Source: 2017 Q1 Beartracks, Tyson Sawyer

I had long desired weight savings of replacing my Odyssey PC680 AGM (Absorbed Glass Mat/Lead-Acid) battery with a light weight Lithium Battery. This would save nearly 10 pounds over the already light weight Odyssey. I was, however, mindful of the risks of lithium batteries. Lithium Iron Phosphate(LiFePO4) is my chemistry of choice for its high current capacity and thermal stability. This is not the same lithium chemistry used in laptops, Boeing airliners, or Samsung phones. Most lithium batteries offered as replacements for 12V, lead-acid starter batteries are Lithium Iron Phosphate. More information on the different lithium chemistries can be found here: http://batteryuniversity.com/learn/article/types_of_lithium_ion
LiFePO4 batteries still have a couple of weaknesses. If excessively discharged, your multi-hundred dollar battery is junk. They need periodic balancing of their cells. Also, though far more robust than other lithium chemistries, they still have limits. Some LiFePO4 batteries I had seen, such as AeroVoltz, required occasional balancing with a special charger and had no ability to disconnect and protect themselves if needed. There are many stories of AeroVoltz batteries being turned to junk from over discharge while trying to start a difficult engine. EarthX was the first affordable LiFePO4 battery I found that had a built-in battery management system (BMS) that balanced the cells and protected the battery. When a friend offered testimonial of how well the EarthX ETX900 was working in his C-180, I ordered one. Here is a link to their website.
The first test was to try starting the engine by disconnecting the on-board battery and using jumper cables to start the engine. This failed. Not enough power was getting to the starter. I noticed that the voltage drop was much more than expected for a lithium battery and suspect that my jumper cables were the problem. After finding
a better set of jumper cables, starting was successful, though a little labored. EarthX batteries are rated to 140F maximum and are not recommended for installation on the engine side of the firewall. My plan was to install it on the cabin side of the firewall. The effort to do this turned out to be much more than expected. To research just how hot the engine side was, I installed some “non-reversible temperature labels” (http://www.omega.com/pptst/TL-10.html). These indicate the highest temperature seen by the label in 5F increments. I installed several on the battery box with the existing PC680 battery and tested them during a Halloween week kayak trip from New Hampshire to Georgia with 80 degree F temps in Georgia. The first indicator on these is 105 degrees F, and it never “triggered”. Seeing this, I changed my plan and installed the EarthX battery in the existing location. Some minor tweaks to the existing box were needed as the dimensions are not identical to the old Odyssey.
The biggest difference is that the EarthX is not as wide. EarthX provided some stick-on foam spacers to solve this.
This also left some space on one side of the battery that let me lower a Bluetooth temperature sensor (from http://www.weatherhawk.com/myblue-t) into the box and monitor temperature in the box while in flight without the need to run any wires from the engine compartment to the cabin.
With the battery installed in the aircraft, without the long jumper cables, starting was clearly snappier than with the Odyssey. Voltage while cranking was also a couple of volts higher. I have had one case of pulling the plane out of a 40F hangar into roughly 0F conditions and having it sit for a little while.
By the time I was ready to start, the carburetor was cold and the air being fed into the engine was very cold. I didn’t give it enough fuel and ended up cranking it quite a bit. On one long crank, the BMS kicked me out for a short time. One of its protection policies is to limit cranking time. After a short wait (it seemed like less than a minute), the battery came back on line and I cranked some more. I don’t know the limits of this battery, but it was clearly cranking strong when the engine finally had enough fuel to start.
One concern with lithium batteries is that they won’t put out much power when cold. The procedure to deal with this is to pull some power from the battery. Its internal resistance will cause the battery to warm up.
Generally, people will either hit the starter for a short pull, or turn on their landing lights for a bit and then shut everything off and wait a minute while the heat in the battery spreads.
Because my battery is in the engine compartment and I don’t start my engine when cold soaked, I haven’t yet needed to test this strategy.
I have found that I see the highest temperatures on the Bluetooth thermometer when running on the ground and during climb out. In the originally installed configuration, I have seen temperatures as high as 132F. During cruise, temperatures were normally not much over 100F. I have yet to explain why the temperature labels have never indicated above 105F while I have seen up to 132F with the Bluetooth thermometer. Either way, 132F is too high given that summer is still to come. It is note worthy that Odyssey lists 113F as the maximum operating temperature for the plastic case version of the PC680. The Battery University web page lists the thermal runway temperature of LiFePO4 as 518F, and classifies the battery as “Very safe even if fully charged”.

The battery was removed and the box modified to include a cooling jacket. It draws air from the rear engine baffle, through a 1 inch SCAT tube, to a partial jacket that covers the lower front face and bottom surface of the battery box. The cool air is introduced at the top of the jacket, flows down the front face, across the bottom and exits at the rear bottom of the box. I figured that these two surfaces were most directly facing the exhaust pipes and introducing the most heat. I have one day of ski flying with this configuration. Temperatures while idling and taxiing on the ground and during climb out are down substantially. Generally well below 100F. While doing multiple back-to back takeoffs that included high powered taxiing through heavy snow, I was able to get temps nearly to 105F. This is a much harder test case than I had done previously.
However, my testing and data collection is ad-hoc at best, so no final conclusions are being drawn yet. I am planning to do some more testing with the temperature labels and placing the Bluetooth thermometer in different locations. I want to get a better picture of what is happening with temperatures and try to understand why I am getting such different readings with the two sensing methods. I will also do long term monitoring as the seasonal temperatures continue to rise. Mounting this battery in the engine compartment is still in question, it may need to be moved. Outside of that, the EarthX ETX900’s performance has been outstanding.
It is about the cheapest 10 pounds you can remove from your airplane without removing a “feature”.