Post Go back to editing

AD9746 + SDP-H1: temperature performances & bandwidth

Hello,

I am currently using AD9746 eval board and SDP-H1 for measuring performances of AD9746 in the following conditions:

     - Output signal on IOUT1P: level of -1 dBFS and frequency from DC to 40 MHz

     - IOUT1N is ground for common mode measurement.

     - External clock frequency injected in J1: -2 dBm / 100 MHz

Context:

I have to perform several measurements:

     - Bandwidth from DC to 40 MHz at -30, +25 and +50 °C

     - SFDR, Harmonics, NSD, output power (among others) at above temperatures at the following output frequencies: 10 kHz, 10 MHz and 40 MHz

My issues:

     - Issue 1: for temperatures from 0°C down to -30°C, output power starts to decrease and spurious ans noise level increase. 

Question 1:  Does AD9746 + SDP-H1 can be used for tests in temperature?

     - Issue 2: DAC output power starts to decrease below few MHz. For instance, I get an attenuation around 35 dB by comparing output signal generated at 10 MHz compare to signal generated at 10 kHz.

Note: the balun (T5 and T7) of AD9746 eval board have been shorted. Additionally I have also tested to tie R34 and R36 to AVDD33 as proposed within AD9746 datasheet.

Question 2: what is DAC output bandwidth?

Thanks a lot for your help. Do not hesitate if you need more information.

Regards,

Pierre

Parents
  • Hi Pierre,

    We usually recommend the DAC output frequency to be ≤ (DAC clock / 4) for minimal performance degradation.

    How does your setup for tests across temperature look like? Ideally, we want to subject only the chip, not the whole evaluation system, to your desired ambient temperature.

    Best,
    Shine

  • Hi Shine,

    Concerning your first answer, actually we have consistent results for DAC frequencies at 10 and 40 MHz. My issue concerns frequencies below 1 MHz, and in this case we are far below DAC clock / 4 !!

              => Do you have an explanation?

     Concerning the test setup in temperature, indeed, we place both AD9746 eval board and SDP-H1 in a laboratory oven.  

              => Do you confirm that these product cannot work within the range -30 to +50 °C?

    Thanks for your help,

    Pierre

  • Hi Pierre,

    We recommend DAC output frequency to be less than 4 times the sampling frequency because of the inherent sinx/x roll off in DACs. You may refer to this app note AN-928.pdf. The easily observable effect of this is the decreasing amplitude of the fundamental tone as it approaches the Nyquist frequency. Since you are operating well below  fs/4, you should be okay.

    The problem might be your measurement setup because the AD9746 eval board and the SDP-H1 were not characterized in the temperatures above. They are designed to be used at room temperature only. By subjecting both boards in different temperatures, instead of characterizing the DAC's performance only, your data is affected by all the components and the boards' response to temperature. A better test setup would be to control the temperature around a small area surrounding the DAC using a thermostreamer. This is why a different board is designed for temperature characterization, so every other components can be placed away from the device under test.

    Best,
    Shine

  • Hi Shine,

    Thanks again for your answer.

    As far as I understand, the eval board and SDP-H1 controller is not consistent with test in temperature (from -30 °C to +50 °C) in our test cases.  

    But concerning DAC output frequencies range, I still don't explain why the compoenent acts like a high pass filter when we decrease DAC output frequency, down to 10 kHz.

    Do you have an explanation?

    In addition, within DAC datasheet, I saw that thers is no measurements below 10 MHz, do you know why? Do you confirm that DAC output frequency can be set down to 10 kHz or DC without, theoretically, any loss compare to a signal generated at 10 MHz?

    Thanks again for your help,

    Regards,

    Pierre

  • Hi Pierre,

    Sorry for the late response.
    Is the output attenuation consistent at different frequencies? What are the results at room temperature?

    If you are using the evaluation board and the DAC outputs are connected to the RF transformers, the output bandwidth is limited by the bandwidth of the transformers. Try connecting DAC outputs to RF amplifiers and see if the results are the same.

    Regards,
    Shine

  • Hi Shine,

    Sorry also for my late answer Slight smile

    The more the frequency is low, the more there is attenuation increases. This phenomenon is observable at ambiant temperature as well as by increasing/decreasing the temperature.

    We have tried to bypass the balun for these measurements without any change. To be sure that balun have no effects on signal attenuation we are removing the balun. I will let you know about the results as soon as we will have tested this new configuration.

    Regards,

    Pierre

  • In addition to my last post, do you have an answer to the following question:

    In addition, within DAC datasheet, I saw that thers is no measurements below 10 MHz, do you know why? Do you confirm that DAC output frequency can be set down to 10 kHz or DC without, theoretically, any loss compare to a signal generated at 10 MHz?

    Thanks and regards,

    Pierre

  • Hi Pierre,

    There is no minimum output frequency but the shape of the output spectrum, hence the amplitude of the output tone will depend on the DAC Mode. Please see Figure 35 and Table 14 in the datasheet.





    Regards,
    Shine

Reply Children