SPI-to-FTDI example - adding digital isolators

        I have a question about the SPI-to-FTDI example.  This example shows how to interface the ADuCM350's SPIH port to an FTDI chip, the FT4222H (using FTDI's UMFT4222EV eval board).  I want to add the Analog Devices' ADuM3150 SPIsolator chip (4-ch of digital isolation using on-chip transformers) to this SPI bus.

      The ADuM3150 isolator chip has two extra low-speed channels, which I'm not using.  It also has a "DelayClk" signal, which I'd like to use because it helps with skew between the various SPI signals.

      My question is:  Can I use this feature of the ADuM3150 with my ADuCM350 SPI master?  That is, can you tell me how I would wire this DelayClk signal to the ADuCM350?  The datasheet and application notes for the ADuM3150 aren't very clear about this type of situation (the application notes mostly talk about standalone A/D chips).  There also isn't any forum for the ADuM3150.  Thank you.

160529_SPI-to-FTDI_isolation_question.pptx
Parents
  • 0
    •  Analog Employees 
    on Jun 7, 2016 8:13 AM

    Hi Anita,

    Apologies for the delay in getting back to you. I am not familiar with these isolation parts and had to find out the information. I got the following information from the applications engineer responsible for the ADuM3150:

    "The DCLK signal provides a copy of the master’s clock with the round trip propagation delay through the digital isolator added in. The round trip prop delay is trimmed at test instead of using an isolated data channel to wrap it back to the master side.

     

    You end up with the delayed clock and miso signals synced up on the master side of the isolator. With SPI, there isn’t a clock input on the master. So to make use of the DCLK and higher clock speeds (up to 40MHz), you need a register available to clock in the data from the MI output pin with DCLK."

     

    I don't think the ADuCM350 can support this feature, as it doesn't have a register available to clock in the data from DCLK. But my understanding is that this is only really required for higher clock speeds (up to 40MHz), so this feature is probably not necessary for the ADuCM350 anyway.

     

    Rgds,

    Eoghan.

Reply
  • 0
    •  Analog Employees 
    on Jun 7, 2016 8:13 AM

    Hi Anita,

    Apologies for the delay in getting back to you. I am not familiar with these isolation parts and had to find out the information. I got the following information from the applications engineer responsible for the ADuM3150:

    "The DCLK signal provides a copy of the master’s clock with the round trip propagation delay through the digital isolator added in. The round trip prop delay is trimmed at test instead of using an isolated data channel to wrap it back to the master side.

     

    You end up with the delayed clock and miso signals synced up on the master side of the isolator. With SPI, there isn’t a clock input on the master. So to make use of the DCLK and higher clock speeds (up to 40MHz), you need a register available to clock in the data from the MI output pin with DCLK."

     

    I don't think the ADuCM350 can support this feature, as it doesn't have a register available to clock in the data from DCLK. But my understanding is that this is only really required for higher clock speeds (up to 40MHz), so this feature is probably not necessary for the ADuCM350 anyway.

     

    Rgds,

    Eoghan.

Children
No Data