pgt400
1 kW
Very nice work! Everyone appreciates your efforts!
Yesa vs Headway vs PSI vs Ebike.ca 12Ah LFP test graph
- Thread starter Doctorbass
- Start date
voicecoils
1 MW
Interesting so, while at 1A or 35A Headway exhibits greater voltage drop then PSI, less flat holding of voltage throughout discharge and higher cell heating, it still manages to deliver more energy in each case.
PSI and Yesa's curve characteristics look so similar I wonder if they use the same cathode powder and manufacturing techniques.
PSI and Yesa's curve characteristics look so similar I wonder if they use the same cathode powder and manufacturing techniques.
Hmm, so if I read these charts correctly, Doc, if the Headway has ~9 milliohms internal resistance, these results indicate the Yesa and PSI have about 6 milliohms internal resistance. It is interesting to see what a 3.5C discharge does to make this apparently small difference seem important.
Willie
Willie
Ypedal
100 TW
Awsome work Doc !!! keep it up ! ( curious to know what happens at the 50,60,70 amps range !!! 
) !
) !
Mr. Mik
1 kW
- Joined
- Sep 3, 2008
- Messages
- 390
I'd be careful when using cells which were specifically supplied by vendors for testing purposes.
How much this applies to Lithium battery technology is something I'm not absolutely certain about, but I have read many posts about how batteries get hand picked. This happens first at the factory, to separate the production run into batches for selling as expensive "Prime Quality" to special customers, and into normal stock or cheap crap.
Then they get sorted again by RC racing enthusiasts who want to match the cells for their packs as closely as possible; some take their testing equipment to the shop and test away until they have selected the best cells out of the entire heap on the shelves!
This is all due to the significant differences between individual cells.
A manufacturer who just sends a randomly picked cell to someone like Doctorbass would need their head read!
So I would not call it fully independent testing unless the cells tested were bought through the usual channels!
How much this applies to Lithium battery technology is something I'm not absolutely certain about, but I have read many posts about how batteries get hand picked. This happens first at the factory, to separate the production run into batches for selling as expensive "Prime Quality" to special customers, and into normal stock or cheap crap.
Then they get sorted again by RC racing enthusiasts who want to match the cells for their packs as closely as possible; some take their testing equipment to the shop and test away until they have selected the best cells out of the entire heap on the shelves!
This is all due to the significant differences between individual cells.
A manufacturer who just sends a randomly picked cell to someone like Doctorbass would need their head read!
So I would not call it fully independent testing unless the cells tested were bought through the usual channels!
Doctorbass
100 GW
headway said:
Hello Victoria,
That will be a pleasure to me to test your 8Ah version.
As well for the 12 or 15Ah too if it's possible.
As i described I always test the best cell of each brand i receive so you can send me just one or preferably 3 samples so i will be able to post the best possible resusts of your products.
Doing a mean result of a batch of sample would take alot of time for testing all items!
so that's why i prefer keeping the best sample of the batch for the posted tests. Just pm me for the 8Ah cell..
Doc
dnmun
1 PW
mik, we called it 'binning' where microprocessors at test would be selected and binned according to the max speed at which they would run. there is always some variability in the process and some parts would run faster than others. i don't think we need to worry that they could select an 'extraordinary' cell.
if they did that then later testing of other cells might show lower performance and then people would worry that they had lost control of their manufacturing process. so i doubt if they would do that in any case. i am so glad they have started tracking the individual cells with the bar codes, i think overall they are doing the right stuff to get the performance up there.
what seemed most important is that they cut IR in half from 17 to 9mohm, and cut 83% of the sag from the older brown cells. at 3.5 C the brown cells were at 2.47V but now 2.95V, and the psi and yesa formulation is 3.05V. so the removal of sag is much greater than the change in the internal resistance. this is not something we expected, thinking IR correlated directly with sag. maybe it does but the scale is not apparent yet. maybe the absolute minimum would be around 2-3mohm.
there does appear to be a relationship between a higher IR and the larger capacity. as though the sag reflects the difficulty the lithium ions have in getting through the matrix of granules in the cathodic structure, so they don't get used up early in the discharge cycle. but we would have to be comparing the same volumes or weight of cathode granules, imo.
if they did that then later testing of other cells might show lower performance and then people would worry that they had lost control of their manufacturing process. so i doubt if they would do that in any case. i am so glad they have started tracking the individual cells with the bar codes, i think overall they are doing the right stuff to get the performance up there.
what seemed most important is that they cut IR in half from 17 to 9mohm, and cut 83% of the sag from the older brown cells. at 3.5 C the brown cells were at 2.47V but now 2.95V, and the psi and yesa formulation is 3.05V. so the removal of sag is much greater than the change in the internal resistance. this is not something we expected, thinking IR correlated directly with sag. maybe it does but the scale is not apparent yet. maybe the absolute minimum would be around 2-3mohm.
there does appear to be a relationship between a higher IR and the larger capacity. as though the sag reflects the difficulty the lithium ions have in getting through the matrix of granules in the cathodic structure, so they don't get used up early in the discharge cycle. but we would have to be comparing the same volumes or weight of cathode granules, imo.
oatnet
1 MW
Doctorbass said:
Doc, thanks for working so long and hard to put this hard data out there for us. While each cell has tradeoffs that suit different applications, the comparisons I have seen so far makes me feel good about the Headway purchases I have made for mine. Yesa has also gone up a couple of notches in my regard.
Has anyone supplied you with current Lifebatt cells? I keep reading posts about quality justifying their high cost but I've seen nothing but Distributer Propoganda "testing", I'd be interested to see just how they actually compare with the the other cells in your impartial discharge test. Ping, and ebikes.ca cells would be interesting too, too bad T-Sky doesn't offer their large-format chemistry in a 10a cell so you could overlay that too.
Hmm, after all your hard work I am eager to give you more, on cells I don't even own.
Sorry about that, thanks for the data!-JD
Ypedal
100 TW
I also sent an " Ebikes.ca " 12ah prismatic LFP cell.. :wink: ( no rush Doc, don't neglect the lady too much or she may lock the lab on you !! )
oatnet
1 MW
Ypedal said:
Ypedal, you are the one person here who might know - are those Ebikes.ca cells the same as the green ones we got from Andy in '07? I thought they looked very similar... those cells have served me very well - although a killed (2) of them through negligence, when I overheated/killed their voltphreaks single-cell chargers and didn't have an LVC to realize it. Now I keep a gap between the chargers
and they do just fine. If the cells are the same maybe I'll buy some replacements from Justin, just to maintain the old x5tracycle in full repair.-JD
PS - Here is thread on those green cells:
http://www.endless-sphere.com/forums/viewtopic.php?f=14&t=2503
PPS: The wood box they arrived in just saw new life (no pun
) holding the vacuum-assist system for my VW conversion.
Ypedal
100 TW
exactly the same..
I don't think ebikes.ca are re-ordering them however ( could be wrong ), but they may have some cells or leftover packs to unload if you email them.
I have a few dead one's too.. when FalconEV sent me this pack, 2 of the cells in parallel were at 0v .. I recharged them but they only lasted a few cycles.. so i'm salvaging cells from one pack . powering my chopper with these and a local customer has some on a cyclone system.
I don't think ebikes.ca are re-ordering them however ( could be wrong ), but they may have some cells or leftover packs to unload if you email them.
I have a few dead one's too.. when FalconEV sent me this pack, 2 of the cells in parallel were at 0v .. I recharged them but they only lasted a few cycles.. so i'm salvaging cells from one pack . powering my chopper with these and a local customer has some on a cyclone system.
oatnet
1 MW
Ypedal said:
Wow. Thanks for the tip!
Thanks to opportunity charges that got me through the day, mine have probably seen 600-900 cycles, including the 44.7mph 90ah discharge test on my Kelly, with no problem but me. I don't ride the bike they are currently mounted on very much these days (after cracking a rib in a fall on a prototype I favor my TidalForce bikes for their balance), but I was still pulling the same amount of WH from them without seeing the voltage drop at the end of the ride.
-JD
Doctorbass
100 GW
Ypedal.. Next test on the list: 12Ah green cell from ebikes.ca!
This is a great luck i have to get acces to every of these different cells in my lab.. I admit thatfor sure, these tests will be helping me to decide on wich i'll choose for my electric car project!
That's why i pay attention to every details about these test! I want a great test job for me too!!
The 12Ah cell is begining the first charge for the first 0.1C discharge!
Doc
This is a great luck i have to get acces to every of these different cells in my lab.. I admit thatfor sure, these tests will be helping me to decide on wich i'll choose for my electric car project!
That's why i pay attention to every details about these test! I want a great test job for me too!!
The 12Ah cell is begining the first charge for the first 0.1C discharge!
Doc
Deepkimchi
1 kW
Yeah - make sure we distinguish between Andy at FalconEV and AndyH
The FalconEV 12 AH cells individually do well, but the solid copper wire connecting the cells was not a good idea, vibration makes the terminals loosen. Changing this out with stranded wire to see if it helps.
DK
The FalconEV 12 AH cells individually do well, but the solid copper wire connecting the cells was not a good idea, vibration makes the terminals loosen. Changing this out with stranded wire to see if it helps.
DK
Toshi
10 kW
awesome. i want to know how my pack rates on your testbench.Doctorbass said:Ypedal.. Next test on the list: 12Ah green cell from ebikes.ca!
This is a great luck i have to get acces to every of these different cells in my lab.. I admit thatfor sure, these tests will be helping me to decide on wich i'll choose for my electric car project!
That's why i pay attention to every details about these test! I want a great test job for me too!!
The 12Ah cell is begining the first charge for the first 0.1C discharge!
Doc
Doctorbass
100 GW
Deepkimchi said:
I saw this problem on the Yesa 48V 15Ah pack too. they sent me a pic of the upgraded pack with the upgraded BMS wires but the link connecting the cells still the same weak desing.. they forgot to put a locknut bettwen the nut and the cooper link!
Doc
oatnet
1 MW
Deepkimchi said:
Yeah, I thought those wire things were strange and had a small contact patch, I built my own connectors from copper strap and with a little locktite never had a problem:
http://www.endless-sphere.com/forums/viewtopic.php?f=14&t=4161
Doctorbass
100 GW
Ypedal said:Awsome work Doc !!! keep it up ! ( curious to know what happens at the 50,60,70 amps range !!!
About the rest of the test for current above 35A, I just wait to get the 40x original IR4110 from Bob I'll built an electronic load with CC discharge to help the CBA to get higher discharge rate!
Otherwise I'll need to discharge at constant resistance mode with my large 0.15ohm load in parallel.. and i may not be able to overlay the results on the existing graph with 1 to 35A to gatr a great comparaison..
Coming soon!
dnmun
1 PW
yes do it, somebody will photoshop it together, pins and needles.
after thinking about the headway cell at 3.5C, and why the internal resistance did not change as much as the voltage sag has decreased, i wondered about another test you could attempt.
if you can measure the internal resistance of the cell at large differences in temperature, we may be able to tease out that part of the equation. at lower temps the ion mobility would be constrained by reduced rates of diffusion of the ions through the electrolyte. versus some quality or structure of the anode not being affected by temperature. internal resistance may not change, but it would be interesting to know.
it should not damage it to freeze it, maybe -20C, and 70C might be too late in the year to get -20C outside. does that make sense or seem worth knowing?
after thinking about the headway cell at 3.5C, and why the internal resistance did not change as much as the voltage sag has decreased, i wondered about another test you could attempt.
if you can measure the internal resistance of the cell at large differences in temperature, we may be able to tease out that part of the equation. at lower temps the ion mobility would be constrained by reduced rates of diffusion of the ions through the electrolyte. versus some quality or structure of the anode not being affected by temperature. internal resistance may not change, but it would be interesting to know.
it should not damage it to freeze it, maybe -20C, and 70C might be too late in the year to get -20C outside. does that make sense or seem worth knowing?
GGoodrum
1 MW
Great stuff, Doc, as usual. One thing I might suggest, for readability sake, is to not overlay quite so much on one chart. It might be a lot easier to read if you had one chart for the 1C tests, another for the 35A tests, etc.
I'm realy looking forward to seeing how the new 12 and 15Ah Headway cells do, along with the new 8Ah as well. Victoria, are these all the same form factor, or are the 12/15Ah versions larger? What is the C rating for these new cells?
-- Gary
One thing I might suggest, for readability sake, is to not overlay quite so much on one chart. It might be a lot easier to read if you had one chart for the 1C tests, another for the 35A tests, etc. I'm realy looking forward to seeing how the new 12 and 15Ah Headway cells do, along with the new 8Ah as well. Victoria, are these all the same form factor, or are the 12/15Ah versions larger? What is the C rating for these new cells?
-- Gary
ZapPat
10 kW
Yes indeed the internal resistance does change with temperature, and we even loose much capacity at -20oC. I've seen a few A123 and other cell discharge curves at different temps to see just how much the internal resistance does get worse as it gets cold.dnmun said:
The cure for this is to have removable packs, and keep them warm until needed. Also have some insulation in your battery compartment for cold times, and/or just use them hard enough so they generate their own heat to keep their performance up. Dont forget the insulation on hot days and hard (high-amp) runs though, since getting real hot is not good for cell life.
dnmun
1 PW
i'm not disputing anything pat, i just wondered if it was possible to calculate the IR for me at extremes of temperature. since he has 3 different chemical formulations at the same age, and history, compared on the same equipment. maybe to establish a baseline for later lifepo4 formulations.
in his original headway thread, i was able to see that IR almost tripled from initial discharge to close to terminal discharge using RLT's graphs. and i could do that too from these graphs like i did last year, but doc wants to use the IR test he established and i thought that was ok since he has good experimental technique and i think the numbers will stick for a long time.
just curious about how lifepo4 works, and why one formulation or cell type maybe is better in different situations.
and i am also curious about the charge cycle, what is the equivalent to internal resistance then? i always think about the review i read where they were talking about the theory of the supersaturating phenomena where the lithium is precipitating out of the electrolyte separated from the anode, not touching it, that has always interested me. what other electrolytes could be used sorta thing.
anyway, that's why i asked, now i included a request for IR over the discharge cycle so it doesn't make it easier for him, plus the high current runs he wants to do.
in his original headway thread, i was able to see that IR almost tripled from initial discharge to close to terminal discharge using RLT's graphs. and i could do that too from these graphs like i did last year, but doc wants to use the IR test he established and i thought that was ok since he has good experimental technique and i think the numbers will stick for a long time.
just curious about how lifepo4 works, and why one formulation or cell type maybe is better in different situations.
and i am also curious about the charge cycle, what is the equivalent to internal resistance then? i always think about the review i read where they were talking about the theory of the supersaturating phenomena where the lithium is precipitating out of the electrolyte separated from the anode, not touching it, that has always interested me. what other electrolytes could be used sorta thing.
anyway, that's why i asked, now i included a request for IR over the discharge cycle so it doesn't make it easier for him, plus the high current runs he wants to do.
Doctorbass
100 GW
dnmun said:
Dnmun, this is faisable but that would take alot of time!!!
.. test over cycles are ... well... hmmm... are... VERY LONG TO ACHEIVE !!!the best would be to test them with normal ebike use and to know the mean current of use... that could happen for me when snow outdoor will have dissapear!
Doc
dnmun
1 PW
ok,using my graphical interpretation on doc's CBA graphs to evaluate internal resistance. in milliohm.
manufacturer---1.11Ah--5Ah--8.85Ah
PSI-----------------6.8-----7.9----16.5
YESA---------------7.6-----8.2----15.6
Headway---------11.2----11.5----15.6
LifeBatt----------23.2----23.2----25.3 (added by edit using the lifebatt graph at 20A from bob's old lifebatt thread)
first 3 done at 22C by doc. if anyone gets some other numbers at different temperatures, try to keep track of where in the discharge it is so we can interpolate, thanks, dm
manufacturer---1.11Ah--5Ah--8.85Ah
PSI-----------------6.8-----7.9----16.5
YESA---------------7.6-----8.2----15.6
Headway---------11.2----11.5----15.6
LifeBatt----------23.2----23.2----25.3 (added by edit using the lifebatt graph at 20A from bob's old lifebatt thread)
first 3 done at 22C by doc. if anyone gets some other numbers at different temperatures, try to keep track of where in the discharge it is so we can interpolate, thanks, dm
oatnet
1 MW
dnmun said:
I had forgotten about that LifeBatt graph. I really want to see Lifebatt/BMI on the same graph in comparison, but maybe this will do for the moment. Here is my crude 20aCBA headway test, compared to Bob's 20a graph using similar hardware:
View attachment 38120 HH23 11572 20a-2c_shrunk.JPG
I was babying the new cell and terminated at only 26v which makes the curves harder to compare, and the lifebatt has a lower nominal voltage, but...
The headway discharged 20a at 3.0v, at 6ah it has dropped .09v, and then trails off .4v to 2.6, where it has supplied 9.19ah.
The lifebatt- discharged 20a at 2.9v, at 6ah it has dropped .08v, and by the time it trails off an equivalent .4v it has supplied @9.75ah.
So based on these crude tests, I theorize that the lifebatt delivers a slightly flatter curve and slightly more ah, but at lower voltage meaning it probably delivers about the same wh as the headway. I know the lifebatt starts at a lower voltage, but if I remember correctly the headway was about 340g, I googled posts describing the lifebatt as 360g, so by weight/volume the headway is delivering slightly better voltage and probably equivalent wh. granted, this is only 2c, I wonder how the voltages compare at 5c.
So please bear with me as I reason out the pros and cons, as I see them, for MY application, a large-vehicle EV Conversion.
The Headway is rated at 5c continous, and the LifeBatt is double that for 10c. For my Bus, I really need at least 80ah/cell simply to have enough KW for a meaningful distance. 80ah at 5c that delivers 400a, more than enough to supply my 500a controller, so I don't really need to pay a premium to get 10c. For vehicles like ebikes and electric motorcycles that can only carry smaller ah packs, but still need high discharge rates, the lifebatt sounds like a good fit - if the packaging is compact enough.
The cycle life of the pack is another issue. A 120v80ah pack yields 9.6kw/cycle, at 350whpm that will deliver 27.43 miles, more than enough for my daily travels. Headway estimates 1400+cycles, so even if I used a full cycle every day, it should be good for 4 years/38,400 miles. LifeBatt claims 3000 cycles so it could be good for 8.5 years of daily full cycles, which is far superior, but for MY application I'm happy to save the money now and replace them with whatever hot new technology is around in 4 years.
Which gets us to cost. One can buy a 320 cells from Victoria to build a 120v80ah pack, shipped air to your door, for $4,960. For comparison to Lifbatt's products, a 144v80ah pack would be $5,952. Lifebatt's 144v10ah pack lists at $3,000, you would need eight of them to build an equivalent 80ah pack, or $24,000 before any 'volume' discounts. ( http://www.lifebatt.com/retail_sheet.asp - 14410-HPS 144V / 10Ah w/VMS, 2 Terminals 24.5 546 182 207 53.9 21.50 7.17 8.15 48 $3,000.00 )
So the way I look at it, LifeBatt's have twice the cycle life, have twice the c-rating, and include nice packaging/monitoring/charging. Headway's provide slightly higher voltage per cell, weigh slightly less, and cost 1/4 as much as LifeBatt, but you need to package them yourself - a learning experience I happen to enjoy and prefer
-JD
