Post Go back to editing

Jitter in AD9144-FMC-EBZ

Category: Hardware
Product Number: ZCU102+AD9144-FMC-EBZ
Software Version: 2021.2

Hi,

I am using ZCU102 + AD9144-FMC-EBZ, in JESD mode 0, with meta-adi linux.

When I use the DAC-FMC-EBZ reference design, I get a jittery sine wave on the scope:

I turn off one of the tones, and put the other at scale of 0.8 and 40 MHz,

and I see that the sine wave is "breathing" - its maximum amplitude is going up and down.

When I draw a histogram of the voltage peak-to-peak of my sine, it is quite wide,

with values from 1.04V to 1.047V. (7mV jitter)

I haven't changed anything in the reference design, and was excepting much better performance.

(Our red pitaya and waveform generators give us much more stable sine wave).

What can be the source for this?

I thought it can be the clock source- is it OK to use a 2GHz sine wave as the source (or does it have to be a square wave?)?

Any idea or direction to get rid of the noise will be very helpful,

thanks!

Thread Notes

  • Hi, On behalf of the EngineerZone Community team, we apologize for the replies to your inquiry that were accidentally deleted and for any inconvenience this deletion may have caused. 

  • Oh, I saved the accepted answer- it seems to be that our clock wasn't stable enough

    (though we are still debating if it is better to use a square clock or a sine-wave clock).

    This was the answer:

    "

    Not sure why this was moved here. I'm pretty sure this has nothing to do with software. 

    erw42 said: 

    I see that the sine wave is "breathing" - its maximum amplitude is going up and down. 

    When I draw a histogram of the voltage peak-to-peak of my sine, it is quite wide, 

    with values from 1.04V to 1.047V. (7mV jitter) 

    I don't think you should look at a high speed data converter like this. 

    Ask google what jitter is: "Jitter is the timing variations of a set of signal edges from their ideal values. Jitters in clock signals are typically caused by noise or other disturbances in the system." 

    So jitter is measured in ps or fs. 

    You typically connect the output of the DAC to a spectrum analyzer and identify all the frequency components present. 

    You will find components other than your fundamental tone. The measure how performant your converter is can be found in the datasheet. 

    Take a look at spurious free dynamic range SFDR. SFDR heavily depends on the Phase Noise (related to jitter) of your converter clock. 

    erw42 said: 

    I thought it can be the clock source- is it OK to use a 2GHz sine wave as the source (or does it have to be a square wave?)? 

    So data converters a “Garbage in – garbage out. (GIGO)”  systems. A bad reference clock will generate bad spurs in the DAC output. 

    And everything mixes with each other. So the amplitude variations you see in the time domain, are actually modulation by these other frequency components. 

    Does this explain things? 

    -Michael  

    "

  • Oh, I saved the accepted answer- it seems to be that our clock wasn't stable enough.

    (though we are still debating if it is better to use a square clock or a sine-wave clock).

    This was the answer:

    """

    Not sure why this was moved here. I'm pretty sure this has nothing to do with software. 

    "I see that the sine wave is "breathing" - its maximum amplitude is going up and down. 

    When I draw a histogram of the voltage peak-to-peak of my sine, it is quite wide, 

    with values from 1.04V to 1.047V. (7mV jitter) "

    I don't think you should look at a high speed data converter like this. 

    Ask google what jitter is: "Jitter is the timing variations of a set of signal edges from their ideal values. Jitters in clock signals are typically caused by noise or other disturbances in the system." 

    So jitter is measured in ps or fs. 

    You typically connect the output of the DAC to a spectrum analyzer and identify all the frequency components present. 

    You will find components other than your fundamental tone. The measure how performant your converter is can be found in the datasheet. 

    Take a look at spurious free dynamic range SFDR. SFDR heavily depends on the Phase Noise (related to jitter) of your converter clock. 

    "I thought it can be the clock source- is it OK to use a 2GHz sine wave as the source (or does it have to be a square wave?)? "

    So data converters a “Garbage in – garbage out. (GIGO)”  systems. A bad reference clock will generate bad spurs in the DAC output. 

    And everything mixes with each other. So the amplitude variations you see in the time domain, are actually modulation by these other frequency components. 

    Does this explain things? 

    -Michael  

    """