Yesa vs Headway vs PSI vs Ebike.ca 12Ah LFP test graph

[Doctorbass]

While doing many measurement with the Yesa cells I decided to compare the results i got with the Headway and the Yesa in extreme conditions (1A and 35A)

I was very surprized to see how much the end of discharge curve could reveal as large difference as this!! and.. the temp increase of the Headway compared to the Yesa at 35A !!!..

The Headway seems having more capacity than the Yesa but with more Vdrop at high current..

For sure these cella are only One sample of each brand and could not reveal a perfect comparison.. but that give us a great idea... both cells was brand new when tested...

I should finish the PSI cell test sent by Ypedal tomorrow so i'll add this cell test in overlay to compare it to the rest..

Ypedal.. how manu cycles they have..?? (I just want to compare Apple with Apple..)

Doc

[Doctorbass]

More details about the YESA test here: http://www.endless-sphere.com/forums/viewtopic.php?f=14&t=8921&p=136533#p136533

Doc

[dnmun]

9mohm -5mohm = 4mohm difference in the internal resistance. at 35A that would be 140mV. looks like you have about 120mV difference so the numbers all seem to be consistent. 3.07V-2.95V at about 20% discharge.

[GGoodrum]

Very interesting, Doc. I thought they'd be closer. I wonder if the prismatic shape of the Yesa cell has anything to do with the difference in heat? Anyway, it will be very interesting to see how both of these stack up against the PSI cell. I have experience with the PSIs but not the Headway or Yesa cells. My sense is that they all should be close, but i don't know.

[Doctorbass]

Very interesting, Doc. I thought they'd be closer. I wonder if the prismatic shape of the Yesa cell has anything to do with the difference in heat? Anyway, it will be very interesting to see how both of these stack up against the PSI cell. I have experience with the PSIs but not the Headway or Yesa cells. My sense is that they all should be close, but i don't know.

Corrected calculation! 3 march 2009

My opinion about heat management and cell shape is still always the same: the more cell shape surface per Watt you have the best it is!

That's why there is many controuversy about small vs bigger cell for EV construction... small cells represent much more work.. but the heat dissipation is better cause heat need to travel less thickness to be dissipated... That's why the Tesla choosed the 2.4Ah 18650 li-Co cells with PTC. 18650 offer more surface per Watt than anyothers...
The pouch cells like ping pack offer large surface but cels are too close to dissipate heat..


in fact the total surface of the:

-Yesa 10Ah cell is: 243.12cm^2
-Headway 10Ah cell is: 165.93cm^2

The Yesa have 32%... more surface



in fact the total Volume of the:

-Yesa 10Ah cell is: 223.72cm^2
-Headway 10Ah cell is: 209.38^2

The Yesa have 6.4%... more volume



The energy density per liter is:

-Yesa: 137.7Wh per liter (30.8Wh at 1C)
-Headway : 153.30Wh per liter (32.1Wh at 1C)

Note:(taking account of the volume required to connect the cell with the screw terminal virtual prism volume and that the Wh are at 1C rate)

Yesa have 10.2% less energy density for 32% more surface.

From now the Yesa seems to demonstrate a better form factor vs energy and heat dissipation. But Russell helped me to put a spotlight right on a parameter that is important and that change that. Prismatic cells are normally installed close together from each other next cells letting no air circulation or room to evacuate heat between cells.. The headway are circular cells.. And the heat can travel between cells easier.

Thanks Russell for this important Notice!

Doc

[oatnet]

Fascinating.

Since the yesa had higher voltage, but the headway had more amp hours, I wonder how the actual watt-hours delivered compare. Could you put a CycleAnalyst between the test batteries and your CBA (as the 3.5c load), then re-run those tests to see how many wh are extracted? Or do you already have a more elegant way of doing that in your well-equipped lab?

-JD

EDIT - I see on your Yesa test page you do have a for more elegant method of deriving wh than the crude test I proposed - is it possible to overlay the yesa and headway wh curves?

-JD

[Doctorbass]

Fascinating.

Since the yesa had higher voltage, but the headway had more amp hours, I wonder how the actual watt-hours delivered compare. Could you put a CycleAnalyst between the test batteries and your CBA (as the 3.5c load), then re-run those tests to see how many wh are extracted? Or do you already have a more elegant way of doing that in your well-equipped lab?

-JD

EDIT - I see on your Yesa test page you do have a for more elegant method of deriving wh than the crude test I proposed - is it possible to overlay the yesa and headway wh curves?

-JD

Here is your answers JD!:

Yesa at 1A : 33.45Wh
Yesa at 35A: 29.12Wh
Headway1A :34.79Wh
Headway35A:29.97Wh


Here is the graph:

[voicecoils]

Very interesting, thanks Doc!

Is it possible to test above 3.5C?

This may be quite good to know for PSI. PSI claims 10C so there should be a considerable difference in voltage sag throughout discharge between Headway & PSI for this, if it is true.

For example, if someone needs 7C cont. discharging it will be good to know whether to have: two headway cells in parallel or one set of PSI? (That comparison would be similar in price! )

[Doctorbass]

Very interesting, thanks Doc!

Is it possible to test above 3.5C?

This may be quite good to know for PSI. PSI claims 10C so there should be a considerable difference in voltage sag throughout discharge between Headway & PSI for this, if it is true.

For example, if someone needs 7C cont. discharging it will be good to know whether to have: two headway cells in parallel or one set of PSI? (That comparison would be similar in price! )

Hello Voicecoil,

I expect to be able to test at 4C or more as soon as i finish the electronic load to upgrade the CBA II to hihgher than 35A.

Doc

[Ypedal]

Qute DrBass " Ypedal.. how many cycles they have..?? (I just want to compare Apple with Apple..) "

The 2 x PSI cells i sent have aprox 10 shallow cycles ( 50 % DOD ) on them, they were used as a boost pack to my 72v Norco for the last 2 weeks before winter hit...

[Doctorbass]

Qute DrBass " Ypedal.. how many cycles they have..?? (I just want to compare Apple with Apple..) "

The 2 x PSI cells i sent have aprox 10 shallow cycles ( 50 % DOD ) on them, they were used as a boost pack to my 72v Norco for the last 2 weeks before winter hit...

Thanks Ypedal.. that's exactly how i already cycled the Headway and Yesa cell so this is nice.. We will really have Apple with apple with APPLE !

Doc

[pgt400]

I would think the PSI cell would blow them away based on their specs. Should hold a higher voltage thru out. A little surprised that the Headway dropped so much voltage at 3.5c...was hoping it would do better. At only $17 each they are still less then half the price of PSI's however.

[Ypedal]

The PSI's were 22.50 $ US each in Taiwan, by the time i paid shipping ( Air , UPS ) duties, brokerage, currency exchange, they ended up 40 $ Canadian per cell.. not that far away from the headway cells..... all things considered.

[dnmun]

these blue sleeved cells are much better than the older brown paper sleeved cells. flip measured 2.47V at 34A last year, and we had about 17mohm IR, so these have improved dramatically, so i am hoping they can still push to improve their cathodic powder and anode coating up to the performance of the top tier cells, such as psi, a123, and yesa. victoria mentioned they will introduce larger capacity too, 14Ah is next i think.

[oatnet]

Here is your answers JD!:

Yesa at 1A : 33.45Wh
Yesa at 35A: 29.12Wh
Headway1A :34.79Wh
Headway35A:29.97Wh


Here is the graph:

Thanks Doc!

I thought the higher voltage would catch the Yesa up in wh, at 1c it is within 1% of the Headway.
So at 3.5c, the Headway yields @3% more wh, and the Yesa yields @3% higher voltage. I think my current vw bus pack layout will help dissipate heat from the headway cells to keep that from being an issue for me.

Hmm ... When formed into a "brick" pack, I'd expect the Yesa would have less exposed surface area than the Headway, I wonder how temps in the center of the packs would compare.

-JD

[Russell]

My opinion about heat management and cell shape is still always the same: the more cell shape surface per Watt you have the best it is!



in fact the surface of the:

-Yesa 10Ah cell is: 243.12cm^2
-Headway 10Ah cell is: 165.93cm^2

That's a difference of 32%...surface


The energy per liter is:

-Yesa: 132Wh per liter
-Headway : 138Wh per liter

Note:(taking account of the volume required to connect the cell with the screw terminal virtual prism volume and that the Wh are at 1C rate)

Yesa have 4.4% less energy density for 32% more surface.

From now the Yesa seems to demonstrate a better form factor vs energy and heat dissipation.

Doc

Doc,

Regarding heat dissipation don't you also need to figure in how the cells will be packaged? The rectangular YESA cells will likely be placed right up against one another however the cylindrical Headway cells will have some air gap between cells. So while an individual YESA cell may shed heat better a fully constructed pack of the Headway cells might actually dissipate heat from each cell more evenly don't you think?

That said I believe I prefer the rectangular YESA cell from an ease of packaging perspective. Two rows of six cells would measure 5.5" (140mm) x 8.0" (204mm) and fit nicely in my rear bag.

Question: Does the 94mm height include the posts?



-R

[Doctorbass]

Doc,

Regarding heat dissipation don't you also need to figure in how the cells will be packaged? The rectangular YESA cells will likely be placed right up against one another however the cylindrical Headway cells will have some air gap between cells. So while an individual YESA cell may shed heat better a fully constructed pack of the Headway cells might actually dissipate heat from each cell more evenly don't you think?

-R

You are totally right Russel, I forgot to take account of that for the prismatic cell placement.

after all.. i only test one cell at a time...

I'll revise the data and update about that .

For the 94mm heai i'll double check tonight and get back to you.

Doc

[Mr. Mik]

For testing purposes you might want to add insulation to the cells, so that you can measure heat production, rather than just the result of heat production minus heat dissipation.

That can give clues for what sort of thermal management or forced air cooling might be needed for a fully constructed pack.

It is also important to continue measuring heat after the end of testing; in my experience the surface of the Vectrix NiMH cells reach peak temperature well after the end of a test (or recharging); that's because it takes time for the heat generated in the cell center to spread evenly.

Depending on physical characteristics of different cells they might also have different "thermal mass"; that means that although they might have the apparent same temperature, one type might contain a lot more heat than the other. That will make a difference when repeated charge/discharge cycles are needed in very warm or very cold weather.
A battery with a higher thermal mass will initially appear to heat up less, but it will also remain warmer for longer, and might accumulate heat if not cooled for extended periods after riding and / or charging.

[pgt400]

The PSI's were 22.50 $ US each in Taiwan, by the time i paid shipping ( Air , UPS ) duties, brokerage, currency exchange, they ended up 40 $ Canadian per cell.. not that far away from the headway cells..... all things considered.

Try and get PSI's for $22 each now that Raymond defected.... Victoria is quoting $16.50 each now, plus $5 shipping each. Thats $21.50 each shipped!

[Doctorbass]

For testing purposes you might want to add insulation to the cells, so that you can measure heat production, rather than just the result of heat production minus heat dissipation.

That can give clues for what sort of thermal management or forced air cooling might be needed for a fully constructed pack.

It is also important to continue measuring heat after the end of testing; in my experience the surface of the Vectrix NiMH cells reach peak temperature well after the end of a test (or recharging); that's because it takes time for the heat generated in the cell center to spread evenly.

Depending on physical characteristics of different cells they might also have different "thermal mass"; that means that although they might have the apparent same temperature, one type might contain a lot more heat than the other. That will make a difference when repeated charge/discharge cycles are needed in very warm or very cold weather.
A battery with a higher thermal mass will initially appear to heat up less, but it will also remain warmer for longer, and might accumulate heat if not cooled for extended periods after riding and / or charging.

+1 on that Mr Mik!

I know about thermal management of lithium cells and the heat increase after the end of discharge was also observed but not as much as i expected.. thatS, around +2 to +5 degree max..

I also can add that at Constant Current discharge the voltage decrease as the cell discharge... and having the same current with lower voltage mean an increase of the IR wich indicate an increase of Heat production enevthou the voltage is lower for the same current.

I agree on the fact that testing a single cell without any other around it is not a as representative as a, ebike pack. But cell in the middle always get warmer than the surounding cells.. So wich condiytion is the best.. a tradeoff between these cell position?.... Also the discharge i dot was done in continuous current discharge wich not represent the same use than an ebike that is more like pulsated discharge peak with a minimum discharge rate...

Discharging by pulse current help the cell to dissipate the heat better eventhough the mean current is the same in both test.

For that maybe i'll upgrade my next test to see if there uis a real difference and if the heat still increase for 2 to 5 degree after the cell discharge cut.

I think this is difficult to reproduce the same conditions of use of ebikes battery discharge. some pack are made in cube.. or prismatic, some are plane shared in many small pack... some have air blowwing on them some have not...
So for that i think keeping the cell alone for the test is the most easy way to keep a baseline for compare cells..

Adding the projected proportional heat increase to the results for every different battery construction way is still the best i think.

Trying to imitate one perticular type of battery construction for a cell test might complicate the comparaison for the overall battery construction comparaison..

At least that's what every mechanical engineer heat simulation do.. they try to keep the baseline conditoins on their test methods and then they add the different test conditions to every simulations.

Doc

[oatnet]

Hmm ... When formed into a "brick" pack, I'd expect the Yesa would have less exposed surface area than the Headway, I wonder how temps in the center of the packs would compare.

Regarding heat dissipation don't you also need to figure in how the cells will be packaged? The rectangular YESA cells will likely be placed right up against one another however the cylindrical Headway cells will have some air gap between cells. So while an individual YESA cell may shed heat better a fully constructed pack of the Headway cells might actually dissipate heat from each cell more evenly don't you think?

JINX - you owe me a coke! I think you said it better though.

-JD

[Doctorbass]

Question: Does the 94mm height include the posts?
-R

Yes.. but i revised the calculation i did an error in my calculations.. thanks for asking

Please note: EVERY CELLS IN THIS TEST WAS THE BEST CELL I HAD FROM EACH BRAND

Doc

[Doctorbass]

Here is the PSI final test graphs:

The result seems interesting.. there is only one strange thing!.. The 30 and 35A look very similar!!.. I doublechecked the parameters and this is still ok..!

Doc

[Doctorbass]

As i promissed: And... Finally the comparative graphs between these 3 cells brand:

i'll elaborate on that tomorrow morning.. now it's time to sleep!! 2:10am here.. work tomorrow

Good night!

Doc

[headway]

YES, We will have bigger cells, 12Ah and 15Ah cells. And they will be planned to come out at the end of March. This is confirmed from our G.M.

Victoria

these blue sleeved cells are much better than the older brown paper sleeved cells. flip measured 2.47V at 34A last year, and we had about 17mohm IR, so these have improved dramatically, so i am hoping they can still push to improve their cathodic powder and anode coating up to the performance of the top tier cells, such as psi, a123, and yesa. victoria mentioned they will introduce larger capacity too, 14Ah is next i think.