AnsweredAssumed Answered

DDS AD9914 Evaluation board linear frequency sweep

Question asked by GerardD on May 9, 2018
Latest reply on May 18, 2018 by KennyG



We are using theDDS AD9914 Evaluation Board to generate a chirp. So far, we haven't succeed.

We run several tests and tried different configurations:

- Single tone. Works

- Linear frequency sweep. Works partially (the amplitude looks like it is modulated) with certain parameters. Working configuration in images 1.1, 1.2 and 1.3.

Image 1.1. External clock -> RF oscillator 25 MHz generated by R&S SMA100B. Lock is detected.

Image 1.2. Example 1 of working linear frequency sweep parameters.

Image 1.3. Example 2 of working linear frequency sweep parameters.



When generating a linear frequency sweep with configuration in images 1.4 and 1.5, we observed:

Image 1.4. Example 1 of NOT WORKING linear frequency sweep parameters.

Image 1.5. Example 2 of NOT WORKING linear frequency sweep parameters.


- Initial and final frequency of the sweep are always the expected values.

- The chirp sweep time is not what we expect. The sweep time of the whole chirp is way shorter than it should be. Image 2 shows the spectogram of two chirps. According to the example 1 parameters, the sweep time should be 500 us. Instead, the sweep time ends up being 400 us aprox. Image 3 shows the PSD of a 0.5 us acquisition. Since the chirp ends before expected, the sweep continues and the lower frequencies have higher energy. It is expected from the spectogram results.


Image 2. Spectogram of 1 us acquisition.

Image 3. PSD of a 0.5 us acquisition.



For ALL of the previous configurations (both working and not working linear sweeps), the amplitude is always modulated. Please see image 4.

Image 4. Amplitude analysis. Is it modulated with time? We don't really understand why the amplitude is not constant during the whole sweep.


We don't understand exactly why the sweep works with some configurations and stops working with others. What are we doing wrong?


Thank you so much!