Paralleling question

[Auhydride]

Hello,
I have a question, searched about it on the forum,
http://endless-sphere.com/forums/viewtopic.php?f=14&t=8226&hilit=paralleling

"You can create self balancing (self correcting) behavior if you place cells in parallel with each other before you connect them in series and that helps to erase the manufacturer defects.
But in absolute terms all cells drain the same energy... the cells at the front of the series and at the end drain the same amount.
If you are really interested in doing the best thing possible you would mate up the strong cells with the weak cells in a parallel situation so that on average the result is "average". But that's more effort."

Basicly, I will have 24 a123 lifepo4 cells and need them in 8 series.
I wanted to parallel them as 3 cell formations first before connecting them in series, thinking It would be easier to balance up 8 cells, instead of 24, would need less multi voltage meters to track them all too.
But an electronics guy that works in the same project insists it will cause batteries to charge each other up and waste energy. But I don't see any reason for them to go out of balance in the first place as I discharge them.

There is also this:
http://endless-sphere.com/forums/viewtopic.php?f=3&t=2872&hilit=+parallel

"I did some more investigating and I think I understand why this is a bad idea. It seems with cells built in parallel, like my proposed 1s4p pack, it would be perfectly fine to charge each 1s4p pack alone. This was actually where I was heading, I was going to get fifteen of the 2A wall chargers so that each 1s4p pack would have its own charger. I could just disconnect the series connections and plug in the chargers to each pack. If I then connect these packs in series for my 15s4p configuration to discharge the packs and I have a bad cell in one of the parallel strings it will not show up as a voltage loss but it will have a loss in capacity. Over time this would probably destroy the entire 1s4p pack before I could spot the problem. It looks like I need to keep them in series first and get the desired voltage and then connect them in parallel to get the desired capacity."

I don't really get it, does this happen because of the reason here:?
http://endless-sphere.com/forums/viewtopic.php?f=14&t=3482&hilit=parallel

"For practical purposes you do want to match the capacities with the same size cells to maximize the cycle life & performance.
The bigger battery will be carrying the brunt of the full load once the smaller one stops contributing current & will tend to wear out faster.
I wouldn't characterize it in terms of safe or unsafe."

So it would happen because weaker cells in the parallel wouldn't be contributing and eventually causing the buddy cells to discharge at higher currents and shortening their life. And because they are in parallel, you can't easy monitor them to see which one reaches cut off voltage sooner, etc.
But other than this, are there any practical issues for "buddy pairing" batteries? What would be the final word on this?
Thanks

 

[dnmun]

you can build it 8S3P. use a BMS to charge and discharge. 24V BMS.

 

[Auhydride]

No idea how to build a BMS though.

 

[michaelplogue]

No idea how to build a BMS though.

http://www.batteryspace.com/index.asp?PageAction=VIEWCATS&Category=1385


.

 

[dnmun]

why would you want to build a BMS if they are so cheap to buy?

 

[chet]

series first, parallel first. I don't think it would make any difference. You should have some sort of BMS for any large battery array.

 

[scoot]

As I was wading through the thread ("The care and feeding of a123 packs"... what an informative thread), I had some what come to the conclusion that paralleling some well matched cells before stringing them together in series might be a good way to go. that way you more or less only have to have a balancing circuit for the single serilly connected string, albeit higher rating on power than seperate balancing/BMS ciurcuits.

My plan, unless someone can advise otherwise, is to use some large foramt cells (e.g, headways or BMIs), pair them up and string them together in series for the voltage I need. I am shootin' for the 2p8s configuration. If I can match the pairs well (say evaluating them with a CBA II or something), then I figure they can sort of effectively balance each other, then use a single BMS, or balancing circuit to balance the serailly connected cell pairs, or "buddies".

I am glad you have asked the question here though. It will be interesting hear any additional feedback that may come our way on this topic

 

[Auhydride]

I'm not tempted to buy a BMS as these batteries will be used for one single race, and once they near cut-off voltage, they won't receive full charging until the end, maybe some regeneration. Also i'm located in Turkey, gotta be careful with the package customs.
Right now, i'm planning to use a multi charger off the R/C chargers market that can balance up to 6 cells. (Charging the pack 4 buddy pairs at a time) so this paralleling would really save me some time for testing and such.
But I do plan to get one for an e-bike gotta admit, I like ready-made stuff.

 

[icecube57]

Make sure the chargers are for lifepo4 and not lipo.

 

[michaelplogue]

Also i'm located in Turkey, ....

Where are you in Turkey? I'm stationed in Ankara.

 

[Auhydride]

Istanbul PM'ed you, would love to talk!

 

[dnmun]

too cool, two guys in turkey. now what's the odds for that?

put your a123's in parallel and then string them together in series.

your cells are worth more than a BMS would cost, they will last forever if you take care of them too.

if you can use a balancing charger that helps a lot, but if the cells reverse in discharge, it's toast. so you could use a LVC like gary built in that thread too.

 

[Auhydride]

Right now, I'm using a sigma lifepo4 balanced charger to balance the cells and big coil of resistance wire to discharge batteries and see which one nears cut-off voltage first, one cell out of 6 cells I bought off that olympics ebay guy is way off balance, rest are doing quite well. My team lost rest of the test batch lol, which had 6 more cells from the same seller, wanted to test them too (the user have found his batteries at a local r/c shop two days after this post, yay.). Couldn't think of anything more practical considering the lack of equipment over here.
Ok, I will do the paralleling for sure.

"too cool, two guys in turkey. now what's the odds for that?"
It's some luck! I don't see the reason not to drive an electric bicycle in the istanbul's solid-state traffic and the gas costing around $7 a gallon , even my old lead-acid scooter provides me a nice transportation for short ranges. But there are so few people that are interested in electric vehicles.

 

[dnmun]

you need a voltmeter. you destroy them by discharging thme below 2V. you actually really wanna put identical batteries in parallel, then hook the series up.

 

[Auhydride]

Sure I'm tracking them with a voltmeter, trying to get a multimeter that can show multiple voltages on the same screen.