Battery Capacity Test

Battery Capacity Test

Pictured below are the electronic loads I used for testing the EM battery. (It's funny, not long ago I literally cut that rack-mounted active load in half because it was so heavy and unwieldy.) Each half is rated at 55 VDC and 200 A (but with a power limit of 750W) and can be used independently. I had intended to sell one half, but now expect I'll keep both. At least now the halves are easier to move around.

The testing was completely manual. The active load was manufactured by AC-DC Electronics (P/N EL750B). I used a Chinese PZEM-051 power meter (small blue glow in the background) a stopwatch and a spreadsheet. I took readings every 5 minutes. I began with one active load and a 10 amp draw, which took about 2 hours to deplete the battery. For the second test, I used both active loads in parallel and drew 20 amps. This discharge took about an hour. Although when riding the actual battery drain is anything but constant, these tests are representative of what to expect – and that is 1 to 2 hours of riding, depending on how “hard” it's being ridden.

Test apparatus: AC-DC Electronics EL750B electronic loads, PZEM-051 power meter, stopwatch, spreadsheet

Test Takeaways

The EM 5.7's battery is rated at 25 Ah. So a discharge rate of 25 amps is called the 1C rate. My 10-amp test was 0.4C (10/25) and the 20-amp test was 0.8C (20/25). Generally, the more slowly energy is withdrawn from a battery, the more it will provide.

Starting with a fully-charged battery and a 10-amp draw required the load to initially consume about 500 watts. The power dissipation diminishes as the battery's voltage drops. I could have drawn up to about 15A and stayed within the 750W capability of a single active load.

Using both active loads in parallel, I could draw 12.5A with each for a 25-amp (1C) test, but I don't see much point in doing that now.

The actual test tables are available as a downloadable spreadsheet, but the main takeaways are:

  • The 5.7's battery provides, at best, only about 80% of its rated capacity. Maybe that should not come as a surprise – it was probably manufactured in 2013. However, it also does not have a lot of cycles on it (as I discovered after learning how to communicate with the BMS).

  • The LED state-of-charge indicator appears more linear at a constant discharge rate than it does in actual use on the bike. (Although the end comes pretty abruptly in both cases.) Also, the data looks slightly more linear than it is because I did not indicate the exact time an LED transitioned. This makes sense as I learned after opening the battery -the SoC indicator is simply a voltmeter having no interaction with the BMS.

  • A fully-charged battery produces about 53 volts. This is a cell voltage of about 4.08 volts (53/13 = 4.08).

  • A fully-discharged battery produces about 40 volts. This is a cell voltage of about 3.08 volts (40/13 = 3.08).

  • When the BMS determines the battery is completely discharged, it will no longer allow it to provide any power and the bike stops dead. This is a safety feature as discharging a cell below a certain point (which varies by cell chemistry and from manufacturer to manufacturer) can cause irreparable damage.

PZEM-051 DC power meter

Measurement current shunt connection

Battery test cable, #2 AWG welding wire

EM 5.7 Discharge Tests.ods