discuss-gnuradio
[Top][All Lists]
Advanced

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: [Discuss-gnuradio] Reading/processing off-air DVB-T transport stream


From: Ron Economos
Subject: Re: [Discuss-gnuradio] Reading/processing off-air DVB-T transport stream
Date: Tue, 15 Oct 2019 18:46:48 -0700
User-agent: Mozilla/5.0 (X11; Linux x86_64; rv:60.0) Gecko/20100101 Thunderbird/60.8.0

The bandwidth of DVB-T is 8 MHz (or 7 MHz for VHF). You can't capture that much bandwidth with an RTL-SDR. The required sample rate is (8000000.0 * 8) / 7 = 9.14286 Msps for an 8 MHz channel, or (8000000.0 * 7) / 7 = 8 Msps for a 7 MHz channel.

You need an SDR that can handle these sample rates such as bladeRF, hackRF, USRP, etc.

Ron

On 10/15/19 18:11, Tom Crane wrote:
I am attempting to read samples of an off-air DVB-T transport stream into an IQ file for offline analysis but can't get it to work at all.

I can run the dvbt_rx_8k.grc example which reads the adv16.cfile IQ TS file produced by Ron Economos, displays the FFT and 16 QAM constellation and produces a valid .ts file which can be played with a video player..

I adapted that example flow-graph for a local off-air MUX with parameters; 64QAM, 2/3 rate, 1/4 guard interval, 8K FFT.  The FFT displays OK but the constellation graph does not work and show the expected 8 x 8 matrix -- instead it shows just a single dot in the middle.

This https://www.mklab.rhul.ac.uk/~tom/tvdx/gr/dvbt_rx_8k_64QAM_chD44-live.grc cut-down demo attempts to sample channel channel D44 live using an RTL2832/E4000 USB stick.  I have also tried using the rtl_sdr dump-to-IQ-file utility and converting the format from 8-bit uint into 32-bit float (complex) using the sox utility, eg.

rtl_sdr -f 658000000 -s 2800000 -n $[280000*60] chD44.raw
sox -t raw -r 2800000 -b 8 -c 1 -e unsigned-integer chD44.raw -t raw -r 2800000 -c 1 -e float chD44-sox.raw

using a sample rate of 2.8M here, and processing the resultant IQ file with grc but the result is the same.

I also tried with a 3.2M sample rate, the maximum permitted value with RT2832 USBs sticks but again without success.

Am I doing something wrong in grc or it is just that these sample rates are nowhere near sufficient for the whole-MUX type of data I am attempting to capture?

I was able to adapt the dvbt_rx_8k.grc demo for a 64QAM file (adv64.cfile also from Ron Economos) and process that OK in grc.

Many thanks
Tom Crane




reply via email to

[Prev in Thread] Current Thread [Next in Thread]