Simple QAM project

GNU Radio
1

I am trying to get data through a QAM-16 Modulator and just display the
data stream on a Constellation Sink. I have 8khz sampled 8-bit
data(char) going into the QAM Mod block. I have the output of the QAMmod
block going directly into the Const Sink. I am not seeing anything on
the Constellation Sink. I put a throttle block in but that didn’t help.
Any thoughts on what could be wrong? I have followed the GNUradio
tutorial: Part 4 by balint256 but didn’t help.
Thanks…DT

2

Can u attach ur GRC file?

-Adeel

3

Attached grc file

4

Tom,

There are few changes required in ur FG
1: PSK/QAM Modulator takes packed byte input (8 bits per byte) instead
of
1 bit per byte
see gnuradio>digital>generic_mod_demod.py
2: Constellation sink work only for PSK modulation M=2,4,8.

To see output constellation of QAM-16 u will have to first downsample it
to
1 sample_per_symbol and then use wx-scope-sink in XY-Mode
Down-sampling can be done using symbol-timing blocks or manually
(delay+keep1inN combination) as in attached FG

-Adeel

5

I am new to linux and GNURadio. I don’g follow where the 1 bit/byte is
coming from. Does the audio source not deliver 8bit/sample or how do you
figure that out? I don’t see anything such as
"gnuradio>digital>generic_mod_demod.py " in my directories. Thanks for
the help.
Tom

Tom,

There are few changes required in ur FG
1: PSK/QAM Modulator takes packed byte input (8 bits per byte) instead
of 1 bit per byte
see gnuradio>digital>generic_mod_demod.py
2: Constellation sink work only for PSK modulation M=2,4,8.

To see output constellation of QAM-16 u will have to first downsample it
to 1 sample_per_symbol and then use wx-scope-sink in XY-Mode

Down-sampling can be done using symbol-timing blocks or manually
(delay+keep1inN combination) as in attached FG

-Adeel

6

On Sun, Jul 7, 2013 at 5:50 PM, tom sutherland [email protected]
wrote:

I am new to linux and GNURadio. I don’g follow where the 1 bit/byte is
coming from. Does the audio source not deliver 8bit/sample or how do you
figure that out? I don’t see anything such as
"gnuradio>digital>generic_mod_demod.py " in my directories. Thanks for the
help.
Tom

Tom,
The audio source will produce single-precision floating-point samples
between -1 and 1. You’ll need to think about how you want to convert
these to digital samples that the QAM modulation can handle. So first
convert it into some quantized representation (i.e., as fixed point).
Then, you should have a stream of ‘packet’ bits that can be directly
inputted into the modulator.

There are blocks called packed_to_unpacked / unpacked_to_packed as
well as a unpack_k_bits / pack_k_bits that you can use to go from
unpacked to packed representation.

Tom