This thread has been locked.

If you have a related question, please click the "Ask a related question" button in the top right corner. The newly created question will be automatically linked to this question.

BQ78350-DEVICE-FW: Unexpected behavior in temperature and Ext Avg Cell Voltage in R3 firmware

Part Number: BQ78350-DEVICE-FW
Other Parts Discussed in Thread: BQ78350, EV2400


We encounter problems in production regarding temperature and Ext Avg Cell Voltage. 

First, we noticed the Gauge's internel temperature could not be calibrated. The temperature needed a too big offset for calibration. We kind of ignored the problem.

But then, we got reports of batteries jumping back to 5% (EDV2) way too early. When we received them, the Ext Avg Cell Voltage reads close to 0. And after removing the BMS from the battery and then reconnecting it, the problem is gone. 

Then we also noticed, those exact boards have the temperature problem. After flashing the .srec, it may or may not heat up to 10 degrees above ambient, causing the unexpected temperature calibration, but at the same time the Ext Avg Cell Voltage is wrong. We checked the hardware multiple times, but no faults have been found. 

We still don't know what's going on, but these problems only appear on gauges flashed with device firmware R3.0

Can you please help us debugging this problem?

We plan to measure the BAT pin input voltage and also the current consumption when in the faulty state.

  • This sounds like a defect with the voltage ADC in the gauge, especially if the internal temperature measurement doesn't work. How many units show this behavior?

  • The ADC was also my guess, but the reading is still fluctuating a bit so there is definitely coming data from the ADC.

    We did not count the number of boards because we didn't think it was a big deal, but I guess it's now approximately 15 boards with this problem out of 50 or 60 this year. 

    Last year we have programmed more than 250 boards without issues, but with R2 firmware.

    The boards are not changed, and are still from the same production batch.

  • The gauge shares the ADC between cell voltage measurements and internal temperature measurements. If just the external measurements were incorrect, I'd suspect a problem with the connection (e.g. cold solder joint) but if the internal temperature measurements are also incorrect, then the likely reason is the ADC.

    I haven't seen issues like this with R3 so chances are that something is wrong with the ADC in the units that show this problem.

  • That could indeed be the case.

    I doubt this is the case. We are also using this chip on another type of BMS, and we also see this problem on that one, with approximately the same failure numbers. They are produced at another date, and actually with another batch of BQ78350 chips. 

    What is the chance of this being related to corruption while programming? We have seen one chip being programmed successfully and after a while, it was dead. Then we needed to program it again with the .srec. Most likely due to corruption. We also have lots of problems with communicating using the EV2400 tool.

  • The gauge won't start back up in gauging mode unless the FW was programmed correctly (it has integrity info that won't be programmed if anything got corrupted). It will stay in FW update mode and can be reprogrammed.

  • If that is the case, it seems to be a bug in the R3.0 firmware. What are the steps TI is going to take to solve this problem and what should we do?

    But anyways, nice to know the gauge cannot have one bit wrong before it starts. 

  • Again, what is TI going to do and what should I do to solve this problem? 

  • I measured the pins on the BQ78350 chip, and the results are interesting.

    VEN (pin 12) is pulsing like it should.

    BAT (pin 7) is low. When measuring ohms with my multimeter, I measure 35 ohms between GND and BAT. My voltage divider connected to the pin has 13K7 pulldown and 300K pull up, so that 'short' is definitely caused by the chip. 

    The temperature is approximately 12 degrees off.

    It seems like the ADC is just working fine, but the port multiplexer seems to hang.

    This time the problem appeared after flashing the .srec for the second time. The first time flashing nothing was wrong.


    I disconnected the battery and then reconnected the battery, so the chip is fully power cycled. Now the input measures 13K7 between GND and BAT. Also, the temperature measurement is correctly working again. When measuring with my oscilloscope, I get a nice pulse on the BAT pin, which was not present before.

    This is the last information I needed to be sure this is your bug and not our bug.