Post Go back to editing

When the ad9739 outputs a DC voltage, the low-frequency noise will change with the DC voltage.

Thread Summary

The user observed that the AD9739's low-frequency noise increases with the DC output voltage. The final answer explains that this is expected due to the AM noise from the master bias generator, which modulates the DAC's full-scale current. The noise is most noticeable at full-scale outputs due to the lack of common-mode rejection. The accompanying answer suggests measuring before the FDA in the schematic for further insight.
AI Generated Content
Category: Hardware
Product Number: AD9739

Hi:

When the ad9739 outputs a DC voltage, the low-frequency noise will change with the DC voltage. 

As shown in the figure below:(The voltage output value has the following three situations: DC = 0V,DC = 0.2V, DC = 0.45V ,The output voltage noise will increase as the output voltage value increases。)

AD9739 output termination circuit (DC Coupling) is shown in the figure below:

Is this result consistent with the AD9739 specification? What may be the cause? Thank you

  • Hi  ,

    Thank you for your interest in AD9739. We'll check this and get back to you as soon as possible. Just a question, have you tried measuring before the FDA in your schematic? 

    Best regards,
    Marco

  • Hello,

    This observation is to be expected since the DAC's low frequency noise is dominated by the master bias generator for the weighted current source array that is central to the DAC architecture.   Note this noise is AM in nature since it is modulating the DAC's full-scale current as compared to PM noise that is due to jitter on the clock signal.   When the DAC code is at mid-scale..........the current is equally divided between the IOUT outputs hence one gets the benefit of common-mode cancellation by an interface circuit (i.e. balun/transformer or diff-to-single-ended amplifier) that has decent common-mode rejection.  When the DAC is cost is at + or - full-scale , all of the DAC's output current is directed to one or the other IOUT outputs hence no rejection of this AM  noise source