A123 cell matching

[Electricdirtbiker]

Hello all,
I read this site everyday and am always amazed at how much information there is. Thanks
I'm building a battery pack for a bike.
I have 10 of the dewalt dc9360 packs that I want to make into 8, 1s 10p packs. (20 cells remaining)
I will try to reuse the orginal cell holders for containing each 10 paralleled cells.
The packs are new and mostly have a 2006 year, but different dates (weeks?)
My question is if the cells all measure out ~2.3ah, is it worth while to match them?
How is matching done?
Are they grouped by aH, IR, or mix and match cells until all 8 packs have the same ah?
When charging the cells with a cc-cv supply, at the end of charge (3.70v), what is a reasonable ending current, >50mA?
Does this ending current have to be the same for all cells? Maybe the cell is so fully charged it doesn't matter?

Thanks again.

 

[efreak]

Welcome E.

Good questions . i am actually going to have to do the same for one of my bikes its been taking forever .
most of the packs that develop the problem cell syndrome, either have been over discharged or improperly charged (not balanced and the discharged)
I was looking in my homedepot and their batts were 2006 so !! I think we should be ok the problems or the solution I should say is in the way you plan on charging the pack that is the question I think .

The best way I have learned is to use a Low Voltage Cutoff circuit a la> Gary Goodrum (GG) and

and use a simple single cell charging method ( Variance of the Dc Module approach) that will ensure a balanced cell eachtime and the lvc will Insure your cells against over discharge

Cc-cv is a bit tricky with these cells and you are so right about having to sift thru tons posts and trying to make sense of it all but look arond and eventuall have to decide on what is best for you

 

[tailwind]

Hi Electricdirtbiker,

So thats a 26.4v x 23Ah pack your building.

Doctorbass did a great job of matching his Makita pack, Konion cells.
http://endless-sphere.com/forums/viewtopic.php?f=14&t=2764&start=0&st=0&sk=t&sd=a

You want all the parallel strings to be depleted at the same time to give you maximum range.

You can match them by adding each individual cell Ah and then have each string total to the same Ah.
Test each Dewalt pack to see if they have any runts or dud cells.

But as Doc points out in his post, a better measure of capacity is the Wh that each string can deliver. This is governed by Ah and Internal resistance.

If you are only using 50% or 80% of the pack on a daily basis I would just match on Ah.
If you want to get as close to 100% as you can then match by Docs method.

Garys LVC is a great solution to protect your pack from over discharge of any string.

50mA or 100mA is ok. You are putting very little energy into the cell at this current so the difference in capacity is minor.

Greg

 

[jeffkay]

My pack of 192 A123's is 12p16s. They were from many different years and some were abused. If you don't match them, at least make sure they all charge to the fullest. I found that if this was done, the little bit left in the discharged pack is negligible...
Jeff K. "Deep Cycle" project

 

[Doctorbass]

I wrote many post about cell matching practice and theory. search for them on that forum.

Cell matching mostly help about cycle life you will get with your pack.
and also the Ah you will be able to get on each ride.

There is a good tread off between Ah matching and IR matching. Me and my friend Eric have wrote a mathlab program that match with 0.01% acuracy all cell in multiple S and P pack.

Search you will find!

Doc

 

[GGoodrum]

I probably have as much experience as anybody with these and what I do is simply match resting voltages, after a charge. If the cells are healthy, they will retain a surface charge of something over 3.50-3.60V almost indefinitely. Cells that I've abused, by over-discharging down to between 1.5-2.0V per cell, tend to bleed off the surface charge right away, and end up with a "resting" voltage somewhere around 3.35-3.45V. These cells typically have lost about 10% of their capacity, but are otherwise okay. I've got a couple of 4P packs where I've simply grouped together four that where weak, together in parallel, and then put those in series with other 4-cell blocks that are healthy. For charging, I still use a bulk charger, but occasionally I will use some Voltphreak 2A individual CC/CV chargers to "top-off" the blocks independently, after a bulk charge.

Anyway, what I'd do is charge the cells first, wait an hour and then measure the voltage for each cell. Make groups of 10 of cells with the closest ressting voltages. Once in parallel, they will all equalize to exactly the same level.

-- Gary

 

[Electricdirtbiker]

Thanks for all your inputs and good ideas.
I will search out how Doc matches his cells and Garys LVC circuit.
Gary, I looked over your LVC circuit and made some sense on how it works. The part I stumbled on is the optical-isolator's CTR, current transfer ratio. What did you determine and why? I'm green to isolator devices. Where can I order the LVC's.
Some of my cells have 1-2 broken spot welds, can I flow 30-40 peak amps through them?
Thanks
Electricdirtrider

 

[K-ray]

Hey Guys,
I've been an avid reader here for some time. I recently
purchased some 4s4p kits from Gary so I'm at this same
crossroad. Great topic! I respect everyone's experience
and appreciate the helpfull community, so I'm hoping for
some clarity... Gary if I understand right on your recent
post, wouldn't a lower capacity parallel block (all weaker cells)
trip a LVC signal earlier (much) than the others? I'd think it
would be preferred to through one weaker cell into each parallel
block?
Thanks, you all are great :wink:
Kevin