DLG / K2 cells

[johnrobholmes]

I have some packs on the way from K2, 48v worth of 16 amp hours. They are using a cell similar to the a123 M1 in size, but each cell is about 3.2ah per cell. Each battery brick is 4s5p, and will take 120a pulse discharge and 64a continuous. With the additional voltage I plan to gear down so that my my peak current requirements drop from 80a to 40a hopefully. Double voltage, halve gearing. As long as I don't goose the throttle it should work out

I see a lot of people getting the Ping made packs, but these cells seemed like a more robust pack for marginally more money. Better construction, known discharge graphs, etc. There is no BMS, but each cell is cycled three times at the factory and matched by voltage and capacity to the other cells, then after construction it is cycled again. Since I will generally use less than half of the capacity during my rides they should last quite a long time and give plenty of punch.


Anybody else use these cells here?

 

[GGoodrum]

I saw these before. I think the company is in Nevada, right? Anyway, I looked at these as a possible replacement for LiFeBatt and a123 cells, but the pricing is close to the same as what you can get a123 cells for, and these are 4C-rated cells, so it didn't seem worth it. the 4s5p "brick" pack is also pretty expensive, at around $250. For a 48V 16Ah setup, you would need four of these, so $1000. that's about double what an equaivalent capacity Ping pack would cost and the charger(s) would be extra. These do have twice the C rating, which is good, but this is as much as a LiFeBatt HPS pack, which I still think is priced too high.

What worries me, howver, is that there is no BMS. I don't care how well cells are matched at the factory, over time their capacities and internal resistances will change, and the cells can eventually get out-of-balance. This is especaially true for the cells buried in the middle of the brick. If you can, see if you can get the cell junctions brought out of the pack to a connector, of some sort. That way you will be able to at least monitor the cells occasionally, to make sure they are staying relatively close. You might also consider using their 4-cell/12V chargers, and charge each 4s5p brick separately, That will minimise the balance problem. Also, make sure you get some kind of warranty. Without a BMS, a warranty is your best protection.

I'm sure you will be happy with the performance. Being able to pull 60A out of these will give you "spirited" performance. Forget about the "burst" mode ratings. On an ebike application, you can only get to about the continuous current rating before the voltage gets pulled down to much. I've found this to be true for pretty much everything, from the low-end Ping/duct tape packs,to the high-end a123 cells.

-- Gary

 

[Microbatman]

K2 cells have been getting quite the intrest.

There are 2 threads at RCGroups.com that are worth reading.

Right now they appear to be the best alternative to A123.

http://www.rcgroups.com/forums/showthread.php?t=868243

http://www.rcgroups.com/forums/showthread.php?t=847214

 

[TylerDurden]

Not too suprisingly, K2 is just outside of LAS, where Valence is; since the head of K2 is a former Valence guy: Johnnie Stoker.

He might actually be the father of the lithum battery...

Method of synthesizing electrochemically active materials from a slurry of precursors

Abstract
A method for making an active material comprises the steps of forming a slurry, spray drying the slurry to form a powdered precursor composition, and heating the powdered precursor composition at a temperature and for a time sufficient to form a reaction product. The slurry has a liquid phase and a solid phase, and contains at least an alkali metal compound and a transition metal compound. Preferably the liquid phase contains dissolved alkali metal compound, and the solid phase contains an insoluble transition metal compound, an insoluble carbonaceous material compound, or both. Electrodes and batteries are provided that contain the active materials.

<snip>
5. A method according to claim 1, wherein the alkali metal ions comprise lithium ions.

6. A method according to claim 1, wherein the alkali metal ions comprise lithium ions, the solvent comprises water, and the transition metal is selected from the group consisting of vanadium, chromium, manganese, iron, cobalt, nickel, molybdenum, and mixtures thereof.
<snip>

http://patft.uspto.gov/netacgi/nph-Parser?Sect1=PTO1&Sect2=HITOFF&d=PALL&p=1&u=%2Fnetahtml%2FPTO%2Fsrchnum.htm&r=1&f=G&l=50&s1=6,913,855.PN.&OS=PN/6,913,855&RS=PN/6,913,855

 

[shinyballs]

They do have a BMS but limited only upto 4 series cells, i.e. 14 volts, which matches there charger. price is around $30

 

[johnrobholmes]

I bought a bunch of a123 packs and got them for $10 per cell. Now you can't find the dewalt packs so cheap anywhere. At just over $10 per cell built into a shape I can use, I think the K2 cells do give a123 a good run for the money. While the cells I chose are indeed 4C, they also have 40% more capacity in the same volume. They have a higher discharge version, but the 4C will be perfect I think.

I will periodically check the parallel banks for balance, probably every 10 cycles once they are broken in. Each 12v chunk will probably be charged seperatly, I have enough hobby chargers to do so. Talking with the rep, he stated that they have many packs in EV use with thousands of cycles and no BMS other than a voltage cutoff. I know some cells will get out of whack, but the beauty of LiFe is that they have a very large safe range of charge voltage with little damage to the cell. Some guys on RC groups have been thrashing their packs with 4.1v/cell charges with no performance decline and marginal capacity decline.

It was between these cells and Lipo. I really didn't want Lipo between my legs at this point. I have been watching Ping packs, and with shipping it seems that the 20ah 48v packs have been going for $650 to $750 or more. Low discharge, soft shell, very heavy packs, not my cup of tea at the moment.


K2 is in Nevada. Excellent comments guys.

 

[Jeff]

I have some packs on the way from K2, 48v worth of 16 amp hours.
Anybody else use these cells here?

I have some recent experience in testing the LFP cells marketed by K2 (formerly Peak Battery) in LV.

Two chemistry variations of the 26650 cell were tested to failure using an automated charge/discharge cycling system running under labview.

The first tested, their pilot model cell was rated to 2.6A, though it could not deliver beyond 2C in order to maximize cycle life.
Advertised impedance was less than 20mohms, actual averaged 28 mohms. Cells tested =20
After ~ 310 cycles to 90% DOD, not exceeding 35 degrees C, internal impedance began to rise sharply to beyond 40 mohms before testing was aborted.

The second tested, the 2.8A cell known as the "E" cell initially indicated an impedance of 31 mohms. At 2C cycle rate to 90% DOD, the cells began to indicate impedance rise at 232 cycles, testing terminated at 274 cycles with internal impedance exceeding 40 mohms.

K2 offered to sell me the EV cells next. I have four "freebies" though I never bothered with testing and kicked them to the curb...
I wound up buying a volume of A123 M1 cells. The A123 were half again more expensive than the K2 cells, but were worth avoiding further grief and used car salesman type pitches from K2.

Oh, and to make matters worse, the pilot and and "E" cells had a plating problem on the positive terminals. It had to be ground to base nickel in order to resistance weld the cells properly. Sheesh.


Jeff

 

[TylerDurden]

Jeff,

Any thoughts on where the cells are produced? How long ago was the testing done?

Not bagging, just wondering if they should be re-visited...

 

[johnrobholmes]

Good info there. We shall see how long my packs last. My charger measures cell resistance, and for my use measuring the resistance as a whole pack will be fine.

 

[Jeff]

Jeff,

Any thoughts on where the cells are produced? How long ago was the testing done?

Not bagging, just wondering if they should be re-visited...

TD, I had to revisit my notes for more info. Peak Battery was very dodgy in telling me initially who assembled the cells.
There were a couple dozen calls with them during the testing cycles, and at one point they did disclose who the maker was and evidently I didn't write it down. Sorry. It was a no-name chinese plant assembling their cells from Phostech licensed materials under their (Peak Battery's) direction.

The first batch of cells tested were done 20 months ago. The improved "E" cells were 16 months ago.

When they shipped the "EV" cells to me for testing, about 9 mos ago, they noted improvements in the separator, and an improved seal. Because these are rated for higher discharge rates along with a slight capacity increase, I suspect they're actually just using a thinner separator with no compound changes.

I included an image of one 6P test pack of the Peak (K2) "E" cells, and a few of the improved "EV" cells that I made quick tests on to see if the newer terminal plating would take a weld. They passed welding OK. And check out the "blast zones" on some of the positive terminals in the 6P test pack. That's what happens when you don't remove their nasty plating completely. It literally blows a hole through the .010" nickel strips and makes quite a bang.

How it started:
I own a TF bike that draws upwards of 60A on hard accel. I was trying to put together a pack of cells that would tolerate the desert heat and deliver the power demanded without an early death. The LiPo's just could not handle the strain nor the heat.

Jeff

 

[johnrobholmes]

Just got in a slew of the cells from K2. Advertised as 19mOhm impedence or less, but so far only 1/4 of the cells are below that. 1/4 of the cells are above 21mOhm, which is more than 10% over the factory ratings. I will be giving them a call about it and see what can be done.


Soldered on connectors and they are charging now. No longer will I have a feeble 8 mile range on my bike!

 

[shinyballs]

johnrobholmes Please post pics of your packs, how's the performance so far... ?

 

[johnrobholmes]

So far they are charging just fine. First pack's resistance was 19.5 mOhm per parallel bank, or 78mOhm total for 4s.