It is not the celllogs though, RC chargers also report it.

Perhaps the most logical explanation is that they never fully charged due to an odd charging method being used. Maybe the end cells were monitored through the main leads not balance wires. Leads that were loaded up, causing a volts drop, that the charger took to be voltage at the cell, not across it's wiring.
 
Couple data points. on makita packs from dr bass, or old nicad packs from tools, it is often the last cell or first cell in the string that fail.... we have talked about the shrink wrap damaging the outer cells on hobbyking lipo packs. i would expect the center cells to fail if it was thermally related to overheating. overcooling, sure it will affect the sag alot.
 
isn't a solution to add a couple more cells at the end to manage the load? I don't want to have faster degradation on my first and last row of cells
 
Prepare to do some mental gymnastics: http://endless-sphere.com/forums/viewtopic.php?f=14&t=62494
 
it could be the cell wiring too

if the connection between each parallel call isn't exactly the same length

and the connection between each parallel pack exactly the same length

the change in resistance can make some packs/cells charge/discharge faster than others



we don't normally see it on ebike batteries because we fully charge so often, but it's a pretty normal phenomenon with big off grid battery packs where it's less common to fully charge / balance them
 
This is the exact same issue I have with my last pack. Cells one and two will be lower. After riding all last summer it came to this after every full drain, cell one would be 2.8v, cell two about 3v and the rest about 3.2v. It would be the same voltage thru about 3.5v so I didn't worry about it since it would stay the same in the voltage that counts. I did notice after leaving my cell log plugged in all winter that it will drain cells one and two which is explained in the thread linked below. I used the same cells to build a higher voltage pack so I am curious if the same thing will happen or not.


thread http://endless-sphere.com/forums/viewtopic.php?f=14&t=20142
 
Having read the thread and attached threads, I see no answer but nobody has explored the 'magnets' angle. Power flow and magnetic fields go hand in hand, and in the case of a series string the fields around the middle of the pack are different to those around the end. Presuming they are arranged is series physically. Not enough pics of packs displaying this odd behaviour exist to explore this idea further. Least not with my level of interest. However, the middle cells have fields by their sides they don't want to cross, so are in a higher flux than the end cells who's fields can stretch out further. This could correlate with users finding the end two cells fall off progressively. Surely using external fields would influence the current flow within an individual cell. So the result could be the weaker field towards the outside of a pack gives an easier route for current, so with a cell rolled up like a swiss roll.........

Oh... swiss roll. I'm outta here. Tea and cake 1, battery thread nil.
 
My guess would be there is more voltage drop in the wiring causing the difference more than the cells themselves.
If the CellLog or voltmeter attached directly the cell terminals independent of the charge/discharge path wires, the readings should be relatively free of wiring resistance effects.
 
The end cells are probably cooler and thus produce less voltage under load.

I've seen significant effect on Panasonic cells in regards to temperature. Even 4°F difference and they will put out measurably more voltage.
That's all your seeing. The load is the same.
 
Hello Nelson. Nice scientific approach. Unfortunately I have seen no pictures of apparently defective packs, just talk. I imagine most are home made though. Proper packs look after themselves, It is only home made packs that are not built like professional one's that get tested. Only where people strip commercial packs at the end of their life do we see proper packs getting tested.

My RC lipo pack is of course fully automated. I have on one occasion done a manual examination and didn't observe this problem. My lowest measurements were somewhat central to my pack. My pack is actually 3 4s bricks side by side with a 20mm gap between them all. 35mm gap above. Heat can be ignored.


I'm really not sure what I would test first. Nothing solid has emerged. No single argument without a fair counter argument. Only magnets, because nobody could be bothered lol
 
Punx0r said:
Prepare to do some mental gymnastics: http://endless-sphere.com/forums/viewtopic.php?f=14&t=62494
A lot of posts on there say the current is the same across all connections no matter where the battery is.
I wont argue against this theory but I have been thinking to solve the problem it flies in the face of this theory.
On my 18650 pack I am planing to add a extra set of 4P on the positive end of my pack so the last set is actually 8P, the way I look at it is the accumulative power and current of all these cells linked up in series is going through the last set of cells so why not make its life easier and have an extra set of paralleled cells there so it ends up 8P, all those connections together with a fatter end lead will just alleviate the stresses.

I noticed on my old Turnigy lipo packs that the main red positive wire is considerably thicker then the negative wire.
 
Back
Top