PHET Lithium Battery Testing

10th August 2007

In July 2007, I imported some lithium batteries from a Taiwanese company called Pihsiang Energy Technology (PHET), to investigate the possibility of building an EV-sized battery pack from a large number of small cells (similar to what is done with the Tesla Roadster). PHET are an official licensee of the Phostech patents for Lithium Iron Phosphate (LiFePO4) battery technology, and produce a range of different LiFePO4 cells.

The ones I am testing are called PE-1150, which are their "high power" 18650 sized cell, rated at 3.2V 1.1Ah nominal. Their specifications indicate they last over 800 charges at 10C continuous discharge, or more than 2000 cycles at normal power levels, which means they could be a great battery for high performance EVs. However, it's important that I confirm this by testing some sample cells before going ahead and spending $10K or more on enough cells for the MX5 conversion!

So for the benefit of others in the same boat, I've documented my findings here.

On the far right you can see the resistor network used as a load for the battery during discharge. The battery is the red cylinder, held between pressure contacts (using conductive silver paste to maintain a good electrical connection). Left of the battery is the 20A 75mV shunt used for current measurement, and below it are two relays for connecting battery to PSU (charging) or resistor network (discharging).

The prototype board includes relay drivers (NPN Darlington transistors), an LM833 op-amp for amplifying voltage drop across the shunt, and two voltage dividers (multi-turn potentiometers) for reducing voltage to the DAQ's required range's (0-2.5V).

On the left is the LabJack U3 data aquisition module, providing 12-bit Analogue to Digital conversion (ADC) for the voltage and current measurements, two outputs for switching the charge/discharge relays, and USB connection to computer.

The power supply has two outputs - the one on the left is set to 12V and is used for powering the relay drivers.

The output on the right is set to 3.70V and used for battery charging, with current limit set for during constant-current portion of charge cycle. Voltage and current were both calibrated using a Fluke 179 multimeter (which is accurate to 0.1%, since the PSU's meters really aren't accurate enough.)

This is just a quick logging application I put together for controlling the battery cycle sequence, logging and graphing the current and voltage, and calculating overall amp-hours and watt-hours in and out of the battery.

The main graph area shows voltage in white and current in cyan for current in to the battery (i.e charging) and orange for current out (i.e discharging). The transition you see in the middle of the screen is the automatic switch from charging to discharging, which occurs when charging current drops to 0.2C.

Below the main graph is realtime feedback of voltage and current (left), then fields showing total current and total watts in (middle) and out (right) of the battery. Left of the graph is a CSV log of the voltage and current, in case I want to process the data in an external application.

Most sampling is done at 1Hz (once per second). One "cycle" involves discharging the battery to 2.0 volts, charging to 3.65 volts, then discharging back to 2.0 volts, measuring energy put in during charging, and energy taken out during the second (full) discharge.