ADRV9364-Z7020 flashing power sequencer

We've designed and manufactured our own custom power-only (5V only) carrier for the ADRV9364-Z7020. However, we haven't been able to successfully flash the power sequencer as per the instructions here: PicoZed Power and Sequencing [Analog Devices Wiki] . Every time we try, we immediately get a read error. Digging a little deeper, we found that the program is unable to read from /dev/i2c-0, which presumably points to the ADM1166's EEPROM.

Is there a quick-fix solution to this issue? We're on a pretty tight schedule.

Many thanks,

Frank

EDIT:

Took a look at the default device tree under zynq-adrv9364-z7020-bob in the latest Zynq-based Linux image and found this:

i2c@41600000 {
compatible = "xlnx,axi-iic-1.02.a", "xlnx,xps-iic-2.00.a";
reg = <0x41600000 0x10000>;
interrupt-parent = <0x1>;
interrupts = <0x0 0x3a 0x4>;
clocks = <0x2 0xf>;
clock-names = "pclk";
#address-cells = <0x1>;
#size-cells = <0x0>;

ad7291@20 {
compatible = "adi,ad7291";
reg = <0x20>;
};

ad7291-brk@2C {
compatible = "adi,ad7291";
reg = <0x2c>;
};

eeprom@50 {
compatible = "at24,24c32";
reg = <0x50>;
};
};

There is no entry for the ADM1166. How can I modify the reference design to include read and write capabilities to the ADM1166's EEPROM?

  • So it turns out the ADRV9364-Z7020 is nothing like the ADRV9361-Z7035 except for in name, which I had mistakenly assumed. In fact, the ADRV9364-Z7020 doesn't have the ADM1166 at all!

     

    Still looking through the schematic now but it seems that D3 (the PGOOD LED, which doesn't light up on the SOM when I plug in my custom carrier) is driven by a valid 3.3V input. However, D4, i.e. CFG_DONE, does light up.

    Is this 5V power-only carrier I made salvageable?

  • 0
    •  Analog Employees 
    on Sep 27, 2017 5:21 PM over 3 years ago

    Hi,

    If CFG_DONE lights up, the only obvious reasons for PGOOD not to work would be:

       - U6 (VDDA_GPO) regulator is not working and keeps the PG_MODULE line low (less likely)

       - JX2 pin 11 (PG_MODULE) is shorted to GND on your carrier (more likely)

    Please check if either of the above are issues on your setup.

    Mihai

  • Thanks, Mihai. We had actually already figured it out earlier after tracing voltages through the entire board - turns out that JX4, pin 9 (VDDA_GPO_PWR) was improperly shorted to GND on our carrier. This resulted in the output of the 2.5V regulator being forced to ground, which caused the ADP7102 to ground PG_MODULE immediately. After applying a tiny piece of tape to the pin to act as an insulator, we were able to successfully boot the SOM.

    Best,

    Frank