BQ76PL455A-Q1: Communication Problem in daisy chain with 12 cards.

Hi Mehmet,

The Use case scenarios app note here: www.ti.com/.../slua785.pdf has many recommendations for daisy chain communication throughout the document. Could you please provide details on your daisy chain isolation schematic? And provide more details on the failure - how many devices can communicate and how many devices do you have total in the stack?

Thanks,

Taylor

We have one master card with MCU on it, and 11 slave cards with bq76pl455. Master card communicates with first slave card over UART. Then daisy chain communication is used between other slave cards. In the daisy chain circuit, we use the design you recommended for noisy environments. There are two common mode filters and one capacitor at the daisy chain output.

Generally 8 of the devices can get an address. For example, 9. 10 and 11. card cannot be controlled from the master as the 8. device closes its higher daisy chain channel. We are using bq Test program to reach the cards.

Hi Taylor, I have given the details of our topology in my previous post. Yesterday, we got measurements directly from the COM input pins of bq76. When we measure the pulse widths at the last card of the daisy chain, we have seen that pulse widths are wider than it is expected. We are using a cable with a 30 pf/m which seems good. The total length of daisy chain cables starting from the first card to the eleventh card is approximately 3 meters or a bit longer. However, we cannot communicate with the last card in the network. I think it is the reason of cable capacitance. And also I must note that the isolation capacitors are 1 nF. We are planning to increase their values up to 2.2 nF but I think it will not solve the pulse width issue. Can you give me other suggestions for me to solve this communication problem? Thanks.

Hi Mehmet,

Can you please reattach the details and schematic? I cannot find the other post with these details. Have you tried a different device as the last card in the stack and have the same result? And is the signal continuously becoming degraded across each card in the stack?

Regards,

Taylor

Hi Taylor,

We have one master card with MCU on it, and 11 slave cards with bq76pl455. The master card communicates with the first slave card over UART. Then daisy chain communication is used between other slave cards. In the daisy chain circuit, we use the design you recommended for noisy environments. There are two common mode filters and one capacitor (1 nF) at the daisy chain output. 

The signal degrades when it passes over slave cards. And it cannot reach in a correct way to the last boards. After changing cable used in the daisy chain, we finally reached 10 of 11 slave cards. But we cannot communicate with the last one. We changed the last board if there will be a problem on it but nothing changed. When we looked at the COM lines at the last board bq76 inputs, pulse widths are higher than 125 ns. I think cable capacitance effects communication. We decided to increase the isolation capacitance value up to 2.2 pF.  But I think this will not solve pulse width problem.

Could you please suggest us a way to solve this problem on the daisy chain.

Thank you

Hi Mehmet,

Yes I understand the description, but what I am looking for is the schematic to see if there is something going on or it just could be the equipment/cabling. If you have the schematic that would be very helpful otherwise we cannot debug this properly. I see that you mention that the total length is 3m across all boards so does that mean the daisy chain cable between each board is about 27cm?

Regards,

Taylor

Hi Taylor,

Ok. I understood. Here is the schematic diagram of the communication part.

And the pack cabling.

Thanks for your help.

Hi Mehmet,

Is this problem solved or still having problems?

I think you check the pulse and it was higher than 125 ns.

What is width between 10 and 11, and between 10 and 9.

You might want to use shorter cable for trouble shooting.

Roger

Hi Roger,

I think we have solved the problem but not yet checked the system under real operating conditions. We increased the value of isolation capacitors and changed the cable with a low capacitance equivalent.

Thanks for your response.