Headway LiFePO4 Battery Testing
29th January 2008

NOTICE: There has been some evidence accumulating that recent cells have not been performing as well as these initial samples. I encourage everyone to do plenty of their own research before purchasing cells! See also this thread on endless-sphere.com for more independent testing data.

Introduction

As part of my ongoing search for good batteries for EV use, here are the results of testing some more commodity Lithium Iron Phosphate (LiFePO4) batteries. This round of testing was two 38120 sized cells from chinese manufacturer Headway (http://headway-cn.en.alibaba.com). This is a effectively an extension to my original battery test report from late 2007 - click here to view the original report.

The Cells

The Test Equipment

Results: Discharging Performance

	38120L	38120P
1C
2C
3C

Notes and comments:

• The lowest load resistance this testbench can manage was 0.1 ohms, which equates to about 3C with these cells, so unfortunately I was unable to test up to 10C as usual. However, performance at 3C covers typical EV use and also gives an indication of how the cell might handle higher currents.
• The resolution of the ACS754 current sensor when used with the LabJack U3 was not brilliant (hence the noise on the current plots in blue), so amp-hour and watt-hour totals may not be precise (especially for lower currents). But, they should still be a pretty reasonable representation and comparison.
• The rows are matched by approximate C rates only - since discharging was done with a resistive load and some batteries have more voltage sag than others.

Results: Charging Performance

The majority of current EVs are charged from single phase power and do not have regen, so the maximum rate of charge is a fraction of 1C. However in the future there will be an increasing need for batteries which can handle high charge current, allowing for fast charging stations, and regenerative braking to improve vehicle efficiency. So it's worth seeing what the maximum rate of charge is for various cells.

	38120L	38120P
0.25C
0.5C

Notes and comments:

• Once again these larger cells were pushing the limits of the existing test bench, and the fastest I could charge them was 0.5C. As it turned out, this did not matter much as the cells didn't display a classic lithium charge curve even at 0.25C!
• The wacky temperature plots on some graphs is due to the air conditioning system. It's mid-summer here in Australia and we've had some scorchers - but I've tried to keep ambient temperature for testing around mid to high twenties (celsius).
• Interesting to note the effect of temperature on the 38120L charge at 0.25C - they seemed to accept charge better when cooled, surprisingly.

Conclusion

The good:

• At a touch over US$0.50 per watt hour, the 38120L are some of the cheapest LiFePO4s I've come across.
• Both cells discharged at up to 3C continuous comfortably, and I have no doubt they could put out 10C briefly - i.e plenty of power for typical road-going EVs.
• The larger format (vs 26650s etc) means fewer cells parallelled up, and much less work to build into a pack.

The bad:

• Comparing 3C results with those from previous tests, cell heating was relatively high and it seems to me that their peak power is well short of the best 26650s and 18650s.
• Charge performance was a bit disappointing, even at 0.25C.. I'd have trouble recommending them for anyone with strong regen or fast chargers.
• 1000 cycle life might be a problem for vehicles needing to be charged every day, though many people running lithium packs would only be charging once per week, in which case calendar life would be the limiting factor, not cycle life.

Final thoughts? Discharge curves between the P and L variants were remarkably similar, I tend to wonder just how much more powerful the P cell is.. Based on these results, it seems like the L variant makes more sense. I would be interested to take them to the next step, and build a pack for a commuter vehicle from 38120Ls sometime.

Comments

There are no comments for this page yet.