AD9249 aperture delay variation

Hi
In AD9249 datasheet I can find the following aperture related specs:
- Aperture Delay
- Aperture Uncertainty (Jitter)

But I can't find anywhere  "Aperture mismatch within device" or "Aperture delay variation Channel-to-channel within the same device". Is it available somewhere?

Regards
Konstantin

  • 0
    •  Analog Employees 
    on Dec 7, 2016 10:33 AM

    Hi Goodsoul,

    Thank you for your interest in AD9249.

    As you have observed, aperture delay variation/matching within one device is not in the datasheet. Unfortunately I do not have this information.

    One datapoint we have is that I believe that the maximum aperture delay variation across many parts is about +/-100ps.

    This is from silicon process variation alone. On a given board, assuming the temperature and supply voltage would be very close to the same for all the ADCs, the main variable part-to-part would be silicon process.

    Please note that these are not guaranteed limits, but guidance on the variation from simulation run at the fast and slow process corners.

    If this magnitude of variation is acceptable for your application, then you will be OK because this magnitude is worse than for across one part.

    What level of variation can you tolerate?

    Thank you.

    Doug

  • Hi Doug
    Thank you for the response.

    The task requires 20ps channel mismatch.

    In datasheet I can see that AD9249 have Aperture Uncertainty (Jitter) - 0.135ps on each channel

    Whether it will be correct to assume, that(for example)

    - Channel x - have 900 ps aperture delay. And real acquire time will be from 899.865ps to 900.135ps.

    - Channel y - have 1100 ps aperture delay. And real acquire time will be from 1099.865ps to 1100.135ps

    Is it correct assumption?

    Regards
    Konstantin

  • 0
    •  Analog Employees 
    on Dec 8, 2016 1:00 AM

    Hi Konstantin,

    In general I agree with your assessment with a small clarification. As you have stated, the Aperture Uncertainty (jitter) is 135fs RMS. This, being a random quantity, is described statistically. The RMS value is equivalent to one standard deviation of the jitter variation. In your analysis you are adding (Aperture Delay) + (Aperture Delay variation) + (one standard deviation of jitter). In reality you would see more than one standard deviation of variation in jitter.

    Practically speaking this makes no difference because the variation from jitter is very small compared to the overall static delay. I just wanted to explain about the RMS jitter term.

    Thank you.

    Doug