is A123 8s2p 4.6Ah enough for 500W bike?

[DrInnovation]

My thunderpower-based Lipo pack is starting to show its age after a year on my bike (bought them in 2008 and used them for UAV... so >500 cycles now). My commute is short but steep so I max out my 500w 24v motor, so use 30A for 2-3 min of the trip that is 6-8% grades, whole trip is only 6min). The high load is probably why my old Lipos are loosing balance quickly Getting harder to keep them in balance and I lost one 3P group and had to replace it with a backup from another pack (which already lost other groups). Now I'm out of spare parts ;-(

Will see how long that lasts but Wanted to start on figuring out my replacement before the lipo pack is totally dead.

Though about replacement lipo, maybe some zippy or tunegy but then felt maybe lifepo4 might be a better choice so I don't have to mess with them as much
So started looking at other choices and found these on ebay:
http://www.ebay.com/itm/A123-16-Cell-8S ... 2ed90008d1

The price built is about the same price as buying the raw cells (in low quantity) from a123
https://www.buya123products.com/goodsdetail.php?i=11
And various posts have been positive on sz_ohmygod's builds, they will even add the JST-XH Balancing connector for me so hard to beat the price.


So looking for feedback on my proposed build. If I'm drawing 550w (assuming maxing out 500W motor plus 10% losses in the system) at 24v nominal will be 22A, which for the 2.2Ah minimum rating is a draw of 10C per cell, and in a 2P configuration should mean 10A/5C, well below the A123 70A continuous rating From the specs (https://www.buya123products.com/uploads ... c1e6e7.pdf) it says at 10A draw, they would say above 3v until almost 2ah per cell is used. 4Ah should be more than I need for my commute. I know many will say go big.. but that is more $$ and weight.


I also want to add a BMS for added protection. Many good posts here about Signalab used in Ping
(which from what I've seen is the same as this
http://www.ebay.com/itm/BMS-for-24V-25- ... 25a2548173

or maybe the
https://bmsbattery.com/bmspcm/330-smart ... ent-20_40a
(which I might be able to reprogram if I go back to lipo)


Anyone have experience with this model?
http://www.ebay.com/itm/PCB-BMS-for-24v ... 28014f4bcc
(which physically is a better fit for my battery box, and has included temp sensors. With the A123 smaller size I think I can get the signalab to fit)

Any comments? Am I underestimating my needs in estimating that A123 8s2p 4.6Ah should be enough for short but hard rides on 500W bike?

Any recommended design changes/tweeks? which BMS's that I should consider?


Background: My setup is 24v and I've tested it working up to 29.6 but like to play it safe and run at well below max voltage.
currently I'm running effectively 7s 6p (really 2p/3p) lipo cells). I generally charge to 4.005v per cell, normal discharge is only down to 3.8v. I've been doing mean-well bulk charging @2a for the pack (<.3a per cell) with monitoring (cell8s). I rebalance the 3s components when any pack is more than .1 off. Balance used to be need less than once a month last summer, increase to almost every week this spring, and last week twice -- wich is why I swapped out a weak group of 3P which I also noted was getting warm to the touch when charging).

[999zip999]

If you keep each cell voltage between 3.45volt and 2.8volt A123 will work perfectly, but 3p ( buffer ) really . They keep balance at these low discharge rate for 1,500 love cycles. Yes for real A123 cells.

[DrInnovation]

eTrike, Thanks for the link. hope there are some left at that price, even used ones will be great for me in a 8s4p or 8s5p config
the BMS is mostly for low-voltage cutout since the a123 fall off a cliff. Could build my own, but for 25$ for the cheapBMS in my list it seems like a reasonable choice. Used to worry about people turning my light/bike on at school and draining it but so far not been in issue.

The motor is an old MAC on a lashout... controller is integrated so not upgradable without a lot of work to remove the potting. This is a commuter bike, not worth upgrading as it gets me to work just fine. Wife already compalined about the 4 blown tires this year.

[DrInnovation]

what kind of tires? I've done 10k with no tire issues other than normal wear on schwalbe marathon plus. I added a touch of slime to the tube which kept small/slow thorn holes from leaking.

they are labeled Crossroads, and are what came with the bike.. first 2 blowouts were because there were no liners on the rims and I guess the added weight of me + SLAs (back then) were too much on the many potholes and pinched it. I did not even know about liners, but learned after the second blowout.
3rd and 4th (more recent ones) were sidewall blowouts -- the side walls of the tires were worn so thin could see daylight through them and small buldges made their way through the sideall and popped. I'd never seen anything like it. Did not realize it on the first one, though it was pinching again, so just put a new tube in and added more tape as a liner. When a second blew in less than week with a very noticable "pop" sound, I knew it was something different. When took it off and could see daylight through the tire, I realized it was the side walls. Bought new tires and tubes, so hopefully I'm passed all that.

Back on task, 4P would just under 4C, 5P would be 3C, much better than 2P hehe. Go for 6 or 8P if you can

I've put over 12k miles on trike(s) using 16Ah(8P) worth of these 26650 A123s. about 20% of that has been at 4C peaks/2C continuous, the rest at 2C peak/1C continuous. minimal balancing after lowering C rate and they still have well over 90% original capacity.

I know bigger is better, but get 8s4P (32 cells for 9.2ah) I'd be looking at 260 + shipping just for the cells which is then above or in the rough price range of lots of other options. I could get 10ah of headway cells assembled kit with PCM for $175 (greenbike kit) (or just headway cells for even less). My hope was that A123's extra power would allow a small 8s2p build to be viable for my short but agressive commute. Part of the argument for the ebike was allowing commuting without paying for parking (which is $450/term), so want to keep the costs down.

Anyone else abusing A123's in low parallel count?

[DrInnovation]

Agree heat is a killer. In this thread
http://endless-sphere.com/forums/viewto ... 0#p1020654
they found no appreciable temp increase for 10A (4C) though it was only a short period (1min @2c 1min @4c), So I felt maybe 4C is still safe usage. My usage will be short (3 min of 6-8% grade), so not a long time of high demand.

Thanks for the added read. I did note that the 1000 cycles @4.68C, is probably just where they stop testing.. as its the same 1000 cycles, for 1.6C (though 4C has a greater voltage drop so must have greater heat too).

Longevity does matter (hence a BMS even though it adds cost). If I could get them at $2 I'd definately go bigger. At $4 for used cells (best price i've seen so far) not sure I'll take the plunge as it depends on how much life they really have left.

[wb9k]

If you can force some air across the A123 cells, 10C continuous for a few minutes should not be a major problem. Won't last as long as a larger pack of the same cells, but should handle your duty cycle just fine I think. I've been able to overheat A123 26650's in testing, but I had to pull 45-50C for about 1.5 minutes to do it. This was with no cooling of any kind. Avoid using the bottom 20% of capacity at these high discharge rates.

[wesnewell]

That's $1.07 per wh. More than 3 times the cost of 20C 5ah 4s hardcase packs. It's your money.

[dnmun]

If you can force some air across the A123 cells, 10C continuous for a few minutes should not be a major problem. Won't last as long as a larger pack of the same cells, but should handle your duty cycle just fine I think. I've been able to overheat A123 26650's in testing, but I had to pull 45-50C for about 1.5 minutes to do it. This was with no cooling of any kind. Avoid using the bottom 20% of capacity at these high discharge rates.

+1 for the education to avoid high discharge rates when the battery is not at full charge. this applies to all batteries. if you want them to last do not push high currents below the 80% SOC level imeo.

[markz]

So you are buying a battery pack from https://www.buya123products.com/goodsdetail.php?i=11
And they are a reputable online company?

[DrInnovation]

If you can force some air across the A123 cells, 10C continuous for a few minutes should not be a major problem. Won't last as long as a larger pack of the same cells, but should handle your duty cycle just fine I think. I've been able to overheat A123 26650's in testing, but I had to pull 45-50C for about 1.5 minutes to do it. This was with no cooling of any kind. Avoid using the bottom 20% of capacity at these high discharge rates.

Thanks.. I was hoping you would weigh in as your posts were past of what got me thinking I might afford a small A123 pack. What's your feeling for a need for a BMS if I do that.. given the high demand will I need regular balancing?

So you are buying a battery pack from https://www.buya123products.com/goodsdetail.php?i=11
And they are a reputable online company?

They are one of the two sources I was considereing for new cells (don't know shipping cost). The other being
http://www.ebay.com/itm/A123-16-Cell-8S ... 2ed90008d1
which is about $8.125 delivered and welded,

Also considering dozentrio's post as in this thread http://endless-sphere.com/forums/viewto ... =9&t=70075 for some used ones). Still waiting on the outcomes of his cell testing -- price with shipping is about $3.60 per used cell.


Thanks for everyone's feedback, and while I might go the used A123 route I'm leaning lipo/lico. Given the earlier feedback of needing 4+ in parallel I was pushed back toward Lipo and dug deeper into analyzing my current equipment and alternatives. Seeing A123 as too expensive I ripped apart my packs that would not stay balanced and two others with a dead group, and my initial testing shows I should be able to salvage individual cells from the recent problem packs so I will repack to have 7s5p of 20C rated lipo, which should be okay for a bit longer.

The increasing pain of taking my bike's battery box apart to balance charge, was part of my problem (its a pain), but I found a 7s lipo/lico singleab BMS on ebay for $35 -- which I bought with my "ebay bucks" that my wife considers "free"). Given the positive things I ready about this company's BMS (used in Pings) I hope that plus the 7s5p rebuild will be a viable solution for a while. And given that I bought that, I am now more invested in staying Lipo/Lico for now. Even looking at doing a solderless 18650 laptop-battery build, or recycled mac-battery build. Have a lot of both already and can get a box of old laptop batteries for much cheaper than even the used A123s. (Might still get the A123s for my robot but that is "work".. so I might still test them out on the bike so I know better for the furture


With respect to rebuilding Lipo packs, If I have excess cells allowing a rebuild of 7s were some are 5P and some are 6P is that a good idea?. Given I'll have a balancing BMS, is maximazing total capacity in a mixed config better or should I just stick with 7s5p and keep them matched insize? I've read plenty of threads were the number of 18650s were not match but some strings had more. (I think that was about matching effective capacity rather than numbers)






T

[markz]

Is

sz_ohmygod (3439 ) 99.5% Positive feedback

a reputable seller?
Or someone you just stumbled upon. His sales and feedback look alright, perhaps a little low (3439), for the payment risk.
$130x2 at a minimum for 52.8V 4.6Ahwith free delievery and those huge anderson connectors, so $520 for 52.8V and 9.2Ah.
Everyone always seems to come around the same price.

[John in CR]

You've been using your ebike for about 8 years, so how much energy does your trip take? That's your starting point. I have lots of A123 M1 cells, and they're great, but they're bulky and expensive. Modern cells are much better. 2p may be fine with your current demand, but to me the pack falls short unless this is ridden as a bicycle with electric assist.

BTW I made a car battery with 24 A123 M1's, which works great. It doesn't even flinch at the 80A peak my starter draws, but my pack only runs for a second or two each time I crank up the car. Next time I'll probably use only 4p or 5p, since 6p is overkill.

[DrInnovation]

Is

sz_ohmygod (3439 ) 99.5% Positive feedback

a reputable seller?
Or someone you just stumbled upon. His sales and feedback look alright, perhaps a little low (3439), for the payment risk.
$130x2 at a minimum for 52.8V 4.6Ahwith free delievery and those huge anderson connectors, so $520 for 52.8V and 9.2Ah.
Everyone always seems to come around the same price.

I did just stumble upon then but checking showed good feedback and there are a few ES post having bought stuff from them and having done well.
But looks like I won't be testing those waters this time. As dozentrio's tests look promising, so if I go A123 I'll more likely go used and wider parallel.

And @John it is only an eletric assist, I want to pedal, just to keep heart-rate < 120 so I don't get to work too sweaty.

[wb9k]

If you can force some air across the A123 cells, 10C continuous for a few minutes should not be a major problem. Won't last as long as a larger pack of the same cells, but should handle your duty cycle just fine I think. I've been able to overheat A123 26650's in testing, but I had to pull 45-50C for about 1.5 minutes to do it. This was with no cooling of any kind. Avoid using the bottom 20% of capacity at these high discharge rates.

Thanks.. I was hoping you would weigh in as your posts were past of what got me thinking I might afford a small A123 pack. What's your feeling for a need for a BMS if I do that.. given the high demand will I need regular balancing?

If you're building an 8S pack, I would use a quality hobby charger to balance it periodically. I'm partial to the Cell Pro 8. I don't think you'll need much, if any, more balancing than would "normally" be necessary [EDIT: but it is a possibility with used cells--not charging at high rates can mitigate the risk]. I would make sure to pay close attention to pack behavior at low SOC to make sure there are no dogs in the bunch that end up biting you with outlier behaviors and possibly an overdischarge situation. This is mostly if you are buying used cells. Use a celllog or similar to keep a close eye on the pack for the first several cycles you run. Once you are comfortable you understand and can more or less predict the way things will go with the pack in your regular riding routines, you can ease up on the management a bit. Some form of automatic LVC on the bike is preferred to prevent damage by overdischarge. If the cells all bottom out around the same time, you can get away with pack-level monitoring of LVC. If you have one really weak cell that bottoms out long before the others, you may need to opt for cell-level monitoring for LVC. If you tend to use the whole capacity of the pack on a regular basis, you may want to go the cell-level route as well.

In cars and other devices with large packs, I'm loathe to go BMS-less. But in little packs like this that can easily be taken out of their installation for bench time I think it's fine if you're reasonably attentive and regularly check top and bottom SOC voltages to make sure nothing's going awry. You can cheaply and easily monitor all cell voltages in seconds....there's no excuse for not at least doing that with such a small pack. Use a charger that cannot possibly overcharge any cell in a reasonably balanced pack, DONT GET LAZY, and this is plenty of BMS for a little LFP bike pack, IMO.

Good luck to you, whatever you end up opting for!

[markz]

Yeah I think a123 is the way to go.
I've dumped BMSB, those bus batteries look promosing.

[markz]

Whats also on my radar is OSNPower, but I've emailed them several times and have got no response.

[DrInnovation]

Figured I'd post a final "what I did" on my own question. Did not get get A123 yet, but to test things out, since I had killed a few of my lipos, I decided to try the ride with 6Ah of 7S lipo (20C rated thunderpower) charged to 28.8v as I would get from A123. I had no problem making it to work and back. When I returned I was still at 27.6v so could easily have made it twice and thus would answer that the A123 woudl easily have handeled the commute.. not sure what the impact would be on its life. But I decided to go with 18650s instead, building a 7s12P pack of Panasonic CGR18650DA cells I salvaged from some Military batteries (BB-2590's).

[quamau]

I'm throwing in my 2c of experience.

I have a similar commute, short and steep, perhaps a bit longer than yours, and I am doing it happily since more than 2 years with A123 32113 cells (with threaded tabs). Nominal 4.4Ah, measured about 4.0Ah new, they were gray market at best. I am more than happy using them at 18A (limited by CA) continuously for approx. 10' ride. Number of rides is about 3-4 a week, so number of cycles is still humorously low, say 500. I check balance every month or so and they are always perfect.

Note, however, that except for a few weeks in the year (i.e. now) I ride at low external temperatures, often below 0°C and very rarely above 15°C.