improvements

I am not satisfied with the results. The sweep signal can be smooth when a lowpass filter is used with a low cutoff frequency. Unfortunately in that case the output is too low. Amplifying it in software works but it is not enough.

I will increase the steps the DAC generates from 10 to 20. This increases the frequency difference between the sample frequency and the sweep frequency. This relaxes the filter requirements. I can even increase the steps to 40 if the suppression is still not adeqate. This way I can have good suppession of the ripple without suppressing the amplitude of the sweep signal itself.




The price I have to pay is loss of the lowest span. Each dac step is 500 KHz. So when 20 steps per sweep are used, the minumum span is 20*500 KHz =10 MHz. This is acceptable as the bandwidth of the IF filter is 1 MHz -60 db. Even 20 MHz as minimum span would be acceptable.