Post Go back to editing

SPI read issue with AD9204-80EBZ eval board


I am currently experiencing an issue while reading data through SPI from the AD9204 chip with an external microcontroller. I connected the micro SPI signals to the eval board connector (J301-2 -> SPI_CLK, J304-2 -> SPI_CSN, J301-6 -> SPI_MOSI, J301-8 -> SPI_MISO, and the ground ) but I cannot see any data back from the AD9204 when sending a read request. The ADC chip runs with the board oscillator (J605 is shorted) and no input signal is provided. As additional info, the reference voltage levels on P103 and P102 are OK and I can see an output running clock on DCOA and DCOB pins.

This is what I can see from the oscillope:

The chip does not react also for different addresses. Is there something that i am missing ?

  • Hi ManuR,

    Here are a few initial questions.



  • Hi Doug,

    thanks for your interest and reply.

    * jumpers on j302 are placed (between 1-2, 4-5 and 8-9)
    * I also measured the spi signals after the level shifters (on J302) and I see signals as expected (same waveform but different scale)
    * I have also looked at the ref documents you sent me, but I believe to follow any reccomandation.

    Maybe I am just missing some major steps that I cannot see (the used protocol is wrong, miss of a specfic voltage supply reference somewhere, any other steps to follow after power-up and before the first read... ).

    Is there a test which I can try to see if the SPI interface is alive (e.g. writing to an AD register which is supposed to cause some external effects ) ?

  • Hi ManuR,

    If you read Register 0x00, it should come back with a value of 0x18 by default.

    One thing I forgot to mention before is the transfer bit. Address 0x08 to Address 0x18 are shadowed. Writes to these addresses do not affect device operation until a transfer command is issued by writing 0x01 to Address 0xFF, setting the transfer bit. Please see the datasheet for more information.

    I wonder if this is part of the issue.



  • Solved. It was probably crosstalk noise on the SPI master side. thanks

Reply Children