Rev B passed the tests! ✅ Rev B passed the tests! ✅

Rev B passed the tests! ✅

The results are in! 🎉

A few weeks ago Rev B landed on the bench and firmware v1 was finished. Since then we have been testing every part of OpenBMS on a real battery, and we can finally say it: Rev B works!

The test setup 🔧

Every test ran on the same setup: OpenBMS Rev B with firmware v1.0.0, a 7S1P Li-Ion pack (Samsung INR18650-20R, 2000 mAh), the 600 W resistive load from the OpenBMS Test Bench and a CC/CV lab charger. A PC logged everything over USB-UART.

OpenBMS Rev B test setup with battery, load, charger and host communication

Learning test (HPPC) 📈

The board ran the learning test on its own, once in the charge direction and once in the discharge direction. It built a model for all 7 cells and saved it in the EEPROM. The measured capacity came out at 2.04-2.05 Ah per cell, right in line with the 2.0 Ah datasheet value.

Fuel gauge 🔋

We ran 7 SOC tests: charging and discharging from 0.24C up to 6.5C, a zigzag cycle, a pulse test and a charge in the freezer. Across all of them, from -12 °C to 61 °C, the SOC error stayed between -1.8 % and +2.8 %.

TestCurrentBattery temperatureMin errorMax errorAverage error
Charge 1C1.95 A23.5 to 30.6 °C-1.4 %+1.0 %-0.1 %
Discharge 2C2.9-4.6 A25.7 to 40.9 °C-0.9 %+0.3 %-0.1 %
Charge 2.7C5.4 A29.6 to 39.8 °C-0.6 %+2.2 %+0.3 %
Discharge 6.5C11.6-15.1 A26.3 to 61.0 °C0.0 %+1.0 %+0.3 %
Zigzag1.9-4.6 A26.2 to 40.3 °C-1.8 %+1.0 %0.0 %
Pulse test5.0-15.2 A28.6 to 51.4 °C-0.1 %+0.9 %+0.4 %
Charge at -11 °C0.48 A-12.4 to -11.3 °C-0.4 %+2.8 %+0.5 %

The error is the fuel gauge SOC compared with a reference from coulomb counting the logged current, averaged over 30 s.

Protections 🛡️

We tested every software protection: over-voltage, under-voltage, over-current (charge and discharge), and over- and under-temperature, including the separate limits while charging. Each one switched the FETs off at its limit.

On top of that, the TPS12000 FET controller protects the pack in hardware, without any help from the firmware. It tripped at 19.9 A on over-current, 16.77 V on under-voltage and 33.1 V on over-voltage.

Cell balancing ⚖️

We pulled two cells out of balance on purpose and then charged the pack. The balancer bled the fuller cells, and the SOC spread between cells dropped from 3.7 % to 1.2 %.

Stress test 💪

Finally, we pushed the power path. OpenBMS carried 19 A continuously for 30 minutes, and the STM32 stayed at 42.4 °C max. Then we switched the main FETs on and off 1500 times at 19 A. No trips.

Continuous discharge at 19 A: current, STM32 and pack temperature

Anything found? 🔍

No board is perfect, and Rev B is no exception. We found two small hardware issues, the pre-charge FET and the current calibration switch. Both have a simple workaround with little or no impact, and both will be fixed in the next revision. Everything else on Rev B works as designed.

What’s next?

We are approaching the end of the project and getting ready for the initial release of OpenBMS. Together with it we’ll publish the full test report and the Rev B bring-up notes, so you can see every plot and number yourself. Want to build your own? Have a look at Get started. Stay charged! ⚡


← Back to blog