• State TI Thinks Resolved
  • Locked Locked
  • Replies 10 replies
  • Answers 1 answer
  • Subscribers 64 subscribers
  • Views 382 views
  • Users 0 members are here
Support feedback
Options
Options
Similar topics

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.

BQ79616-Q1: Daisy chain communication troubleshoots

José Gabriel Oliveira Pinto
Prodigy 10 points
Part Number: BQ79616-Q1
Other Parts Discussed in Thread: BQ79616

Tool/software:

Hi,

 I am currently designing a modular BMS with a master-slave topology. The master PCB has a microcontroller and a BQ79600 and the slave PCB has two BQ79616s to monitor and balance 22 cells in series. The PCBs are connected via the daisy chain pins of the BQs.

 

However, I have successfully communicated with the slave before, directly through the UART and through the daisy chain, now only the UART communication works. Through the daisy chain I can only wake the slaves and then the communication fails when I send a read command.

I can communicate with the TI BQ79616 Evaluation Module (EVM) using my master board, the same one used to communicate with my slave board, and the same software.

 

In the attachment there are two photos with the COMMH of the slave board and the EVM, the voltage in my slave board is slightly lower (blue line), is this difference relevant for it not to work?

 

Also attached is the schematic of the slave board, which is based on the schematic of the evaluation board.

 

Does anyone know where the problem might be coming from?

Best regards.

 AMS_Slave.PDF

  • 0
    TI__Expert 5170 points

    Jose,

    Although we do recommend a transformer between PCBs, it appears your waveforms are very similar.

    I have a few questions:

    - Where exactly on the schematic are you probing? We recommend probing COMLP-COMLN.

    - Are your current and voltage references typical (ie. the device is not drawing a huge amount of current, or you are applying a large amount of voltage)?

    - What are the maximum and minimum voltages that you are reading?

    Regards,

    Ben

  • 0
    Prodigy 10 points

    Hi Benjamin,

     

    I also attach the schematic of the master PCB where a transformer is used to isolate the communication.

     

    To answer the questions:
    - I made a mistake regarding the probing location, we are indeed probing between the COMLP-COMLN. And for reference, the purple signal is probed after the transformer, between the isolated COMLP-COMLN. The subtitle of the images is also swapped, as refered in the text, the image where the blue line have a lower amplitude is the comminication between the Master and Slave PCBs.
    - At the moment I am feeding the BQs through a power supply and at 30V they are drawing about 20mA each, with a LED indicating that the BQ is awake, i think this is a normal value.
    - The voltages I am reading through the UART are around 2.7V powering the BQs through an external power supply with 30V and a resistor ladder to simulate the cells voltages.

    BMS_Master.PDF

  • 0 in reply to José Gabriel Oliveira Pinto
    TI__Expert 5170 points

    Jose,

    For the minimum and maximum voltages, I was interested in the voltages on the COM pins, as shown in the oscilloscope shots. Could you also measure before the transformer, as close to the pins as possible? Additionally, could you tell me what commands exactly you are sending? 

    Regards,

    Ben

  • 0 in reply to Benjamin Luty
    Prodigy 10 points

    Hi Benjamin,

     

    Sorry for the misunderstanding, the minimum and maximum voltages in both master (on BQ79600 COMMH pins) and slave (on BQ79616 COMML pins) sides can be seen in the figure below, as well as the communication period.

     

    Pictures

     

    The commands I am sending are those described in the BQ79600 software design datasheet. The sequence is shown in the picture below. On the first read command of the auto-address sequence, the BQ79600 sets the SPI_RDY line to zero and never sets it to one. The same sequence works perfectly with the EVM, but not with my hardware, but it worked with both before.

     

     

    Regards,
    Jose

  • 0 in reply to José Gabriel Oliveira Pinto
    TI__Expert 5170 points

    Jose,

    It looks like the commands sent are okay.

    Here in section 7.5 of the datasheet you can see the range of values that are accepted as logic zero by the device:

    I cannot tell for certain but it seems like the voltages on the COMM lines are going sufficiently high and/or low to be registered as proper communication. It is possible that during the WAKE tone/ping the voltage is low enough for long enough that the device wakes, but for the communication it is not reaching the proper levels.

    Regards,

    Ben

  • 0 in reply to Benjamin Luty
    Prodigy 10 points

    Hi Benjamin,

     I measured both the COMMH and COMML voltages to GND and summed them on the oscilloscope to calculate the common mode voltage. The math channel is calculating CH1+CH2, this is, the double of the common mode voltage. As you can see in the pictures, the values of the math channel on the EVM board are between 4.82V and 5.22V and on the slave board between 4.74V and 5.22V. Dividing the values by 2 to obtain the common mode voltage, on the EVM board we get values between 2.41V and 2.61V and on the slave board values between 2.37V and 2.61V. I think these values respect the requirements.

     Regards,
    Jose

  • 0 in reply to José Gabriel Oliveira Pinto
    TI__Expert 5170 points

    Jose,

    Can you read the fault registers from the BQ79600? Can you read CVDD as well? Additionally, I would like to confirm this is not an issue unique to the IC, so if possible could you switch the IC with the one that was working in the EVM? If you cannot switch the IC could you try with a different device?

    Regards,

    ben

  • 0 in reply to Benjamin Luty
    Prodigy 10 points

    Hi Benjamin,

     

    I can read the error registers of the BQ79600 just fine and they all come back as no errors detected. The CVDD is at 5V after the wake ping. I have not swapped the IC with the one on the EVM board, but I have soldered another board with a new BQ79616 and the problem is the same.

     

    I am also able to communicate with the second BQ79616 of the slave board in daisy chain with the BQ79616 EVM board, bypassing the first BQ79616 and connecting the master COMMH directly to the COMML of the second BQ79616. I think the problem is on the first BQ79616, but I still haven't found the problem.

     

    Kind regards,
    Jose

  • 0 in reply to José Gabriel Oliveira Pinto
    TI__Expert 5170 points

    Jose,

    When attempting to communicate with the first 616 could you probe on both sides of the transformer to see if that is causing some sort of disruption?

    Regards,

    Ben

  • 0 in reply to Benjamin Luty
    TI__Expert 5170 points

    Jose,

    Could you also tell us exactly how you are bypassing the first device? Specifically, could you point out where on the schematic you are making the connections?

    Regards,

    Ben