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.

CDCE62002EVM: Configuration and programming

Part Number: CDCE62002EVM
Other Parts Discussed in Thread: CDCE62002, ALLIGATOR

Hi, I have the CDCE62002EVM, I want to use it to generate a 1Ghz clock. I've been trying to familiarize with the board but:

1- Any time I try to change the frecuency, I use the CDCE62002EVM GUI 1.2.1 frequency planner, set the desire frequency apply the calculated settings and click Write, and then /PD Calibration and nothing, the output goes flat. I push the reset button and goes to the factory default configuration and the signal appears again. What are the steps sequence needed to change the output frequency? any frequency

2- I read that the CDCE62002 can generate a square signal (clock) but the signal I get with the factory settings are sine wave. How can I change to the square clock signal?

  • Hi Juan,

    Please refer to section 6 of the EVM user guide which details step by step on how to control the EVM and write to the part.

    Default Config i am seeing on the EVM GUI and the Datasheet is 125 MHz LVDS on out 1 and 156.25 MHz on LVPECL out0. Both of these waveforms are square wave.

    Best,

    Sandra  

  • Hi Sandra,

    I've generated the defaults values of the EVM. My issue is why even though the EVM is supposed to generate a square signal, I am seeing a Sine wave on the oscilloscope. What am I missing? since the EVM user Guide doesn't have a troubleshooting section.

    In the guide says that saving the configuration is optional if the desired behavior is to start the EVM without programming it at startup. So, if I write the configuration in EVM using values obtained using the frequency planner of the GUI without saving it in the EEPROM, it should work. My issue is that I configure the board using the frequency planner to a 60Mhz frequency, without saving the EEPROM, and saving the EEPROM, and the result is the same, there is no signal. 

    Thank you for your help,

    Juan

  • Hi Juan,

    Can you share with me how you're measuring the clock into the scope? 

    Also based on your scope set up i would expect the waveform to look more like a sine wave than a square wave. 

    Can you update the scope to mimic below image? This image is from CDCE62002 EVM out Y0 with a math function to turn P and N channel to a single wave from. 

    Best,

    Sandra 

  • This is how I'm measuring the signal using the default configuration:

    and I get the sine signal

    I open the GUI tool, then I open the frequency planner tool and write the EVM with the Write button, and then press the PD calibration

    The signal disappears

    I need the square signal to use it as a clock to drive a custom counter, and obviously I need to change the signal frequency.

    Thanks,

    Juan

  • Hi Juan,

    I have tested the config with 25MHz LVCMOS input + 100MHz LVPECL on Y0 and Y1 disabled and could not replicate your issue. Attached below is the profile i had created for this test. You can import it on your GUI/EVM and test it out.

    Steps i have taken are using the frequency planner to write 25MHz input and 100MHz output on Y0 with Y1 disabled. Following that i performed a "/PD Calibration" then i performed a write to the EVM by selecting the "Write" button this is to ensure that this config has updated on the EVM. 

    https://e2e.ti.com/cfs-file/__key/communityserver-discussions-components-files/48/CDCE62002_5F00_2025_5F00_12_5F00_29.ini

    • I noticed that you're probing Y1, do you also observe signal loss on Y0?
    • Additionally, I see that your GUI is configured to 25MHz input clock from PRIREF but the image you shared of the EVM doesn't show a PRIREF input. Please make sure to feed in a clock to PRIREF or switch the reference to SECREF to use the on board 25MHz crystal. Not feeding in a PRIREF can cause signal loss on the output since its the GUI is configured to use PRIREF input clock.   

    A couple of points on the scope and measurement methods:

    1. Your scope image shows a 200MHz bandwidth which is filtering out the harmonics of the frequency. In order to get a sharp square wave you'd need a scope bandwidth (BW) that is 20 times the carrier frequency. For example measuring a 100MHz signal you'll need a scope with a 2GHz BW.    
    2. Looks like you're using banana/alligator clips, this can also contribute the signal distortion.
    3. Note that the ground pin of the scope would need to be connected to ground rather than the N channel. You can connect it to the external part of the SMA for a ground connection. 
      1. Note since you'd only be probing the one of the channels you'll observe a half swing
    4. I'd recommend using a scope probe rather than jumper cables for better performance measurement and a higher BW scope.   

    Let me know any issues you face following these update.  

    Best,

    Sandra

  • Hi Sandra, thank you for your advice. The issue I had is solved with your advice. Now, I have a few remarks and a question.

    I changed the frequency to 20 Mhz since I have a 200Mhz Oscilloscope. 20 Mhz is the lowest frequency using no external components or signals with the EVM.

    The signal I obtained was like this:

    I looked at the loop filter tool, and one of the parameters are out of range

    Here is the Closed Loop graph:

    I'm guessing some of the harmonics are being filtered. So, I'll have to use the external filters configured with the SW9 and SW10 to improve the signal?

    Do I have to include a termination hardware to improve the signal?

    If I want to use the signal generated with the EVM as the clock signal to drive a counter (MC100EP016A), is there any additional recommended hardware to include in between?

    Thanks,

    Juan