I have a few questions about EasyWave and the SDG2042X.
I'm trying to generate the following signal.
f = cos(2pi*3.9e7) + cos(2pi*4e7)
Could you describe where the spurious signals appear? Are they IM3/IM5? Or are they at some other frequency?
I don't have a spectrum analyzer, but I would suggest trying the following:
1) Use the internal CH1+CH2 mode, and check how clean the output is. By this, I mean set CH1 to 39 MHz and CH2 to 40 MHz. In the Utility->Output Setup->Wave Combine, set CH1 := CH1 + CH2
Is the result clean? If not, it's likely that the analog output amplifier. If they are, then you should be able to get the output displaying properly.
If it actually is able to cleanly output the sum, then you need to tune your ARB better. The goal is to have an integer multiple of both 39 MHz and 40 MHz periods in the waveform. Firstly, you should decide on the sample rate. Unfortunately, we are limited to 300 MSa/s. It would have been nice to choose a sample rate which is an integer multiple of both fundamental frequencies an use TrueARB mode.
So, we must use DDS mode. I would suggest using Matlab to generate the ARB sequence, and output it in the CSV format that the generator can load. First, choose some number of samples per period of the 40 MHz sine wave (48 is probably a good number). Create a list of the samples of a sine wave with that many samples per period, and the list for the 39 MHz at the same sampling rate.
I wrote a quick Matlab script, which I'm attaching, though I don't know if it'll produce anything better that what you've been able to do. Note that for this particular pair of frequencies, you should play the ARB at 1 MHz in DDS mode.