Post Go back to editing

AD7264 is not initiating conversion

Dear All,

Could you please give us some hints on what might be happening with AD7264 that we have. The device is configured in control register mode (not pin driven). the Vdrive is 3V.

Basically we have AD7246 on a custom board and the SPI control/data transfer is implemented inside a FPGA via custom IPs. We can read/write to registers of the AD7264 fine. You can see an example diagram from a network analyzer (e.g. written 0x191 to the control register) and the output data on MISO (from the AD7246) matches that. This configuration is supposed to set the PD0 to high and enable calibration with unity gain.

This however does not start any data conversion on the ADC. Please see below. Nothing is written to MISO line from the ADC. Maybe it is in power down mode but I thought that doing 0x91 to the control register will set ADC to ON? 

For reference I attach the timing diagram from the SPI controller on the FPGA. It is probably difficult to see but the clock cycles are following the data sheet recommendations.  

The SPI clock is running at 30.72MHz but also tried lower clock with no much help.

Any ideas what could we be doing wrong? 

Please ignore those glitches in the network analyzer output. 

Any comments will be much appreciated. 

Many thanks

Milos

PLOTS:

Writing 0x191 to control register. 

  

Listening for data on ADC MISO.

Timing diagram for SPI registers read and write

ADC connection

Edit Notes

adding tags
[edited by: Lluis at 3:30 PM (GMT 0) on 3 Aug 2020]
  • Hello,

    You are right that if the AD7264 is powered up in control register mode, writing to the device powers it up. You are using the internal reference, are you also waiting the 240us required?

    Have you checked reference is powered up through the test points?

    Are you getting 0x1FF independently of the input voltage?

    Does powering the comparators on make any difference?

    Regards,

    Lluis.

  • Hi Lluis,

    Thanks for the reply.

    I managed to get something going in the meantime. It looks like when I was setting the control register for read immediately after i was writing to it the control register was getting set to all ones and thus probably switching off the ADC (bug in software).

    Now in the process of evaluating what it outputs.

    Thanks

    Milos