Re: [Discuss-gnuradio] [USRP-users] usrpN200 + TVRX2 50-860 MHz Rx x 2 for DVB-T

[Bogdan Diaconescu]

The snr was determined only by looking at uhd_fft display and making the difference between the plateau of the signal and the noise floor. The maximum level I could obtain moving the antenna in the lab close to the window was between 45dB-50dB and of course it is just a calitative information only as the USRP was not calibrated for precise measurements.

On longer term it is necessary that the program itself will measure the SNR as the OFDM synchronizer uses maximum likelihood algorithm for now which is dependent on SNR. I'm working on an SNR estimator based on the pilots that are sent on the OFDM together with the payload and that will also display the SNR for the end user.

One other SNR measure that is usually employed for characterization of constellation type of signals is MER (Modulation Error Ratio) which should be easily added as a gnuradio block.

Bogdan

 

On Friday, May 9, 2014 12:09 AM, Nasi <address@hidden> wrote:

you said 'SNR was above a certain level'.
How much was it?
How did you measure the SNR?


Thu, 8 May 2014 12:46:17 -0700 (PDT) от Bogdan Diaconescu <address@hidden>:

Hi, I did experiment with real life signals from a local DVB-T transmitter and it worked when the SNR was above a certain level. That was with wbx card.

I also experimented with the same signals using a Philips TV room antenna containing integrated RF amplifier and it actually worsen the SNR than just VERT900 antenna. When using the highest gain on the antenna the signal was totally buried in noise. In the end the best antenna was the small stock antenna that came with rtl2832 stick.

Bogdan

On Thursday, May 8, 2014 10:25 PM, Nasi <address@hidden> wrote:

Hi Bogdan,

Did you try real TV transmission? Off air reception?


Thu, 8 May 2014 12:22:37 -0700 (PDT) от Bogdan Diaconescu <address@hidden>:

Hi,

I did not try N200 but I do not think that you'll be able to use sampling rate of 9.14msps (which is more precisely 64/7) with N2xx. The sample rate does not depend on the dautherboard you are using but on the motherboard.

Bogdan

On Thursday, May 8, 2014 10:08 PM, Nasi via USRP-users <address@hidden> wrote:

Hi,


Thu, 08 May 2014 20:18:56 +0400 от Nasi <address@hidden>:

HI all,

I want to collect TV data using usrpN200 + TVRX2 50-860 MHz Rx x 2 for DVB-T .

Here are my questions:

- Is it possible to receive directly with a sampling rate of 9.14 Mps for DVB-T? Or should I resample it from 10 Mps? I was resampling with daughterboard XRCV 2450 since it cannot produce 9.14 Mps.

- Is it possible at all to get off air TV boadcasting using above equipment?


--
NE

_______________________________________________
USRP-users mailing list
address@hidden
http://lists.ettus.com/mailman/listinfo/usrp-users_lists.ettus.com

--
NE

--
NE