Tritium IQ BMS

From Black Metal Forest
Jump to: navigation, search

A Battery Management system is crucial to a safe and well performing Lithium Battery pack. is a beautiful rescorse of information on BMS desgin and helped me to pick out a BMS perfect for our solar car application.

The Tritium IQ BMS is an very well throughout system thorough and thorough. Precises and redundant ADC measurements, flexible settings, and smart common sense design. The manuals were a little lacking in some areas, namely there was hardly a mention of the configuration utility which is complex enough to warrant it's own manual.

The Main Issue

After we finished assembling our pack, our battery pack sat dormant for months because of bazaar electrical issue. The BMS would begin its careful start up routine, only to abort very quickly. Due to the timing of the abort, the problem presented itself with a load of plausible red herrings. Wrong configuration settings? Not enough current for contractors? High voltage isolation problem detected? No capacitive motor controller on the output? "Smart Mosfet"s on the relay outputs having their protection tripped somehow? In total I spent at least a full week of work on the issue. I made a custom CANbus message decoder with an arduino in order to decode the raw BMS data. Tritium would not answer my emails at the time. In the end I got a hold of an engineer on the the phone and he solved the problem with a few questions!

The answer

It is complex. So the BMS has two power inputs, one for the logic and one for the high current relay power. The BMS begins it's startup at the moment it senses the 12V secondary line has power at its input. The BMS microcontroller activates mosfets that in turn route the secondary 12V powerline into the main relays. Without my knowledge, the grounds of the microcontroller and the ground of the secondary supply are not connected on the PCB. The micro can not control the Mosfets correctly with out a common ground reference, for for a quick second, right when the secondary supply is plugged in, the voltage difference would be correct but then float out of range, and the Mosfets turn off. I had my Oscilloscope hooked up during my trouble shooting but diden't know what I was looking at... The control voltage would go low to turn the Mosfets on, but then would rise up like a capacitor and turn the Mosfets off again.

Our tritium BMS is Hardware version 5. In previous versions I was told that the grounds were connected together, but because people also would have the two power rail grounds connected together somewhere else in the vehicle, it made a "ground loop" that would cause the circuit to bug out. Their solution was to disconnect the grounds from each other and require the user to connect them. The manual says way at the bottom in the CANbus section that the two power rails have to have there grounds connected together someware on the vehical. Aw well it was my fault in the end