This thread has been locked.

If you have a related question, please click the "Ask a related question" button in the top right corner. The newly created question will be automatically linked to this question.

CDCE62002: Is it possible to use this device without an XTAL input?

Part Number: CDCE62002

I am working on a design in which we would prefer not to include a crystal, but only a 16 MHz single ended LVCMOS reference clock.

Is it possible to program the device ahead for this option, then use it on a board which does not have a crystal input?

A related question - is there any way to program the device without first providing a 25 MHz crystal input? From the datasheet it is not clear which input clocks the SPI interface, but I am assuming it comes out of the smart multiplexer?

Thanks!

Rob

  • Dear Robert,

    CDCE62002 support an LVCMOS input reference. 16 MHz is covered. The SPI is a seperate port and device is programmed using pin 7,8,17 and 18. I hope this helps.

    Regards, Simon.
  • Hi Simon,

    The default programming expects a 25 Mhz clock or 25MHz Crystal.

    If I provide neither of those inputs, is it possible to still communicate over spi and program the device? Does the device rely on one of the other clock inputs in order to register the spi commands and shift them into eeprom?

    Thanks,

    Rob

  • A clock or crystal input to the PLL is not required to program the device via SPI interface.

    Alan
  • Alan,

    That has not been my experience.  I tried on both a custom board and the evaluation module. Here were the results:

    With no Crystal, the device does not respond to spi commands, and does not output a clock, even if the correct input clock is fed into the clock in of the smart multiplexer.

    If a clock of almost any frequency  (1-25 MHz)  is fed into the AUX IN (xtal in) pin, the device will become responsive to SPI and correctly output as configured.

    My theory is that the AUX IN pin is used to drive an internal clock domain which shifts values out of EEPROM at boot time, and also reads the SPI interface .

    It would be nice in the future if this were documented more clearly in the datasheet. Without a crystal or clock into AUX IN, this chip does not function, EVEN if the smart multiplexer is set to ignore AUX in.

    Thanks,

    Rob

  • That should not be the case.

    Can you do a register read back without a crystal input?

    Note that there will be no output clock if the VCO calibration does not complete successfully (outputs will byte muted/ squelched)

    Alan
  • Alan,

    A register read without the crystal gives all 0's on the dev kit, and all 1's (FFFFFFFF) on my custom board.

    Thanks,
    Rob
  • Can you please share your schematic and SPI write & read commands to review?
  • I'm sorry I can't share my schematic. However, I can tell you that when a clock is provided to AUX in, my SPI commands are successfully registered and the EEPROM can be programmed / read into RAM upon power up without any issues.

    You can see the behavior i'm describing using the Eval kit. All you need to do is remove the crystal and attempt to read registers over SPI. The device is non-responsive (and consumes almost no power) unless a crystal or clock is supplied to AUX in.

    Thanks,
    Rob
  • Dear Robert,

    I am sure you are correct.

    Is there anything we can do help you with this behavior?

    Regards, Simon.
  • Simon,

    I think this should be made more clear in the documentation, assuming it is how the device is designed to operate. If this is per design, the manual should mention that a Crystal or AUX input is required to operate the device.

    Thanks!
    Rob
  • Rob,

    We tested this on the CDCE62002 EVM without a XTAL on AUX input and CLK on PRI input. We were able to successfully communicate with the chip by reading SPI registers. I wonder if there is a RESERVED or TITSTCFG bit(s) that were modified from the default and stored to the EEPROM which may be causing unexpected behavior? Could you check that any reserved/test registers were not modified from the factory default; otherwise, correct the discrepancies and repeat the experiment?

    Regards,
    Alan
  • Alan,

    I will try this later this week and get back to you.

    To be clear - you had no clock inputs to the device at the time of reading from SPI?

    Thanks,

    Rob

  • Yes,  we had no clock inputs to the device at the time of reading from  SPI.