RFSG - Generate Dual Tone Modulation

[Regulator]

PXI:  5670 RF Vector Signal Generator 2.7GHz (5621 ARB and 5610 Upconverter)

        5421 ARB

My goal is to generate a dual-tone modulated signal with 90Hz and 150Hz modulating 300MHz carrier.  This was built-up from the examples to include the two tones, but I couldn't get the second tone to be at the frequency I told it to (it was only a multiple of the first frequency-see images).  By increasing the number of samples by 3 before generating the sinusoids I think I got it to work, but I can't explain why I was getting so many spurs, nor why the "coerce frequency and samples subVI" (the one right after the property node) doesn't take care of this.  Why do I have to hard code this?  Will this work?  Any information would be great!

[JoshY]

Hi Regulator,

How did you change the actual Number of Message Samples output from FindWaveformParamsForFrequency.vi? The only differences I see between the front panels are the Modulation Index and the Minimum Number of Samples, and neither of these should directly affect that output. However, I believe this is the reason why- a larger number of samples will create a better waveform. Also, the modulation index is reduced from 1 to .3.

Regards

Josh Y

[Regulator]

I do see that the max and min samples are different in my comparison, but these changes did not change the output of the FindWaveformParametersForFrequency.vi.  So we agree that this shouldn't have changed anything.   

In the first set of images I show what the FindWaveformParametersForFrequency.vi came up with after taking in the IQ rate (100MS/s), Quantum (4) {which is read only}, and the sample range and I didn't get what I expected.  The "# of samples" seemed to be short, which may be why I had the distortion (Discontinuity maybe?).

The change that I made in the second set of images is shown in the block diagram at the output of the lower FindWaveformParametersForFrequency.vi, where i took the "# of samples output" and multiplied it by 3 to get more samples of the FGEN. 

From your reply it seems okay to be putting in 3 times the samples of the message signal into the RFSG to modulate the carrier.  Is that what you were saying?

[JoshY · Accepted solution]

Hi Regulator,

I believe this is a sampling issue. You take the number of samples from the 90 Hz signal rather than the 150 Hz signal. The 150 Hz signal needs more samples in order to accurately represent it, so this number of samples passed from the 90 Hz will not suffice. When you multiply the number of samples by 3, you are simply oversampling. This will be a bit more taxing when doing the modulation, but it is okay as long as your machine can handle it.

Regards