BQ79656-Q1: Ring Architecture: Only One CMU Responds in Reverse After COM Break

Hi Vashali,

Could you please send your autoaddressing and switching to reverse procedure? Please ensure the CMU5's Top of Stack register is removed, as well as the address in reverse direction is written. if CMU5's TOS register is not removed, it will not send commands to CMU2-4.

Best Regards,

John

Hi John Du ,

Thank you for your response.

• After detecting a communication break between CMU1 and CMU2,I clear the COMM_CTRL register on all stack devices using a reverse broadcast to ensure the previous Top of Stack setting is removed.

• I switch the bridge direction by setting DIR_SEL = 1.

• I perform dummy writes to re-synchronize DLLs on all CMUs.

• I enable auto addressing in reverse  direction via reverse broadcast.

• I assign addresses sequentially using broadcast writes to DIR1_ADDR.

• Then, I configure all CMUs as stack devices, set the bridge as base, and assign Top of Stack to the last CMU in the reverse path.

  Let me know if I'm missing any step in the reverse setup.

Hi John Du 

Is there any additional configuration?

Thanks!

Hi Vaishali,

I do not see any issue in these steps. Are you able to get any readings via singleread/stack read? Also given that CMU1 and CMU2 break, the Top of Stack assignment will be one less than before. Could you please send me a picture/diagram of your setup? I would like to see if there are any hardware related issues involved.

Best Regards,

John

Hi John Du ,

•  Both single read and stack read commands work only for CMU5. CMU4, CMU3,CMU2 are not responding.  
•  I’ve attached a picture of my setup. The break is between CMU1 and CMU2. 
• COMH of bridge is connected to COML of CMU1, and COML of bridge to COMH of CMU5.Thanks!

Hi Vaishali,

The setup should be fine given that the base is able to fully communicate from the north direction. Is it possible to check on startup, without any line breakage, that the device is able to communicate in reverse? 

Understanding if the device can communicate in reverse or not would give insight on if this is a hardware/procedure issue. If the device is still not able to communicate in reverse despite no line breakage, then there could be a potential issue of the VIF signals. If possible can you provide the logic analyzer files. 

Best Regards,

John

Hi John Du 

• Yes, on startup without any line breakage, I am able to communicate with all 5 CMUs in reverse direction.  .

• Sure, I can share the logic analyzer data.
  Do you want the capture during reverse communication without any line break? 

  Thanks!

Hi Vaishali,

I would like the logic analyzer data when device is failing to communicate with CMU 1-2 breaking, where it cannot communicate in reverse.

Thanks,

John

Hi John Du ,

I’ve captured the logic analyzer data for the reverse communication scenario where the COMH of the bridge is not connected to the COML of CMU1  

Please find the attached file.

5226.digital.csv

Thanks!

Hi John Du ,

I have captured the logic analyzer data for the case where there is a communication break between CMU1 and CMU2.

3326.digital.csv

Hi Vaishali,

I'm not able to interpret the csv currently with only 0s and 1s, is it possible to send some type of data that has either uart/spi or even pseudocode? If using Saleae, preferably the .sal file itself.

Best Regards,

John

Hi John Du  ,

I have uploaded the original Saleae .sal file that captures the reverse communication failure between CMU1 and CMU2.

Combreak.zip

Thanks

Hi John Du ,

     I wanted to check if you were able to look at the .sal file I sent for the reverse communication issue between CMU1 and CMU2.

Thanks,