# Sampling with TVRX daughterboard and Nyquist frequency

**URL:** <https://www.ruby-forum.com/t/sampling-with-tvrx-daughterboard-and-nyquist-frequency/214842>\
**Category:** GNU Radio\
**Created:** [December 13, 2011, 5:45pm UTC](https://www.ruby-forum.com/t/sampling-with-tvrx-daughterboard-and-nyquist-frequency/214842 "2011-12-13T17:45:20Z")\
**Posts on this page:** 3\
**Page:** 1

<div class="post-metadata">

**Author:** ![dubstep](https://www.ruby-forum.com/letter_avatar_proxy/v4/letter/d/cdc98d/32.png) [@dubstep](https://www.ruby-forum.com/u/dubstep)\
**Post date:** [December 13, 2011, 5:45pm UTC](https://www.ruby-forum.com/t/sampling-with-tvrx-daughterboard-and-nyquist-frequency/214842/1 "2011-12-13T17:45:20Z")

</div>

Hi all,

I use a TVRX daughterboard to acquire VHF signals. I believe (but I may  
be  
mistaken) that the TVRX daughterboard downconverts RF signals to a  
baseband  
of 6MHz bandwidth before going to the ADC. The function usrp\_rx\_cfile  
(which I use to acquire signals) has a minimum decimation factor of 8  
for a  
sampling rate of 64Ms/s. So it gives a maximum sampling rate of 8MHz.

Nyquist theorem tells us that to acquire properly a signal occupying a  
band  
between 0 and 6MHz, the sampling rate shall be at least 12MHz, which is  
then unreachable by the TVRX daughterboard.

So, are my figures wrong, or am I missing something ?

Thanks for your lights,

Cheers  
Cyril

---

<div class="post-metadata">

**Author:** ![cguirguis](https://www.ruby-forum.com/letter_avatar_proxy/v4/letter/c/f08c70/32.png) [@cguirguis](https://www.ruby-forum.com/u/cguirguis)\
**Post date:** [December 13, 2011, 5:53pm UTC](https://www.ruby-forum.com/t/sampling-with-tvrx-daughterboard-and-nyquist-frequency/214842/2 "2011-12-13T17:53:22Z")

</div>

On Tue, 13 Dec 2011 17:44:14 +0100, Cyril C. wrote:

> Hi all,

> I use a TVRX daughterboard to acquire VHF signals. I believe (but  
> I may be mistaken) that the TVRX daughterboard downconverts RF signals  
> to a baseband of 6MHz bandwidth before going to the ADC. The function  
> usrp\_rx\_cfile (which I use to acquire signals) has a minimum decimation  
> factor of 8 for a sampling rate of 64Ms/s. So it gives a maximum  
> sampling rate of 8MHz.
> 
> Nyquist theorem tells us that to acquire  
> properly a signal occupying a band between 0 and 6MHz, the sampling rate  
> shall be at least 12MHz, which is then unreachable by the TVRX  
> daughterboard.
> 
> So, are my figures wrong, or am I missing something  
> ?
> 
> Thanks for your lights,
> 
> Cheers  
> Cyril

The underlying  
motherboard samples at a fixed rate–64Msps for USRP1/B100/E1XX, and the  
N2XX family samples at 100Msps, so Nyquist isn’t violated at all. The  
FPGA downconverts that sampled 6Mhz to baseband, which is then run  
through CIC decimators and half-band filters. In general, a Gnu Radio  
flow-graph deals with complex-baseband signals, which “straddle” 0Hz  
(that is, for a signal of B bandwidth, it stretches from -B/2 Hz to B/2  
Hz).

---

<div class="post-metadata">

**Author:** ![cguirguis](https://www.ruby-forum.com/letter_avatar_proxy/v4/letter/c/f08c70/32.png) [@cguirguis](https://www.ruby-forum.com/u/cguirguis)\
**Post date:** [December 13, 2011, 7:14pm UTC](https://www.ruby-forum.com/t/sampling-with-tvrx-daughterboard-and-nyquist-frequency/214842/3 "2011-12-13T19:14:33Z")

</div>

On Tue, Dec 13, 2011 at 8:44 AM, Cyril C. [removed\_email\_address@domain.invalid](mailto:removed_email_address@domain.invalid)  
wrote:

> Hi all,
> 
> I use a TVRX daughterboard to acquire VHF signals. I believe (but I may be  
> mistaken) that the TVRX daughterboard downconverts RF signals to a baseband  
> of 6MHz bandwidth before going to the ADC. The function usrp\_rx\_cfile  
> (which I use to acquire signals) has a minimum decimation factor of 8 for a  
> sampling rate of 64Ms/s. So it gives a maximum sampling rate of 8MHz.

> Nyquist theorem tells us that to acquire properly a signal occupying a  
> band between 0 and 6MHz, the sampling rate shall be at least 12MHz, which  
> is then unreachable by the TVRX daughterboard.
> 
> So, are my figures wrong, or am I missing something ?

Nyquist says that you need 2 samples per second for every 1 Hz of  
bandwidth  
you want to sample. However, we use complex (I and Q) samples, which  
count  
double. So if you want to sample 6 MHz of BW, you would need 6 MS/s  
complex or 12 MS/s real. You are best off oversampling by about 25%, so  
the 8 MS/s complex of a USRP1 or B100 device can give you the full  
bandwidth.

Matt