0
\$\begingroup\$

I am doing a basic amplitude modulator, and I'm using as input signal for now a Gaussian noise generator, then I modulate it with a carrier fc=2e9Hz. When using Simulink along with a RF bandpass filter, I got an error saying that there is missmatch between the filter sampling time (Badnpass filter Ts = 2.5e-9) and the input signal the noise I guess (Ts= 2.5e-11), as it can be seen in the picture attached.

block config

I have tried as well to set the same sampling time in the gaussian generator and in the filter, but then I got another error saying that the sampling time in the filter is to low compared to the center frequency... Which I don't know what it does mean, since if I set it lower, then I would have aliasing issues.

May someone enlighten me a bit?

Thanks in advance!

Regards

\$\endgroup\$
  • \$\begingroup\$ Did you try setting the sample times the same? \$\endgroup\$ – Andy aka Mar 31 '15 at 9:47
  • \$\begingroup\$ I have tried as well to set the same sampling time in the gaussian generator and in the filter, but then I got another error saying that the sampling time in the filter is to low compared to the center frequency \$\endgroup\$ – ndarkness Mar 31 '15 at 12:44
1
\$\begingroup\$

if I set it lower, then I would have aliasing issues.

You said you have a carrier of 2 GHz and a sampling period of 2.5 ns.

You already have aliasing issues.

You need a sampling period of less than 0.25 ns to avoid aliasing.

It may be that you are deliberately undersampling in order to use the sampling process to down-mix your signal. That is a perfectly acceptable technique, although I'm not familiar enough with Simulaink to know if there's any isses with simulating it.

On top of that, if a "Gaussian Noise Generator" is in fact a white gaussian noise generator, then you have an infinite signal bandwidth. (I'm not clear if you are using this as a stand-in for a message signal or to model channel noise)

Try low-pass filtering the noise source down to a small fraction of the carrier frequency before using it to modulate the carrier. If you keep your 400 MSa/s sampling rate, you need to keep the signal bandwidth below 200 MHz to avoid aliasing.

\$\endgroup\$
  • \$\begingroup\$ Hello The Photon, I know that I am having aliasing if I use that sampling period, the thing is that if I change it to 2.5e-11,for instance, I got another error in the same block saying "saying that the sampling time in the filter is to low compared to the center frequency", and for now you are right I don't want to undersample to down-mix my signal. \$\endgroup\$ – ndarkness Apr 1 '15 at 7:05
  • \$\begingroup\$ Concerning to the the Gaussian Noise generator, we could keep it as I am using it as a message signal. Then, do you suggest me filtering the source with a low-pass filter and a bandwidth of 200MHz before modulating the carrier? Could you explain me a bit more this idea? \$\endgroup\$ – ndarkness Apr 1 '15 at 7:09
  • \$\begingroup\$ Main point is, white noise has infinite bandwidth. So if you try to use it in place of a message signal, it's not suprising the tool thinks the sampling rate is not sufficient to measure it. \$\endgroup\$ – The Photon Apr 1 '15 at 16:10

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.