AD9746/9783: DAC analog signal peak power attenuation issues on AD9746/9783 Evaluation boards.

Dear Experts,

My customer is evaluation AD9746 and AD9783 with those evaluation boards and found some of the following technical issues with AD9746/9783 evaluation boards. Could you please review the following descriptions and then reply to the following questions?

1. Issue overview:
1) AD9746 Evaluation board:
My customer wants to make AD9746 output the analog signal peak power at 0[dBm] but cannot do so. For the details, please see below and the attached file.

2) AD9783 Evaluation board:
My customer can make AD9783 output the analog signal peak power at 0[dBm]. However, the DAC code for the 0[dBm] analog signal power in AD9783 was different between the calculation and the measurement. For the details, please see below and the attached file.

I made the attached pdf tile. At first, you should review the attached document and then return to the following questions and so on.


2. What configuration my customer did set and what my customer saw:
2.1 AD9746/9783 Evaluation environment:
Please refer to the environment block diagram on the p.3 of the attached document.


2.2 AD9746 Evaluation board:
1-1) The configurations: pp.5-7 of the attached document.
1-2) The spectrum measurement of the above configurations: p.8 of the attached document.
Note: When "Full-Scale Adjustment" in the AD9746 was set to 0x1F9, the analog signal pear power of the AD9746 was -13.86[dBm].

2-1) The configurations: The additional configuration from the pp.5-7 of the attached document was to set "Full-Scale Adjustment" in AD9746 to 0x3FF. Please see the p.9 of the attached document.
2-2) The spectrum measurement of the above configurations: p.10 of the attached document.
Note: When "Full-Scale Adjustment" in the AD9746 was set to 0x3FF, the analog signal pear power of the AD9746 was -9.86[dBm].

3-1) The configurations: The additional configuration from the pp.5-7 & 9 of the attached document was to set "Amplitude" in AD9746 to -20[dB] (Full Scale)..Please see the p.11 of the attached document.
3-2) The spectrum measurement of the above configurations: p.12 of the attached document.
Note: When both "Full-Scale Adjustment" was set to 0x3FF and "Amplitude" was set to -20dB (Full Scale) in the AD9746, the analog signal pear power of the AD9746 was -29.85[dBm].


2.3 AD9783 Evaluation board:
1-1) The configurations: pp.14-15 of the attached document as the defaults
2-2) The spectrum measurement of the above configurations: p.16 of the attached document.
Note: When the DAC code in the AD9783 was set to 505 (dec), the analog signal pear power of the AD9783 was -1.70[dBm].

2-1) The configurations: The additional configuration from the pp.14-15 of the attached document was to set the DAC code to 1023 (dec). Please see the p.17 of the attached document.
2-2) The spectrum measurement of the above configurations: p.18 of the attached document.
Note: When the DAC code in the AD9783 was set to 1023, the analog signal pear power of the AD9783 was 2.33[dBm].

3-1) The configurations: The additional configuration from the pp.14-15 & 17 of the attached document was to set the DAC code to 697 (dec).
3-2) The spectrum measurement of the above configurations: p.19 of the attached document.
Note: When the DAC code in the AD9783 was set to 697 (dec), the analog signal pear power of the AD9783 was 0.00[dBm].

4-1) The configurations: The additional configuration from the pp.14-15 & 17 of the attached document was both to set the DAC code to 697 (dec) and to set "Amplitude" to -20dB (Full Scale) in AD9783.
4-2) The spectrum measurement of the above configurations: p.20 of the attached document.
Note: When the DAC code in the AD9783 was set to 1023, the analog signal pear power of the AD9783 was -20.00[dBm].


3. Questions of my customer:
Could you please reply to the following questions.


Q1: Even if “Full-Scale Adjustment” was set to 0x3FF in AD9746 Evaluation board, the analog signal peak power of the AD9746 was -9.86[dBm].

Is it reasonable? If it is not reasonable, why is it?

Note: Please refer to the p.10 of the attached document. My customer expected that the analog signal peak power of the AD9746 would be about 0[dBm] with the "Full-Scale Adjustment" at 0x3FF.


Q2: What should my customer do with AD9746 Evaluation board in order to make the AD9746 output the DAC analog signal peak power at 0[dBm]?

Q3: My customer calculated the DAC code for the analog signal peak power of AD9783 at 0[dBm]. The DAC code which my customer calculated in AD9783 was 512 (dec) for the analog signal peak power of the AD9783 at 0[dBm]. However, my customer measured the analog signal peak power of the AD9783 at 0[dBm] when the DAC code in the AD9783 was set to 697 (dec).

Why is the difference of the DAC codes of the AD9783 between the calculation and the measurement for the analog signal peak power of AD9783 at 0[dB]m]?


Thanks and regards.

AD9746_Signal_peak_power_attenuation_issues_R001_20170306.pdf
Parents
  • Dear larrywelchusa-san,

    Thank you so much for your kind suggestion.

    I am so sorry for my much late response due to my customer's late feedback.

    When my customer did set the bit resolution for AD9746 to 16 bits in DPGDownloader PC software tool, he saw the analog signal peak power at about 0[dBm] as desired. When "Full-Scale Adjustment" was set to 0x2B4 with 16-bit resolution in AD9746 Evaluation board, the analog signal peak power was about 0[dBm]. It was okay. But, he still has the following of some questions and concerns.

    Could you please reply to the following questions?

    Q4: However, my customer's calculation on the analog signal peak power is still different from the measured one.

    When "Full-Scale Adjustment" was set to 512(dec) in AD9746 Evaluation board, the analog signal peak power should be 0[dBm]. It was calculated by my customer and understanding of my customer. However, the analog signal peak power with "Full-Scale Adjustment" at 512(dec) measured by him was not 0[dBm].

    As described in the above, when "Full-Scale Adjustment" was set to 0x2B4 with 16-bit resolution in the AD9746 Evaluation board, the analog signal peak power in the AD9746 Evaluation board was about 0[dBm].

    Why was there some difference between the calculated analog signal peak power and the measured one?

    Could you please advise my customer about how to calculate the analog signal peak power for the "Full-Scale Adjustment" value set in the DPG Downloader PC tool?


    Q5: According to the following description of "AD9741/3/5/6/7 Evaluation Board Quick Start Guide", "DAC Resolution" set in the DPGDownloader software should be set to the data width of the DUT as AD9746. However, you requested us to set the DAC resolution to 16 bits different from the AD9746 with 14 bits.

    p.4 of AD9741/3/5/6/7 Evaluation Board Quick Start Guide:
    "The evaluation boards for the AD9741/AD9743/AD9745/AD9746/AD9747 are identical except for the DUT installed. The operation of the eval boards is also the same except for the vector generation "DAC Resolution" field and the Data Width fields as shown below. These fields should be configured to match the DUT."

    We know that your request is just an experiment. Could you please update the DPGDownloader PC software tool or advise my customer about any way to meet the above requirement?


    Thanks and regards.

Reply
  • Dear larrywelchusa-san,

    Thank you so much for your kind suggestion.

    I am so sorry for my much late response due to my customer's late feedback.

    When my customer did set the bit resolution for AD9746 to 16 bits in DPGDownloader PC software tool, he saw the analog signal peak power at about 0[dBm] as desired. When "Full-Scale Adjustment" was set to 0x2B4 with 16-bit resolution in AD9746 Evaluation board, the analog signal peak power was about 0[dBm]. It was okay. But, he still has the following of some questions and concerns.

    Could you please reply to the following questions?

    Q4: However, my customer's calculation on the analog signal peak power is still different from the measured one.

    When "Full-Scale Adjustment" was set to 512(dec) in AD9746 Evaluation board, the analog signal peak power should be 0[dBm]. It was calculated by my customer and understanding of my customer. However, the analog signal peak power with "Full-Scale Adjustment" at 512(dec) measured by him was not 0[dBm].

    As described in the above, when "Full-Scale Adjustment" was set to 0x2B4 with 16-bit resolution in the AD9746 Evaluation board, the analog signal peak power in the AD9746 Evaluation board was about 0[dBm].

    Why was there some difference between the calculated analog signal peak power and the measured one?

    Could you please advise my customer about how to calculate the analog signal peak power for the "Full-Scale Adjustment" value set in the DPG Downloader PC tool?


    Q5: According to the following description of "AD9741/3/5/6/7 Evaluation Board Quick Start Guide", "DAC Resolution" set in the DPGDownloader software should be set to the data width of the DUT as AD9746. However, you requested us to set the DAC resolution to 16 bits different from the AD9746 with 14 bits.

    p.4 of AD9741/3/5/6/7 Evaluation Board Quick Start Guide:
    "The evaluation boards for the AD9741/AD9743/AD9745/AD9746/AD9747 are identical except for the DUT installed. The operation of the eval boards is also the same except for the vector generation "DAC Resolution" field and the Data Width fields as shown below. These fields should be configured to match the DUT."

    We know that your request is just an experiment. Could you please update the DPGDownloader PC software tool or advise my customer about any way to meet the above requirement?


    Thanks and regards.

Children
No Data