I've been messing about with a function which exports the tone
parameters and sends them to a UDP program which then programs a DDS.
Not for LF/VLF, this is for a multiplied transmit chain on SHF. The GFSK
transitions in FT8 etc are added later in wsjt-x but for this lash-up, I
took the values from itone and nsym between genft8_() and gen_ft8wave_()
in widgets\mainwindows.cpp and sent then using QUdpSocket
writeDatagram() to a simple receiving server on another device, which
generates the timed sequence by sending commands to a DDS board. I
haven't tried it with QRA64 or JT4 yet.
That approach avoids having to mix anything at signal frequency on 47 or
122 GHz, and permits arbitrary multiplication factors, but you might be
able to use a similar technique to program a DDS at baseband.
A different, more generic approach has been tried by Andy G4JNT. He
takes the audio tones and interprets those into frequencies and sends
the codes to a DDS to reproduce the original. Andy's system is entirely
agnostic about the software using to generate the tones, where I have to
patch in my extract-to-UDP source into wsjt-x.
Very early days, and I'm too tied up with machining radio bits to do
much coding at the moment. However, a new AD9912 DDS board just arrived....
Neil G4DBN
http://g4dbn.uk
On 12/21/2020 12:25 AM, lstosk...@cox.net wrote:
Seems foolish to build a whole SSB rig to convert the audio signal
from WSJT to RF when at low frequencies some direct synthesis should
be possible. Or am I dreaming?
N0UU
_______________________________________________
wsjt-devel mailing list
wsjt-devel@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/wsjt-devel
