K2 Energy Lithium Iron Phosphate 26650 Cells
By Ken Myers
Original posting date: June 23, 2008
Updated July 23, 2008

     K2 Energy provides a Lithium Iron Phosphate type cell. www.peakbattery.com

     Steve Hill, of Robotic Power Solutions (AKA www.battlepack.com), provided some 26650 2500mAh cells (non-CID version) for me to compare directly against the 26650 A123 Systems, Inc. 2300mAh cells. The K2 Energy cells are described as; LFP26650P Hi Power Rechargeable 3.2V 2500mAh. There is a thread on RC Groups about these cells. Everydayflyer (Charles) is also testing them and has reported his results in the thread.

     Here is what K2 Energy says about their 26650 cell:
"... This cell can handle 42 Amps continuous discharge with over 50 Amps 30 second pulses. The cell is rated at 3.2 nominal volts and 2500mAhrs. (snip) The internal impedance is less than 9 milliohms. This battery is 26.2mm D x 65.2mm H and weights 82.5 grams (with a card board sleeve)."

     It is important to note that suppliers can claim whatever they want about their cells.

     So far, Everydayflyer, using a CellPro 10S has found the internal impedance for the individual cells in a 4S pack to be;
Test one: 12.5mOhm, 11.4mOhm, 14.4mOhm, 12.9mOhm
Test two: 12.8mOhm, 12.1mOhm, 14.6mOhm, 13.0mOhm

     For comparison, here are the numbers he got from a 4S A123 Systems, Inc. 26650 2300mAh pack
Test one: 6.1mOhm, 6.0mOhm, 8.9mOhm, 6.3mOhm
Test two: 5.5mOhm, 5.0mOhm, 8.4mOhm, 5.4mOhm
     He noted that cell #3 has always had a higher resistance.

My Size and Weight Measurements

K2 diameter 27mm, length 65.1mm
A123 diameter 26.5mm, length 65.6mm
K2 individual cell weights with tan cardboard-like sleeve: (I have four cells.)
Cell 1 82.7g
Cell 2 82.2g
Cell 3 82.8g
Cell 4 82.8g
Resting voltages as delivered:
Cell 1 3.28v
Cell 2 3.28v
Cell 3 3.28v
Cell 4 3.27v
Note: The button end is positive on the K2 Energy cells, which is the reverse of the 26650 cell from A123 Systems, Inc.

3S1P Pack weights:

     Cells #1, #2, and #3 were made into a pack using 2 APP connectors, power leads, node/balance leads, tape and hook and loop fastener strips. Everything was done to duplicate, as closely as possible, the A123 pack I've been using.
A123 240.5g (8.48338 oz.)
K2 267.85g (9.44813 oz.)
Difference: 27.35g (0.964742 oz.)

Testing and Comparing of the K2 & A123

     The new K2 Energy pack was first charged at 2.5 amps (1C) using an unmodified AF109 charger with termination done manually with the resting cell voltages being 3.732, 3.738, 3.719 about an hour and a half after the initial charge. The pack was discharge using the AF 109. The pack was then charged at 5 amps (2C) and 2.526AH was returned to the pack in 32 minutes, again terminating manually. The pack was discharged using the AF 109. The pack was then charged at 7.5 amps (3C) in 21 minutes with 2.492Ah returned to the pack. The pack sat overnight. In the morning, I was very surprised to see the Cellmeter-8 read all three cells' voltages exactly the same at 3.334v!
     I noted one thing while charging on the AF 109; the K2 pack gets warmer, much warmer, than my A123 pack when it is Zip charged. This indicates higher cell impedance than the same size cells from A123 Systems, Inc., which was verified by the tests done by Everydayflyer noted earlier.
     With 3 charges and 2 discharges on the new K2 pack, I decided to start static testing. Comparative static testing was done using the Son of Swallow power system; Hyperion Z3019-10, Master Airscrew standard wood 10x7 prop and a Castle Creations Phoenix-45 with low timing.
     My Hyperion Emeter can store 5 data points at a time. For each static motor run, at full throttle, I collected data every 5 seconds with shut downs and run-ups to full throttle between each one. I did three data gathering runs for each pack so that 15 data points were gathered between 5 and 75 seconds, recording the data to a spreadsheet for each 5 collected data points. The graphs show the behavior of the packs with the slight resting, to record the data, between each 5-point collection. When viewing the graphs, it is easy to spot the “break” between the data collection runs.
     RPM is not shown on the graph because, when RPM is changed to RPM/1000, the KRPM is approximately the same as the voltage and does not present well.

Test 1

Averages for 5 through 75 seconds:
A123: 8.62v, 33.58 amps, 8628 RPM, Watts in 289.60
K2: 7.88v, 29.41 amps, 7978 RPM, Watts in 231.67

     In my article "Effects of Temperature On the Battery", I noted that battery temperature effects the performance of the battery. I decided to run the second test starting 15 seconds into the motor run to allow each battery a chance to heat up under load.

Test 2

Averages for 15 through 85 seconds:
A123: 8.57v, 33.43 amps, 8558 RPM, Watts in 286.64
K2: 8.10v, 29.87amps, 8092 RPM, Watts in 239.23

Getting the Performance Up for the K2 Enerby Cells

     I decided to see if I could get the performance up for the K2 cells. For a plane that flies on the wing, like the Son of Swallow, pitch speed is a good indicator of relative performance. I compared the data that I had gathered with the Master Airscrew 10x7 standard wood prop for the A123 pack and K2 pack. Test two indicated a pitch speed of 56.7 mph for the A123 pack and 53.6 mph for the K2 pack.
     I statically tested two alternate props. The first was a Master Airscrew 11x8 standard wood prop. The average data; 7.49v, 36.53 amps, 7176 RPM, 273.59 watts in, pitch speed 54.4 mph. The second prop was a Master Airscrew 10x8 standard wood prop that yielded; 7.75v, 33.36 amps, 7626 RPM, 258.42 watts in, pitch speed 57.8 mph.
     After posting my first graph to the RC Groups thread, Steve Hill asked me to run the test again, which I did the following day.

Test 3 (Day 2 of testing)

Averages for 5 through 75 seconds:
A123: 8.61v, 33.33 amps, 8602 RPM, Watts in 286.88
K2: 8.14v, 30.39 amps, 8216 RPM, Watts in 247.50

     These numbers indicated that the K2 pack was improving with use.
     I was concerned that the 11x8 might be too much prop load for my goal of staying in the 35-amp draw range, so I reevaluated the 10x8 Master Airscrew standard wood prop. With the K2 pack, the average numbers were now; 7.96v, 34.7 amps, 7776 RPM and 276.07 watts in. The new data now indicated a pitch speed of 58.9 mph.
     These numbers indicate similar performance to the SOS using the 10x7 and the A123 pack.

Subjective Flight Testing

     I do not have any onboard data gathering system. Flight-testing was done at the Skymasters' Electric Fly on June 14. It was a beautiful, sunny day with the temperature in the high 70-degee F range. The wind was about 10 mph. I first flew the Son of Swallow using the A123 pack to "get the feel and performance" of the plane fresh in my mind. After the initial flight with the A123 pack, I landed and immediately changed to the K2 pack. I did not change the prop at this time. While flying with the K2 pack I immediately noted that it required a higher throttle setting in level flight to feel the same and was just a little bit off of the performance of the A123 pack in the vertical, yet it was still good, just not quite as good.
     The K2 pack was charged and the prop changed to the Master Airscrew 10x8 standard wood. With the prop change, the plane "felt" at least as good in the air as when flown with the A123 pack and 10x7 prop. After another recharge of the K2 pack, it was flown again.
     When the K2 pack is used with the 10x8 and the A123 pack with the 10x7, there is about 30 watts in difference in favor of the A123 pack. I could not really tell the difference in the air.

     The following Monday, I did a couple more static tests. Test 4 used the A123 pack that was charged on Saturday at the meet but never flown and the K2 pack that had been charged earlier Monday morning. The load was the 10x7 prop that I had used for all my previous testing.

Test 4

Averages for 5 through 75 seconds:
A123: 8.43v, 32.62 amps, 8458 RPM, Watts in 275.21
K2: 7.79v, 28.62 amps, 7904 RPM, Watts in 222.89

     For the next test I used the 10x7 with the A123 pack and the 10x8 with the K2 pack. Remembering that both the A123 pack and K2 pack "put out more" when warm, I tested them with the packs "hot off the charger."

Test 5

Averages for 5 through 75 seconds:
A123: 8.68v, 33.90 amps, 8672 RPM, Watts in 294.20
K2: 7.77v, 33.92 amps, 7620 RPM, Watts in 263.50

     Using the data from tests 3 and 5, I graphed the relative pitch speeds.

     I am continuing to fly the SOS using the K2 3S1P pack and Master Airscrew10x8 standard wood prop to check the life of the K2 cells. I have decided to charge the K2 for the rest of its life at 7.5 amps, as waiting longer than 20 minutes is "just too long", when at the field! This should provide a good "torture test" for them as well.

     On the 17th of June I did a couple of discharge tests to try and establish a baseline capacity for the pack. Unfortunately, I do not have a CBA discharger. I first used my SR Batteries charger and discharged to 7 volts at a 1-amp discharge rate, as high as the unit will go. I found the results of the test, not the cells' performance, unsatisfactory with the capacity of 2224mAh and the bounce back resting voltages of the cells, after the load was removed, were 2.795, 2.616, 2.652. Next I charged the pack again and used the unmodified AF 109 charger to discharge the pack. The AF 109 uses a 1.25-amp rate, and I was surprised to find that it took the pack down to 7.07 volts, almost the same as the SR charger. It showed 2405mAh taken from the pack and the bounce back resting voltages were 2.791, 2.668 and 2.658, which was very similar to what I found using the SR charger.

     At this time I am not satisfied with the tests I'm using for capacity, but have recorded the results here so that I have a record of what I've seen.

06/21/08 Flying Update
     The Son of Swallow was flown four times using the 3S1P pack made up of the K2 Energy cells and the Master Airscrew 10x8 standard wood prop. The performance is good, equal to the A123 pack using the 10x7 prop, but the power system did require a little higher average throttle setting than with the A123 Systems, Inc. 26650 cells. This was demonstrated by the higher average current draw when compared to the A123 pack, which results in approximately the same flight time as the A123 pack. It is also shown by the approximately 135-degrees F/57.2-degrees C to 140-degrees F/60-degrees C battery temperature upon landing on an 80-degree F/26.7-degrees C ambient day. Because the pack was "coming down" hotter, I did wait for the K2 pack to cool some between each 7.5 amp charge. The K2 pack's temperature, after each charge, was about 105-degrees F/40.6-degrees C. All flights were taken with the pack "hot off the charger." In many ways the K2 pack's characteristics reminds me of the Emoli pack I flew two years ago.

06/23/08 Balance Update
     These cells seem to be staying in balance very well. I meausred the cell voltages of the resting pack that had been discharged with the Son of Swallow's last flight on Saturday, June 21. The cell voltages were; 3.234, 3.234, 3.230.

06/25/08 Flying Update
Five flights on June 24. About 80-deg ambient temperature. The pack was consistently at 140-deg F, or higher, after a 6 minute timed flight. I waited for the pack to cool before recharging between each flight.
     Even though I had only partially charged the pack at home and finished at the field, the first flight had noticably less power available. Once the pack had "warmed up", the performance was up on flights 2 through 5. I was consistently putting in about 2000mAh according to my AF 109. On flight 5 I decided to bump the timer up to 7 minutes. I went into a loop at about the 6-minute mark and the power was obviously down at that point. I just landed without flying to the 7-minute mark. Even though these cells have a higher listed capacity, 2500mAh, their "usable" capacity is about the same as the cells from A123 Systems, Inc., which are rated at 2300mAh.
     The last flight was about 2 p.m. yesterday and using the CellMeter-8 the cells read 3.122v, 3.094v, and 3.088v on the morning of June 25.

06/28/08 Update - testing done on 06/27/08
     I had not been able to get much more than 2000mAh, according to my AF 109, returned into the K2 Energy pack at the field when charging at 7.5 amps. When this testing was done, it was four days since the K2 pack was flown. I checked the resting cell voltages (3.120v, 3.091v and 3.085v). I decided to balance the pack with my A123 Blinky. The resting voltages, after being AF 123 Blinkied, were; 2.971v, 3.001v and 2.952v.
     The pack was then charged at 2.5 amps (1C). The ambient temperature at the end of the charge time was 88-deg F and the pack's temperature was 93.1-deg F. The AF 109 indicated that 2329mAh had been returned to the pack. Before the start of the test using the MA 10x7 wood prop, the K2 pack's cell voltages read; 3.319v, 3.401v and 3.372v. After the test run, the pack's temperature was 110.5-deg F.
     I captured full throttle data every five seconds on the Hyperion Emeter and recorded the data to the spreadsheet between each five data captures. The test sequence was run until the Castle Creation's Phoenix 45 hit the 6v LVC that I had it set for. I forgot to gather the voltages of the cells after the test sequence.
     I had left my A123 pack partially charged. 902mAh were Zip charged into the pack according to the AF Whattmeter. The ambient temperature was still 88-deg F. The A123 pack's temperature was 95.3-deg F after the Zip charge. The same test was run using the MA 10x7 wood prop to the LVC for this pack. The A123 pack's temperature after the test was 107.5-deg F. The voltages after the test sequence were 2.811v, 2.863v, 2.803v.
     A second set of data was collected for the K2 pack. The prop was changed to the MA 10x8 wood and the K2 pack recharged at 7.5 amps. 2083mAh was returned to the pack. Its temperature after the 7.5 amp charge was 103.1-deg F. The voltages after the test sequence were 2.982v, 2.998v, 2.957v.
     The graph below shows the volts. The voltage of the K2 pack with the 10x7 load and 10x8 load can clearly be seen to taper off in voltage after about the mid-point of the run.

     I have not previously shown RPM, because it does not present well with the other data, so here is a graph of the RPM, for those who may be curious. The RPM obviously parallels the volts.

     I had noted earlier that both packs fly the Son of Swallow for 6 minutes of very aerobatic flight when the MA 10x7 wood is used with the A123 pack and the MA 10x8 wood is used with K2 pack.
     The K2 pack with the MA 10x8 wood averages 32.24 amps for the 150 seconds, while the A123 pack averages 33.45 amps for 150 seconds.
     I noted in my first flight report that I did not care for the way the plane flew with the K2 pack and MA 10x7 prop. The reason for that can be seen in the following graph that shows the Watts in.

     Based on Watts in per pound, it looks like the A123 pack with the 10x7 should be far superior to the K2 pack with the 10x8, but flight testing has proven this NOT to be the case! Why? Pitch Speed. Here is a graph showing the pitch speeds.

     The preceding two graphs, and my flight experiences, cleary demonstrate that, for planes that "fly on the wing" and that are "relatively clean", the pitch speed, when used in the ratio of pitch speed to stall speed, can even override the watts in per pound rule of thumb.
     The Son of Swallow has a pitch speed to stall speed ratio of 4.09:1 when using the A123 pack with the 10x7 prop, 4.15:1 when using the K2 pack with the 10x8 prop, and 3.73:1 when using the K2 pack with the 10x7 prop.
     I honestly cannot perceive any difference in the flyability, performance or duration of the Son of Swallow whether I am flying the K2 pack with the 10x8 or the A123 pack with the 10x7.

     The next graph is similar to ones that I presented earlier and shows the volts and amps, but this time all the way to the 6v LVC.

     Finally, I measured the resting voltages for the two packs approximately 18 hours after they were depleted in the tests.
K2 - 3.004v, 3.028v, 2.983v
A123 - 2.947v, 2.990v, 2.944v

Note: The following flight tests were completed using the K2 Energy pack with the CC Phoenix 45 at low timing, using the Hyperion (mislabeled) Z3019-10 and the Master Airscrew 10x8 standard wood prop.

07/04/08 Flying Update
     Three flights on the pack today in the SOS. Sunny with temperatures in the low 70-deg F range all morning and early afternoon.
     The first charge was done at the field so that the pack would be "warmed up" before the first flight of the day. All charges were done at 7.5 amps and there was a ten minute "cool down" time for the pack between flights.
Charge #1: pack temp. beg. chg. 73.6F, after chg. 99.2F, input 2200mAh, after flight 125.2F
Charge #2: pack temp. beg. chg. 101.0F, after chg. was missed and mAh, I was talking, after flight 137.8F
Charge #3: pack temp. beg. chg. 106.4F, after chg. 117.2F, input 2100mAh, after flight 131.7F
Pack voltages after about one and a half hours from the third flight were; 3.203v, 3.202v, 3.194v.
     Each flight was a timed 6 minutes and the LVC was never reached during any of the three flights. The charging was done manually using the AF 109 charger and the stated mAh were shown in Ah on the AF 109 screen.

07/05/08 Flying Update
     Four flights on the pack today in the SOS. Sunny with temperatures in the low to mid-70-deg F range all morning and early afternoon.
     Again, the first charge was done at the field so that the pack would be "warmed up" before the first flight of the day. All charges were done at 7.5 amps and there was a ten minute "cool down" time for the pack between flights.
Charge #1: pack temp. beg. chg. 73.0F, after chg. 106.1F, input 2040mAh, after flight 131.6F
Charge #2: pack temp. beg. chg. 113.7.0F, after chg. 118.1F, input 2000mAh after flight 135.2F
Charge #3: pack temp. beg. chg. 105.8F, after chg. 118.6F, input 2010mAh, after flight 143.6F
Charge #4: pack temp. beg. chg. 117.3F, after chg. 121.5F, input 2140mAh, after flight 148.6F
Pack voltages after about one and a half hours from the fourth flight were; 3.142v, 3.145v, 3.121v.
     Each flight was a timed 6 minutes but today the LVC kicked in near the end of the forth flight, cutting it just slightly short of the six minute mark. I can tell I was pushing it harder on the fourth flight, as the pack temperature after landing was the hottest of the day. The ambient temperature had not risen much as it was only about 75-deg. F at the end of the fourth flight. The charging was done manually using the AF 109 charger and the stated mAh were shown in Ah on the AF 109 screen.

07/06/08 Flying Update
     Good day of flying with temperaturer in the mid-70's F. Only got three flights, even though I was at the field quite a while. Helped a friend find a plane that went down in the woods, and that took quite a while.
     The first charge was a partial charge at home at 2.3 amps and then topped at the field at 7.5 amps. Put in 1250mAh at home and another 800mAh at the field
Charge #1 see note above, after flight 131.1F
Charge #2 pack temp. beg. chg. 108.3 F, after chg. 123.4, input 2100mAh, after flight 138.0F
Charge #3 pack temp. beg. chg. 109.8, stopped charge, looked for plane and then finished charge, after flight 136.5F

07/11/08 Flying Update and flight time clarification
     Flew the SOS three times after completing the Mid-Am preparations, but did not record any data.
     I've noted that I've been flying this pack for 6-minutes. That means that I start the timer on my Hitec Eclipse 7 just before taking off, then throttle up and take off. The timer on the transmitter beeps a few times at 5 minutes and 45 seconds. I generally finish any maneuver I'm doing and then set up for a typical "pattern" type landing, in no hurry. Wheels down varies from a little less than 6 minutes to a little over. If the grass permits easy taxiing, I generally taxi back to the flight station that I'm using. The typical flight, to keep "pressing the pack" includes a lot of verticals; stall turns, consecutive loops, loops with snaps, vertical eights, etc.

07/23/08 IR test
     Since I have been beta testing the K2 Energy 26650 2500mAh LiFePO4 cells, I wanted to get a direct comparison of the internal resistance to the A123 26650 2300mAh LiFePO4 cells I had been using. Based on my comments from above, I knew that it was higher for the K2 Energy cells because of the pack's higher measured heat after charging and flying, but I wanted to know by how much. Using a FMAdirect CellPro 10S answered that question!
     I ran three charge cycles at the same time on the two 3S batteries, one at 2.3 amps, one at 5.0 amps and one at 7.5 amps, using the 3.65v A123 default on the charger at the time I tested. By checking the saved .TXT files via the spreadsheet method I confirmed that the K2 Energy cells had a lot higher internal resistance, when compared to the A123 cells.

     The spreadsheet shows that the IR for the K2 cells is more than three times greater than the A123 cells when charged at the same rate at the same time. This clearly confirms why the K2 pack gets "hotter" during charge and discharge, and why it is "harder" to get energy out of the pack. It also explains the greater voltage depression of the K2 pack under load.

To Reach Ken Myers, you can land mail to the address at the top of the page. My E-mail address is: KMyersEFO@mac.com
EFO WEBsite: http://members.aol.com/KMyersEFO/