a123 Prisimatics vs HobbyKing Lipo

[oatnet]

I am trying to do a fair comparison between a123 LiFe and HobbyKing LiPo. I have extensive experience using LiFe, and I am using an example from a pack I can demonstrate at will. I do not have direct experience with LiPo, so that example is based more on published stats. I admit a strong bias towards LiFe, but I am truly seeking a fair comparison, so I challange you to keep me honest by correcting any assumptions I have made in error.

Life and Lipo come in different voltages, making it is tough to make a pack of the same V/Ah, so I tried to align the packs in the same number of watt-hours. I did not include shipping from China in the costs, since both packs are close in weight.

================================
LiFePO4:

I have made (3) purchases of 16ah prisimatic a123 cells from Cell Man at $37.50/cell. A fully packaged 24v (8s) subpack weighs 7.4185 lbs, and a full 20s pack weighs 18.63lbs. For much of the discharge cycle, the pack stays @64 volts (@3.2v/cell) , and I have verified I am able to pull 16ah from the cells. 16ah x 64v = 1,024w, just over 1 KW.

a123 Pack:
Watt hours: 1,024
Weight: 18.63lbs
Size: 9" x 6" x 6.25" = 337.5 Cubic Inches
Cycles: 1500-2000
Cost: (20) x $37.50 = $750.

================================
LiPo:

I used a "Turnigy 5000mAh 6S 20C Lipo Pack" as a basis for comparison:
http://www.hobbyking.com/hobbyking/store/uh_viewitem.asp?idproduct=9176

It is listed at $44.20 each, and rated as weighing 913g. I used the 5ah rating and Guestimated (please correct me here) a discharge voltage of 3.83v, so I have made the assumption that a 6s pack would discharge at 23v, and contain 115 watt-hours. (9) of these LiPo packs would supply 1,035wh, a good comparison for the 1,024wh a123 pack.

Turnigy Lipo Pack:
Watt hours: 1035
Weight: (9) x 813g = 8217g, 18.13lb.
Size = (9) x 5.98" x 2.01" x 1.97" = 212.87 CU
Cycles: 300-500
Cost: (9) x $44.20 = $397.80

================================
Comparison:

Size: Lipo is 67% of the volume of LiFe. Lipo is the clear winner, if VOLUME is a design criteria.

Weight: The Lipo is 2.68% or 0.50 lb lighter. However, the a123 pack is complete to the andersons, and the LiPo builds I have seen have complex wiring harnesses added on that might narrow this gap. It is pretty close to a draw on weight so I'd call this a tie, YMMV.

Cycles: LiFePO4 will last 4x the cycles of LiPo, clear winner.

Cost: At first blush, LiPo appears cheaper. However, you would need to replace the packs (4) times to match the cycle life of the a123's, at a cost of $1,591.20, more than double the cost. Some might argue that the cost of entry is cheaper with LiPo, but it looks like you have to buy additional hardware to charge and balance it, I use a $50 Vicor Multipack PS to charge my a123 packs.

Vendor: a123 is quality powder but likely b-stock cells. They have performed to spec and the only failure was due to my inattention. I don't know who makes the HK lipo, and there are reportedly a percentage of DOAs, puffys, and short-lived packs.

================================
Summary:

Instead of lipo being significantly lighter, both packs weigh roughly the same. Lipo is truly 37% smaller than a123, but costs twice as much for the same cycle life. LiFe has an inert cathode that is not succeptable to thermal runaway, and it can be charged more simply without balancers.

Comments welcome, please be kind - this demonstrates the core logic I have based my current opinion on, and if there is a flaw in it I am open to learning. If you change my mind on a data point, you could change my LiPo opinion.

-JD

 

[neptronix]

Sorry, but you got a mistake in there. Hobbyking's page states:

Pack Weight: 793g

The rated charge and discharge voltage are 4.2v/cell to 3.0v/cell. Nominal voltage is 3.8v/cell.
( in real life you want to stop discharging at 3.3v/cell, at that point, you have about 1% SOC left. )

Also, are those 16AH cells still in production? If not, maybe a comparison to a cell you can actually purchase these days is best. 20AHs are popping up..

 

[oatnet]

Sorry, but you got a mistake in there. Hobbyking's page states:

Pack Weight: 793g

The rated charge and discharge voltage are 4.2v/cell to 3.0v/cell. Nominal voltage is 3.8v/cell.
( in real life you want to stop discharging at 3.3v/cell, at that point, you have about 1% SOC left. )

Cool I'll update that. What do you think is a real-life voltage per cell under a 35a load?

-JD

 

[oatnet]

Also, are those 16AH cells still in production? If not, maybe a comparison to a cell you can actually purchase these days is best. 20AHs are popping up..

The 20ah cells are even lighter, so the 16ah pack favors lipo more. Plus I have direct experience with the 16ah cells, and a pack still in my hands I can refer to. My 20ah cells are still in the box.

 

[oatnet]

Sorry, but you got a mistake in there. Hobbyking's page states:

Pack Weight: 793g

Cool I'll update that. What do you think is a real-life voltage per cell under a 35a load?

-JD

I see two weights listed!
Under the product config table I see WEIGHT: 913g STOCK -14 QTY....

At 793g, the equivalent LiPo pack would weigh 15.75lb, 2.88 lb less, and 15.5% lighter.

If someone could weigh a configured pack of 3-4-6 of these 6s packs , including wire harness/fuses etc like my LiFe pack, I'd be glad to use real-wolrd data instead.

-JD

 

[neptronix]

Cool I'll update that. What do you think is a real-life voltage per cell under a 35a load?
-JD

Funny you ask, i have ran 36 amps peak on 5ah 20C 5S for a while...

If i remember correctly, voltage drop is 2.75v across a 10S 5ah pack, so that's 0.275v of voltage drop per cell.
Internal resistance is average 2 mOhm per cell, according to the iCharger anyway.

On 20AH with a 2C load, i get about 0.8V of drop for the 10S pack, thus 0.08v drop per cell.

Now an equivalent A123 pack would have 2 more cells, so the calculations a bit different.

Hope this data helps make the comparison a bit more fair.

BTW you're probably looking at a shipped weight VS real weight for the lipos, you should assume the lower value. My 5s packs are listed as 666g and weight just about that, so i'd take the lower weight and assume it's correct. ( plus some of the weight is wrapping, wires, circuit board, solder etc etc.

 

[oatnet]

Funny you ask, i have ran 36 amps peak on 5ah 20C 5S for a while... If i remember correctly, voltage drop is 2.75v across a 10S 5ah pack, so that's 0.275v of voltage drop per cell.

Now an equivalent A123 pack would have 2 more cells, so the calculations a bit different.

Is the .275 off 4.2v peak charge, yielding 3.925v/cell? Or is resting voltage at 10%DOD a little lower?

I listed real-world data on the a123 pack, so the calculation was simply dividing voltage by 20 cells.

 

[Hillhater]

The main issue i see with the larger 16-20Ahr cells is the real world availability.
It seems to be a major problem finding a trustworthy supplier currently !

 

[mr.electric]

Cycles: LiFePO4 will last 4x the cycles of LiPo, clear winner.

Cost: At first blush, LiPo appears cheaper. However, you would need to replace the packs (4) times to match the cycle life of the a123's, at a cost of $1,591.20, more than double the cost.
-JD[/quote]
You have extensive experience with LiFePO4. Have you ever seen a pack get to 1500 cycles. I have found that some other factor inevitably kills a pack before it actually reaches very high cycles. Bms failure, wiring issue, physical damage, water damage, vibration breaking wires etc. I would be happy with 500-1000 cycles. Plus I think the cells run out of calendar life relatively quickly. 2000 cycles would be a full discharge charge cycle every day for over 6 years. That ebike would have gone through countless electrical components controllers, motors throttles and mechanical components tires, wheels, bearings, chains, even seats.
I agree that good cycle life is important but I think lipo and lifepo4 can both match the life time of the majority of the other ebike components.
If you built a car engine capable of 500k miles few people would take advantage of it. The car would hit the junk yard after 150 -200 k because of the various other compiled issues. Also people inevitably upgrade because of technology improvements even if the old stuff still works.

 

[EBJ]

I'd like to see some performance specs. Like charging and discharging rates.
I hear a123 can put out something close to 30C, but I haven't seen a BMS that would allow it to do so.
Most of the BMS I see tend to be limited to 60-80A continuous. What BMS are YOU using? and what is the discharge rating ?

I think you should also include the price of the chargers required. And maybe some BMS options for either battery system.
Example: Good BMS w/ cheap charger = $ ? No BMS w/ great charger = $ ?
for both Lipo and a123

 

[Red_Liner740]

I dunno where this 300-500 cycles for Lipo comes from....maybe if u do 4.2V to 3.0V cycles....i'm sure there are tons of people that top off at a lower voltage and end discharging at 3.6V (80% discharge) that get a ton more cycles than 300.

 

[neptronix]

Is the .275 off 4.2v peak charge, yielding 3.925v/cell? Or is resting voltage at 10%DOD a little lower?

Very good question.
That measurement was taken somewhere between 4.0 and 3.8v.

Below 3.6v/cell, the sag gets pretty extreme.

 

[lesdit]

I'm wondering what your method was to rate the capacity of the 16ah a123 cells ? Can you describe the voltage limits used, and if a particular charge technique was employed ?
Thanks

 

[oatnet]

The main issue i see with the larger 16-20Ahr cells is the real world availability.
It seems to be a major problem finding a trustworthy supplier currently!

It is a major problem finding a trustworthy, long term supplier for any ebike commodity... Stock comes and goes - even the very Turnigy pack I am comparing to is out-of-stock at the moment, 14 on backorder. Plus both the Turnigy and the a123 16ah are prisimatic, whereas the M1 cell is cylindrical with additional packaging, so it is not a good comparison.

Further, I have been reading the threads here, where people are getting a123 prisimatics from Mavizen (spelling?), FalconEV, and other suppliers. I purchased mine from cell_man in (3) purchases, spread out over more than a year, so from my perspective they have been commonly available every time I have gone looking. Further, cell_man recently posted that he might have supply of the a123 prisimatics in the future.

Plus I think the cells run out of calendar life relatively quickly. 2000 cycles would be a full discharge charge cycle every day for over 6 years. That ebike would have gone through countless electrical components controllers, motors throttles and mechanical components tires, wheels, bearings, chains, even seats.

You are correct, my packs outlive ebike mechanicals. Then I move them to a new ebike, and I don't have to buy a pack for the new ebike, that saves the money I mentioned. Heck, I am still getting use out of the Ocean Cheer pack I bought in 2007, which was the first LiFePO4 pack published on Endless-Sphere. It has been used on a whole bunch of builds, although it is so 1st-gen heavy, I really just use it on the bench as a test pack these days.

LifePO4 has not been around ebikes long enough to answer the calender question with empirical evidence. I have some M1 packs I built 3 years ago (from USED dr. bass dewalt cells), haven't cycled that often, but they still are yielding the anticipated watt hours. I also seem to remember reading an a123 document that projected a 10-year calendar life, I wonder if I can find that again.

However, remember that cell degradation is an continual process, you don't hit 300-500 cycles and your lipo falls of a cliff. after 150-250 cycles, lipo cells are half-degraded, a123 is just getting broken in.

If you built a car engine capable of 500k miles few people would take advantage of it. The car would hit the junk yard after 150 -200 k because of the various other compiled issues.

Disagree. At 200k, the car would be made into a NYC taxi. If there are motors in the junkyard with life in them, people will pull them out and use them. Plus folks like Luke would buy them out of the junkyard, and because the engine was so overbuilt to last 500k, he'd tweak it to pull 2000hp and 1500ft-lbs of torque and put it on a Moped with helium in the tires. :lol:

I'd like to see some performance specs. Like charging and discharging rates.
I hear a123 can put out something close to 30C, but I haven't seen a BMS that would allow it to do so. Most of the BMS I see tend to be limited to 60-80A continuous. What BMS are YOU using? and what is the discharge rating ?

30c performance is an interesting subject, but I want to keep this topic on track. a123 specs have been discussed to death here - with a few clicks of the E:S Search button, you will be able to satisfy your curiosity.

However, when you dishcarge at 30c, you drain the pack in 2 minutes. How far can you commute in 2 minutes? My personal belief is that c-rating is largely irrelevant in EV's - unless you are building an exhibition bike that only runs for 2 minutes. I like my bikes to be able to run for at least a half hour - that is a 2c continous discharge rate. By the time I add enough AH to give the bike useful range, the c-rating is not a factor.

BMS' continue to be problematic for all chemistries, I think the BMS' being developed by some of the wizards here on E:S are probably the best bet. I have never run a BMS on LiFePO4 because it is not at risk of Thermal Runaway when overcharged/discharged. I use a CA to track wh used, and that has worked flawlessly for me. I did recently add a 4s pack that was half charged to a 18s pack that was fully charged, and toasted 2 cells - but that was an installation error, not an operational error. A BMS would have cost me more than those two cells.

I think you should also include the price of the chargers required. And maybe some BMS options for either battery system. Example: Good BMS w/ cheap charger = $ ? No BMS w/ great charger = $ ? for both Lipo and a123

I did it for a123, maybe someone with direct experience can detail the components for LiPo. Nowadays I use $50 Vicor Megapack Power supplies for charging. When I build a pack, I parallel all the cells and charge them to 3.65v, and let them rest; repeat a few times. Then I assemble the pack, don't overdischarge it, and it stays pretty much balanced. I may re-balance it when I move it to a new ebike, I have a bank of voltfreaks single-cell chargers for 2a/cell balancing, and an aluminum plate with vicor DC-DC converers for 20a/cell balancing.

I dunno where this 300-500 cycles for Lipo comes from....maybe if u do 4.2V to 3.0V cycles....i'm sure there are tons of people that top off at a lower voltage and end discharging at 3.6V (80% discharge) that get a ton more cycles than 300.

I am happy to update that number with real-world data if someone with direct experience will quote it, but those are the numbers I have seen discussed on LiPo threads here, and R/C threads elsewhere. Further, I left off an important pro-LiFe data point - I see a lot of reports of lipo with 50+ cycles in getting puffy or low v cells - I have not seen a corresponding fail rate in cylindrical or prisimatic a123's.

Very good question. That measurement was taken somewhere between 4.0 and 3.8v

Given that I prefer LiFe, I tried to be conservative, and it sounds like the 3.83v-under-discharge I used for lipo is actually too high, swaying the numbers pro-lipo. I know we have a lot of highly experienced folks who have been promoting lipo here, hopefully one of them can join the thread and help me ballpark that number better.

I'm wondering what your method was to rate the capacity of the 16ah a123 cells ? Can you describe the voltage limits used, and if a particular charge technique was employed ?

The cell_man cells come with the yield written on them. I tested one on a CBA, probably set the low end at 2.2v, and cutoff the test when it reached the cell_man rating before reaching 2.2v. I did another test where I pulled 16ah from a 16s pack, rated with a CA-SA and its calibrated shunt. I do bulk charging and I probably pulled it off the charger and went for a ride. I think Luke and some other folks took a more scientific approach with the 20ah cells.

-JD

 

[Hillhater]

Where can you currently buy genuine A123 , 20Ahr pouches ? :?

 

[neptronix]

C rate does matter. It means that the cells perform better even at lower discharge rates. If i run my cells at 10C constant, they perform horribly even though they are way under spec. They just get warm and turn battery charge into heat.
It adds up if you are using smaller packs, like for example if you wanted to drag race etc.

I have heard of people retiring their lipo packs at 600 cycles if treated well.
In regards to that, i worry about calendar life more than cycle life since i am a fair weather ebiker and not a daily rider. Thus, A123 makes no sense to me, other than for the safety factor.

 

[EBJ]

Where can you currently buy genuine A123 , 20Ahr pouches ? :?

x2

 

[jonathanm]

This thread confirms my current thoughts - as a tinkerer, LiPo is fun and exciting, if I was building anything commercially, or even building something for a not technical friend, it's probably not the way to go......yet.

 

[EBJ]

This thread confirms my current thoughts - as a tinkerer, LiPo is fun and exciting, if I was building anything commercially, or even building something for a not technical friend, it's probably not the way to go......yet.

I would say it depends on the product being sold. If it's advertised as High-performance and it's dangerous to begin with, then i don't see a problem w/ using lipos.
Example: Extreme electric dirtbike
Some things have an inherent risk, I think it's just that most people don't associate "electric bicycles" as a risky sport. If it was turned into an "extreme" sport, then I'm sure we would see more being sold with Lipos. *and a million warning stickers all over it. haha.

 

[jonathanm]

I would say it depends on the product being sold. If it's advertised as High-performance and it's dangerous to begin with, then i don't see a problem w/ using lipos.
Example: Extreme electric dirtbike
Some things have an inherent risk, I think it's just that most people don't associate "electric bicycles" as a risky sport. If it was turned into an "extreme" sport, then I'm sure we would see more being sold with Lipos. *and a million warning stickers all over it. haha.

Well that's true enough......I can just imagine selling high powered lipo dirt bikes and my lawyers and investors making customers sign a 10 page disclaimer at time of purchase, lol. A motocross bike or a rice rocket is a pretty dangerous thing and they sell plenty of them......but having said that they are only dangerous through human intervention.....the LiPo can be dangerous through *lack* of human intervention....big difference......

 

[EBJ]

I would say it depends on the product being sold. If it's advertised as High-performance and it's dangerous to begin with, then i don't see a problem w/ using lipos.
Example: Extreme electric dirtbike
Some things have an inherent risk, I think it's just that most people don't associate "electric bicycles" as a risky sport. If it was turned into an "extreme" sport, then I'm sure we would see more being sold with Lipos. *and a million warning stickers all over it. haha.

Well that's true enough......I can just imagine selling high powered lipo dirt bikes and my lawyers and investors making customers sign a 10 page disclaimer at time of purchase, lol. A motocross bike or a rice rocket is a pretty dangerous thing and they sell plenty of them......but having said that they are only dangerous through human intervention.....the LiPo can be dangerous through *lack* of human intervention....big difference......

With things like Gary's ONE-PLUG charging using the Hyperion charger, LVC's that cut throttle inputs and Alarm out, and fire-proof bags for safety, Lipo's become much harder to "mess-up" and much more simple to use. I imagine in such a situation manufacturer defects could be the only thing to worry about. "just follow the rules and no one gets hurt"

 

[jimw1960]

Where can you currently buy genuine A123 , 20Ahr pouches ? :?

x2

If you want pristine cells, a123rc.com was a reliable source for a good price from China, but they seem to have stopped selling the 20AH cells. They do have the 15Ah cells, which might be better for some folks because they are smaller and lighter. Free shipping, too.
http://a123rc.com/goods-473-Excitingly+Powerful+A+123+15ah+PRISMATIC+CELLS.html

If you don't mind B-grade cells with the tabs cut, there is an ebay seller from China who has the 20Ah pouches for like $28 per cell including free shipping. Check out his ebay ads; he shows you how to terminate them without the tabs (good enough for an ebike):
http://shop.ebay.com/xueming0061/m.html?_trksid=p4340.l2562

If I may rant, it bugs me that you have to buy these cells from China when we have a perfectly good A123 factory right here in the USA making them.

 

[Hillhater]

Thanks jimw, its good to know there are sources, even if that are not "pristine" and tab-less .
do you know why they show two distinctly different 20Ahr cells ?..
the APP72161227, and the AHP70165227 ... ??
These are obviously different physical sizes, but why the two versions ?...both genuine A123 ??
Which is the most recent , ...which is "preferred" ?

 

[EBJ]

Where can you currently buy genuine A123 , 20Ahr pouches ? :?

x2

If you don't mind B-grade cells with the tabs cut, there is an ebay seller from China who has the 20Ah pouches for like $28 per cell including free shipping. Check out his ebay ads; he shows you how to terminate them without the tabs (good enough for an ebike):
http://shop.ebay.com/xueming0061/m.html?_trksid=p4340.l2562

If I may rant, it bugs me that you have to buy these cells from China when we have a perfectly good A123 factory right here in the USA making them.

I also wonder why the seller states only a 3C discharge 10C burst ? I was under the impression a123 had awesome 30C discharge ratings?
The ratings he is giving for these seem closer to the Headway LifePo4 cells. Perhaps a safety spec to save his ass from people demanding performance from "neutered" cells ??

 

[jimw1960]

I also wonder why the seller states only a 3C discharge 10C burst ? I was under the impression a123 had awesome 30C discharge ratings?
The ratings he is giving for these seem closer to the Headway LifePo4 cells. Perhaps a safety spec to save his ass from people demanding performance from "neutered" cells ??

I think they are downplaying the C-rate on these because of the tabs. The connectors he is using to terminate them are only good for limited amps. If you tried to put 30 C through those ring terminals, it would probably cook the terminal in short order.