Re: [Discuss-gnuradio] ssb - question

[Jeff Long]

Unless you're building the IQ modulator with real signals for 
educational purposes, go with the flowgraph Ron mentioned. 
Multiplication by a complex sine is much easier. You can also use a 
Rotator or Frequency Xlating Filter to do similar things. The bandpass 
filter can be built in to the Xlating Filter.

On the additional peaks, the second highest one is about 50 dB lower 
than the desired one. From a real hardware point of view, it might as 
well not be there (same as 1mW vs 100W). If you use a bandpass filter, 
you can make it arbitrarily small, but filtering should be done based on 
requirements, not just to make a plot look nice.

- Jeff

On 09/08/2015 01:30 PM, Przemek Lewandowski wrote:
> Thank You Jeff for answer.
>
> I have fixed Sampling Rate (of course you were right with it) and
> realized that Hilbert Taps was wrong. When Sampling Rate was default
> 32k, default taps for Hilbert Transform is 64.
> So for 48k Sampling Rate for Hilbert Transform should be 96 ?! (Im not
> sure about what im writing, but this configuration works as you can see
> in image 1 - upper plot)
> image 1 - http://postimg.org/image/jme38k5hx/
> image 2 - fixed and simplified flowchart -
> http://postimg.org/image/ukpcqqu39/
>
> And after all that, I can see there a lot more peaks. So should I filter
> them ?
> What with that second peak near LSB Peak ?!
>
> Greatings, thank you in advice
> Przemek Lewandowski
>
> 2015-09-08 18:32 GMT+02:00 Jeff Long <address@hidden
> <mailto:address@hidden>>:
>
>     Your first problem is that you're using a 32k sample rate with a 20k
>     sin/cos. This will actually come out as a 12k sin/cos, due to
>     aliasing. Try the same thing again with a 48k sample rate and see
>     what happens.
>
>     Either bandpass filtering or a Hilbert transform can be used to
>     create SSB.
>
>     Not sure what your flowgraph does, but it looks fun.
>
>     Jeff.
>
>
>     On 09/08/2015 08:48 AM, Przemek Lewandowski wrote:
>
>         Hi everybody
>
>         Im trying to make SSB modulation with Hilbert transform. And Im
>         using IQ
>         modulator
>         here is schema:
>         http://postimg.org/image/gms5srret/
>
>         And I have a question. In Frequency domain there should be one
>         peak, and
>         on on my plot there is one - that is good, and second that is
>         smaller/suppressed.
>
>         Is something wrong ?? or is it normal ??
>
>         Is Hilbert method for creating SSB signal is good ?? or can it
>         be made
>         easier ???
>
>
>         thank you very much.
>         Przemek Lewandowski
>
>
>         _______________________________________________
>         Discuss-gnuradio mailing list
>         address@hidden <mailto:address@hidden>
>         https://lists.gnu.org/mailman/listinfo/discuss-gnuradio
>
>
>
>     _______________________________________________
>     Discuss-gnuradio mailing list
>     address@hidden <mailto:address@hidden>
>     https://lists.gnu.org/mailman/listinfo/discuss-gnuradio