Steveo's Dewalt A123 Individual 200+ cell testing...

[steveo]

During the last few weeks I've been doing individual cell testing for all my Dewalt cells individually. I had charged up & balanced each pack, then i let the chemistry settle for a few days.

I've been using the sr960 Mega power Charger/discharger/balancer for my testing (I have this charger modified like doctorbass did for 4 wire sensing to get more accurate results)

Majority of my cells are in the 2200mah+ rating with the exception of a few cells & packs ranging 2100mah - 2199.

All cells that i have tested have fallen for the internal resistance mark, about half dozen at 11mohs, and the rest ranging from 13-12mohms

Here are some photos of my progress

Cells about 60% all tested

This is how i bent the tabs to latter sodder when assembly of the packs!

Lower Voltage Cells (Anything below 2v was placed aside)

Here is an example of a cell discharged beyond 2volts, but THE READINGS ARE GOOD BASED ON TESTING!

another example, this one has lot some capacity.

=============================================================================================================================
=============================================================================================================================
I had a few questions to clarify, as per there are many experts out there that have already done this 8)

Low discharged cells
What should i do with the cells discharged pass 2v but still give excellent readings? can i still use them? what is the down side if i do use? Has there life been impacted alot, how many cycles could they have possibly lost say they where that way for about 1 months time ? ( i usually get cell packs every month )

Battery pack configuration to go with
I am looking the build either 5p 30s battery pack setup; or .. I've been thinking about going 23ah with 10p, 30s and using the dewalt black balancing side covers to hold them together 23ah each 2.3ah pack, it will be difficult but defiantly possible to mount

Connections after all testing completed
I will be using 10awg for my connections & 45amp Anderson connectors
Now I'm Still Looking for suggestions for my battery pack building *********HELP*********

My objective here is to try to keep the balancing side covers to hold the cells together if Possible!!!

-i could solder the tabs to the 10awg directly, I would have to cut all the side covers for this.
-I could use copper brad equivalent to 10AWG
link >>> http://cgi.ebay.com/ws/eBayISAPI.dll?ViewItem&item=230239533920
-I could use like a metal, steel rod, something i could solder to and could handle discharge of min 45-60amps.

Any Feedback from everyone would be awsome.

thanks
Steve











`

 

[Link]

:lol: You're as bad as the good Doctor himself.

 

[Doctorbass]

During the last few weeks I've been doing individual cell testing for all my Dewalt cells individually. I had charged up & balanced each pack, then i let the chemistry settle for a few days.

I've been using the sr960 Mega power Charger/discharger/balancer for my testing (I have this charger modified like doctorbass did for 4 wire sensing to get more accurate results)

Majority of my cells are in the 2200mah+ rating with the exception of a few cells & packs ranging 2100mah - 2199.

All cells that i have tested have fallen for the internal resistance mark, about half dozen at 11mohs, and the rest ranging from 13-12mohms

Here are some photos of my progress

Cells about 60% all tested

This is how i bent the tabs to latter sodder when assembly of the packs!

Lower Voltage Cells (Anything below 2v was placed aside)

Here is an example of a cell discharged beyond 2volts, but THE READINGS ARE GOOD BASED ON TESTING!

another example, this one has lot some capacity.

=============================================================================================================================
=============================================================================================================================
I had a few questions to clarify, as per there are many experts out there that have already done this 8)

Low discharged cells
What should i do with the cells discharged pass 2v but still give excellent readings? can i still use them? what is the down side if i do use? Has there life been impacted alot, how many cycles could they have possibly lost say they where that way for about 1 months time ? ( i usually get cell packs every month )

Battery pack configuration to go with
I am looking the build either 5p 30s battery pack setup; or .. I've been thinking about going 23ah with 10p, 30s and using the dewalt black balancing side covers to hold them together 23ah each 2.3ah pack, it will be difficult but defiantly possible to mount

Connections after all testing completed
I will be using 10awg for my connections & 45amp Anderson connectors
Now I'm Still Looking for suggestions for my battery pack building *********HELP*********

My objective here is to try to keep the balancing side covers to hold the cells together if Possible!!!

-i could solder the tabs to the 10awg directly, I would have to cut all the side covers for this.
-I could use copper brad equivalent to 10AWG
link >>> http://cgi.ebay.com/ws/eBayISAPI.dll?ViewItem&item=230239533920
-I could use like a metal, steel rod, something i could solder to and could handle discharge of min 45-60amps.

Any Feedback from everyone would be awsome.

thanks
Steve

`

Hello Steveo!

Wow, i'm impressed! you did a nice job.. it turn out that you've been hit by the Doctorbass-o-mania...

About the low cell that get good measurements results, i have no exact opinion, but i still think they are as good as others... They have a solid reputation about their strongest endurance! Only cells that was reversed and that display low voltage after that could give low measurement and then need to be rejected.. usually cell under 2050mAh had bad past experience...

Oh I forget!!!

Please wait if you can before to asssemble your pack!!

I will get the completed version of my cells matching software!!!

My friend Eric is working on it to update it and to get better statistic calculation and is also based like the DNA network calculations!!

The goal is to assemble multi parallel pack that have all the same capacitance and internal resistance!!

The perfect exact matching over 160 cells would take more than twice the age of the universe to calculate using a simple P4 quad core computer if each possibility would be calculated... but this is not possible from now :lol: so we will accept some error margin and add some filter in the calculation to simplify the matching.


No software like that exist on the earth!! since used cell matching would have been done....

That will be a premiere!!.. and will be used HERE for the first time!!


I'm pretty confident about the results!

We established some rules about the matching calculation in the software.. that will accelerate the processing.

we use Mathlab version 7

in the pic below, the pack id is each parallel pack results about RI and Capacity

From now, the red curve represent the state of all cell that had been placed randomly in all parallel group... and the blue curve represent the final optimization.

on the left , the blue graph represent the error between each parallel pack over each iteration of the optimization (on the X axe) at the end of each curve, the repetitive data that flaten the curve is due to the toping results about the exchange of cell over different parallel pack that entered in oscillation and always get the same results and can't be ameliorate. like you know, as this curve fall, as the error decrease between each paralle pack.

 

[paultrafalgar]

I've never used Mathlab. My operating system is Debian Linux. So I looked for a Linux equivalent of Mathlab and found Octave. There is a repository of functions written for it, which are supposed to be compatible with Mathlab here:
http://octave.sourceforge.net/packages.html#main
I hope someone finds this useful.

 

[steveo]

Hey Doc

Thanks for the fyi for the LV cells i have =)
What i can do is build a few packs just using those cells but only the ones with best capacity to see how the pack behaves!

Doc i'm defiantly waiting for the software to help me with this, that would be of huge help for me here! Do you know how long it will be.. i estimate approx 1-2 weeks for my completion between work etc. i hope software is completed by then :wink:

-steveo

 

[deVries]

Oh I forget!!!

Please wait if you can before to asssemble your pack!!

I will get the completed version of my cells matching software!!!

My friend Eric is working on it to update it and to get better statistic calculation and is also based like the DNA network calculations!!

The goal is to assemble multi parallel pack that have all the same capacitance and internal resistance!!

The perfect exact matching over 160 cells would take more than twice the age of the universe to calculate using a simple P4 quad core computer if each possibility would be calculated... but this is not possible from now :lol: so we will accept some error margin and add some filter in the calculation to simplify the matching.

No software like that exist on the earth!! since used cell matching would have been done....

That will be a premiere!!.. and will be used HERE for the first time!!

I'm pretty confident about the results!

We established some rules about the matching calculation in the software.. that will accelerate the processing.

we use Mathlab version 7

in the pic below, the pack id is each parallel pack results about RI and Capacity

From now, the red curve represent the state of all cell that had been placed randomly in all parallel group... and the blue curve represent the final optimization.

on the left , the blue graph represent the error between each parallel pack over each iteration of the optimization (on the X axe) at the end of each curve, the repetitive data that flaten the curve is due to the toping results about the exchange of cell over different parallel pack that entered in oscillation and always get the same results and can't be ameliorate. like you know, as this curve fall, as the error decrease between each paralle pack.

Hello Doc,

This certainly got my attention!

Any status report you can offer about this method?

Also, I came across this company which seems to have interesting ideas about battery management. Comments? ...

http://www.youtube.com/watch?v=oiRhChE6rt8

http://www.intersil.com/cda/deviceinfo/0,1477,ISL9217,0.html

http://www.intersil.com/Applications/printdoc/E-Bike.asp?device=ISL9217

Hey Steveo,

What did you decide to do to balance your packs or sub-packs?

TIA