Finalizing a123 LiFepo4 pack configuration, advise ?

[dragonfire]

As my SLA´s ae bloated i am going ahead with converting my Goped ESR750 EX from BB´s hr12-12 to Lifepo4.

In order to keep on using the stock onboard charger which is seemingly capable to charge LiFePo and even Lipo ( except balancing) in needed modes according the manufacturer i want to make packs that are cheaper than the oem ones ( 2200 dollars for a 32a upgrade kit...) but am wondering if i can use the 2,3ah´s per cell LiFePo´s instead their 1,1ah cells ( same voltage) and can save on the second pack of parralel cells the OEM version does have ( or should i get that a123 "26550"-Cells" ( Dewalt) setup for even more range ?).

this is the factory kit´s specifications:


14 cells (7s2p) at lbs of weight and 25,9v charged voltage ( allthough the onboard charger is reported to put out 24volts...? )
8ah per pack and 90 amps current draw capacity.

this in parralel gets 16ah capacity, a third pack added on 24 ahs and with 4 packs it shoudl be at 32ah, i figured i will want a 3s1p pack of those 26550 cells as a seperate booster pack to get voltage more closely to the 31 volts cutoff.

i am also not fully decided whether to use salvaged dewalt/26550 cells at 4.50€ "as is", or roughly 6€ per cell allready packed/ checked and with balancing wire from china ( includes custom and shipping fees, hence "roughly 6€") or ne cells at 12€ per NEW dewalt-cell respectively 7,50€ for the 1100 mah ones ( also new and allready with soldering tabs).

I am also yet undecided if going for a 16 ah or a 32 ah setup, as the 16ah setups seemingly allready get 10+ miles in range under heavy load this may allready uffit my purposes.



People have tried anything from 8s1p to 7s7p and got range increase, i mostly wonder if i should get the 1100 may cells or the beefier 2300 mah ones, even more considering the lesser the number of cells the fewer risk from not actively ruinning balancers on them.

oh´and should i install a safety switch on the cells that cut of to prevent overcharge and maybe also over-discharging if the controller is allready set to cut of at 20.4 volts per bank ?

 

[dragonfire]

i essentielly still need to know whether i should better strap up 7 or 8 cells in series in order to get the best charging state out of them. as for paralleling, i am also yet undecided if i should strap 2 or even 3 of them in parralel on each 24 volt bank or if 1 or even 2 of them per bank ( 2 banks each 24 volts in total) would be way to go ?

the maximum draw of the motor is reported to be at approx 115 amps peak, cruising current al like 30 amps and trafficking at like 70 amps. divided through the number of the banks ( 2) the load per pack will be at 57 amps peak, theorethically still within the capability of the dewalt a123 cells but would 2 or even 3 in parralel double or even tripple my range effectively / would one or two in parralel be totally to few compared to the SLA´s that come stock ( rated at 12 amps but it´s more likely 5 amps at each of it´s 4 12 volts batteries they would, but still: they are sooo tiny in seize and i have 0 experience on them yet) ?

 

[scoot]

I believe you will need to go 8 in series. That is what I will use in my esr battery upgrade. The lifepo4 batts are typically rated at 3.7 v max charge voltage. 9x3.7 = 29.6v. last night I did a time study of the esr charger during the last portion of charging cycle. My cycle analyst meter showed that the tail of the charge was reading 30.2 v for more than an hour, but the current readings were quicky and steadily declining during that period. I'll post some actual readings later so you can see what I am talking about. keep in mind this is while charging the SLAs... not sure what would be observed with the lifepo4 batteries. But i sense we would be ok with the stock charger and a 8s setup.

One of the fellas out on the goped nation forum did a bang up job with the a123s just like you are planning. he did not even use any voltage protection or balancing circuits. He just connected all of the cells in parallel once a week to let them balance that way and made sure he never over dicharged the them during use. he was however using an high powered external charger. I would prefer, as you do, to use the stock ESR charger... I am not seeking rapid charging like he was.

 

[dragonfire]

thanks, i was figuring out it should be 8 cells but another gpn member allready used exact these cells in a 2008 thread and he was running them 7s7p ( yes, 7 parralel) and mr patmont stated that the charger with batteries unplugged should put out 24 volts at each lead.

i was mistakening the number of ah´s with the numbers of cells in the goped specs at first

i got a quote on 8s4p from hobbyking.com , he sells used and matched ( simila resistnace in ohms) a123 cells singular and in packs, he also sells the packaging material and wiring.
4 packs of 8 a123 26550 cells are a 296 Dollars shipped ( exclusing custom tax and sales tax, another quarter on top of this as for germany).
the packs are ready to install with rc-connectors and a secondary balancer cable that i can use for reading out the singular cells manually, too.

i was reading several knowledgeable comments about the a123 cells, all saying that the batteries will remain within >0,2volts in balance "automatically", if a singular pack drops faster then cause of one or two dying cells that are pulled through with so a checkup every some weeks during season should fit the "balancing" purpose.

As for low voltage protection i may will find a cycle analyst in europe or will go with a rc battery watcher from Graupner that has a buzzer that can be set as liked, i´ll set it at 16.5 most probably, if i run fewer than 4 packs a little higher in case of sags in after heavy load.
This will require some testing with my lipo powered drill in the pocket to assemble the deck fast in after putting load on the setup, i also think about a low-resistance thermoelectric module to be placed close to the brushes and another one close to the batterie both running on different frequencies via microphone input to my xo laptop, i hope i can add on a magnetig measurement module somewhere aound, too , to gather data about altering loads for regen and if i ever get to do a delta/wye mod, hopefully the magnetic nterferences won´t be too hard to accomplish/ filter.

i still want a booster pack setup to be feasable for the purpose and close to 30 watts is allready plenty.

though, still undecided if i will try and add it on th esr, the charly or a project 7 inch airtire setup ( some knockoff with usable quality) that nicely fits the two still good sla´s but misses controllers.

i´d like to try a regen braking setup in the form of switching to regen mode when going of the throttle on a brushed engine or then switching over to neutral and regen circut on a brushless high-rpm outrunner, but this will make me want a booster to overvolt it even more.
when a addidional brake is applied, a regen brake as kind of "third brake" brake can act stepless but still hardly adjustable if i run a number of 2,5v small capacity supercapacitor cells as a fixed minimum braking current load and a conveniently splitable number of bigger capacity similar cells ( alternatively the really low resistance LiPo cells but i´d hate to have to care for their risks even in smaller series), their capacity whill then be used to feed a connection loop through the supercaps to the drive controller when a specific current draw is put on the batteries so i can still pull strong even if the sl´s have allready dropped, maybe i will figure out how to smoothly parralel this both different cells without hurting the sla through high discharge without to much fuzz i´ll try that but i don´t believe so. i should dig more about dc-dc converters and the sequence of usage of supercaps in the prius and alike cars.

hopefully this will lead to a form of "parking brake" by switching the polarity when parked and the third brake is into "lock", too.
so i can still have my booster pack somehow

 

[dragonfire]

As for the ESR onboard charger, the draw at the wall plug is at 72 watts stable, considering a maximum efficiency grade of approx. 85% , 30 volts per bank seems just right.
the dewalt/ 26550 a123 cells do have the same 4.2 volts critical limit, so the onboard charger seems to be just right and as charging in steps of 1c, 2c and 3 in pulses balancing seems to be seldom necessary, even more if the cells are out of the same production run/ badge ( used cells other than from dewalt direct returns will bear alike risks, too).

Nice to see how the ESR seemingly is capable to go with lipo´s and both stock controllers.

By chance, do you have a Go-prammer you´d lend out in axchange for a gift and safety deposit ( or an idea of the manufacturer/ cable wiring scheme so i can crimp a connect cable to my laptop´s rj-45 plug and communicate with the drive controllers eprom in 8 bit mode, maybe as simple as currupting downloadable Kelly controller software written in basic) ?