bms saves battery, now what?

[Living E Caveman]

Theoretically I will be having a bms tell me an old lithium battery is fading, but only 1 cell at a time will fail for a while. And my pack is easy to disassemble and reassemble. What do poeple do when a cell fails, by bms recognition or just testing with a multimeter? Replace some cells?

 

[2007blueprius]

I think it's a bit more complicated than that, I am in the process of dealing with tis issue, I do have 2 fading blocks, each are made out of 4 paralel cells, I am not sure what exactly my problem is but I do belive I may have to change those blocks to say the least, depending on your configuration you may have a different experience, I belive in my case the problem is largely caused by a pourly designed management system,
when rebuilding a battery, you want to have the blocks buit out of simmilar capacity cells and preferably all blocks of simmilar capacity, your total capacity will be determined by the weakest cell.
in my case I am dealing with 20 ah, considering the bad ones when I find them get replaced, they will hopefully still have some capacity, I could use to make something for a different application ( my sons power wheels, or razor )
hope this helps

 

[dnmun]

i was thinking it would be useful for people to record their charging cycles with the cellog 8s and downloading the charging data onto files to use to see if there are profiles during charging that change and if you could graph it and have multiple series or samplings of data over the lifetime of the battery then maybe something would turn up in the data. collecting data during charge keeps it nominalized so it can be formatted for analysis. who knows what's in there.

if you had an an extra channel on the cellog 8S then maybe do some sort of hack to create a voltage on one of the open channels and have that voltage be the shunt voltage on the charging current so the data would be correlated voltage and current. that is the easiest way to keep the current data with the voltage data and record it.

but if you look in the spec sheet for this BMS you will see where they actually can deliver data files to you directly from the BMS:

http://www.bestekpower.com/fuelgasgauge/BMS-D141.html

if you get to the bottom you can see the USB cable adapter for programming it. but if you look at the third to last category of data tha the BMS collects is "self discharge current" and i think that would be a really good parameter to focus on for following battery health.

i sure wish i could do the data com stuff and displays. that would be neat to have real time voltage and state of charge data for each cell.

i am still gonna order one get it for 15S lipo.

 

[2007blueprius]

well said, the sistem I am using in some sense has much of that built in, they are easily/cheaply available used now, since the setup had other weakneses and many owners did not know what to do with it, the dcto dc converter had an act for frying, and if you remember my first post its rather odd how they set up the charging, I'm pretty sure it has a lot to do with the baterypack beeing in the condition I got it, therefore many owners gave up.
however one nice thing about the BMS I have ( althou I would rather it kept the pack properly balanced instead ), it does have a usb port that connects to a display unit which also houses a sim card, I havent's really looked into it but I do get live data as far as how voltag/amp draw, individual cell voltage, some other stuff too there are 4 temperature sensors in the packs which pug into the bms but for some reason it is recomended not to, that would also be displayed and logged, in th eright hands this would be a decent setup for all that tinkering, I think the problem I have it is undersized for my batery pack, and it can not keep them balanced right, the logic is a bit funny too, it would probably do well on 1/4 the pack

 

[dogman dan]

The simple version is, find out what your cells internal resistance is.

The ones that have abnormally high resistance are the ones to replace. If they are all high resistance, then replacing a few cells won't help that much.

Alternatively, capacity testing will generally also pinpoint which cells are trashed.

 

[Living E Caveman]

Thank you all so much. I am getting a cell log, its coming up on my list of must see technology. And yes, since I am using a multi-meter right now to get the feel of my cells and can test resistance very easily, that will be the way I choose to monitor which batteries are failing, together with checking voltage before and after some rides. The weaker batteries are the only ones I want to keep an eye on, like with a cell log.
I will also be using these methods to possibly rebuild used lithium batteries into (hopefully) useable packs. If cheap cells come along.
I knew I wanted to join in these forums, I'm glad I finally did. Endless Sphere has been like college for me, starting a small E-bike business from scratch for over a year.
Its crazy that the most technologically advanced country on earth doesn't HAVE any college courses about ultra light E.V.s? (except in Oregon) I've loved them since the 80's and finally can afford them! DONE! Talk about a reluctant male, what about a reluctant bicyclist? More people must find this out. I think its as natural as going back to being a kid, like when you are living your second childhood and that's just fine.

Anymore input on this will be greatly appreciated.

Thanks everybody! Now go for a joyride somewhere new. get lost! happy summer

 

[jerrypa]

The simple version is, find out what your cells internal resistance is.

The ones that have abnormally high resistance are the ones to replace. If they are all high resistance, then replacing a few cells won't help that much.

Alternatively, capacity testing will generally also pinpoint which cells are trashed.

This help me too. Thank you. Could I measure the batteries resistant with a standard multimeter? Can you say what are typical good values?

Is there easy way to know is the BMS broken?

My cell are rectangular. I think it's not matter if I use other shape of Li-ion battery. Only important thing is that electrics values of new cell are right?

FROM:
http://www.endless-sphere.com/forums/viewtopic.php?f=14&t=50365&p=744887&hilit=+Conhismotor+battery#p744887

I have same kind off 36V battery JL-JD21-10A-36/10-A001 that was bought on 2011.

Do the battery do these when it is overloaded.
- Battery random dies and get back to live when turn the key off and back on? Usualy almost on a hill when max power has taken out.
- Sometimes battery works well after happening but sometimes (after lot load) it start to work in a moment.

Do the battery have own safety circuit that shut down the battery when voltage is too low? I think so.

Battery
http://www.conhismotor.com/ProductShow.asp?id=41

Where i can buy new cells?
Cell Specification 3.7V 2AH
Cell Combination 5-parallel 10-series
Cell Size 18650
Cell Quantity (parallel*series) 50

Its seems like to be warming problem? Because if error comes after lot use it will not work right away but after less use it will.

Edit:
All battery couples voltage was 4V after driving. Perhaps the bms is broken. Where can i find new one?
ps. Fuse isn't at use.

Controller is 250W or 350W)

 

[Hillhater]

The simple version is, find out what your cells internal resistance is.

The ones that have abnormally high resistance are the ones to replace. If they are all high resistance, then replacing a few cells won't help that much.

Alternatively, capacity testing will generally also pinpoint which cells are trashed.

This help me too. Thank you. Could I measure the batteries resistant with a standard multimeter? Can you say what are typical good values?

Well, the MM will be needed, but you cannot just put the MM across a cell and measure its resistance.
you need to measure the voltage "sag" on each cell as a normal discharge load is applied.
Co, use the MM to measure the voltage of the cell ar rest ( no load), then measure the voltage change as the load ( Amps) is applied.
You will also need to measure the load (Amps) accurately at the same time.
Then do a little simple math, ..IR = V/I
Where V = voltage "sag" ( difference between the resting V and the loaded V
..and I = the current (amps) under load.
This will help understand the process..
http://endless-sphere.com/forums/viewtopic.php?f=14&t=48739

 

[jerrypa]

So as i thought that i have to measure the voltages under load. I think that i don't have to calculate the resistance, voltage fall will tell a lot. Because current (I) is same (Figure3: kirchoff's law) in this case to find bad parallel. It is easy. R*I=U

But Krichoff's law pushes me to open the connection of parallel to find the bad cell. This make it bit challenging.
(HVACfun.com picture: Like R1 is the bad one cell resistance. Voltage over R1 ja R2 is same but resistances are different.)

 

[Hillhater]

Obviously for a constant current, Internal resistance will be directly proportional to voltage drop, so no,..you dont have to calculate the IResistance to locate the bad cells.
But ..remember the volt drop will change depending on the actual current load. !
And yes, you will have to separate the parallel cells in order to measure the IResistance of the individuals.

 

[jerrypa]

I measured the resistances easy way. And I am not wiser yet. Few seems to be weaker..mmm.

Here is the reason why I have to separate the parallel cells for better results (when i have time). Parallel resistance do not have big differences even the other (R1) resistance is wery big. (HVACfun.com picture: Like R1 is the bad one cell resistance. Voltage over R1 ja R2 is same but resistances are different.)
2 parallel resistance
1/ R1+1/ R2=1/R
1/ 0,06 ohm + 1/ 0,06 ohm =1/R => 0,03 ohm
1/ 0,30 ohm + 1/ 0,06 ohm =1/R => 0,05 ohm
1/ 100000 ohm + 1/ 0,06 ohm =1/R =>0,06 ohm

Meaby the battery is death? Because the internal resistance is almost 10 times bigger than new one.

 

[Hillhater]

Meaby the battery is death? Because the internal resistance is almost 10 times bigger than new one.

Unless you measured the Iresistance on the new battery at the same current, in the same way, then its not easy to compare.

Are those readings from parallel groups ?
Some of those IR readings are double the lower ones..start there splitting the groups down and testing individual cells.

 

[jerrypa]

Now i found few laptop batteries. Maybe this give little life for the battery. Fast review show that there are couple good(read better) Sanyo cells (i hope ).

Cell type
1. SF US18650GR T 8A1130H26D
2. UR18650ZT R1122 P37C SANYO
3. ICR18650-26C SAMSUNG SD1 B23
4. UR18650FM M50A SANYO
5. UR18650FM R1122 P30C SANYO
6. LGABB41865 K167F166A2
7. UR18650FM M46A SANYO
8. LGABB41865 K181F307A4

 

[jerrypa]

charger