My NoV for 500kHz came through yesterday, authorising 0dBW ERP, modes
FSK, PSK, MSK and TBD just as I'd asked for. It also allows ASK, but
didn't specifically request that :-)
This morning got the 7m Tee antenna resonant using an old Decca
variometer with a few more turns added, and made a 550kHz Low Pass
filter for the Tx. Have just now tested the Tx + ant at 120 Watts RF
from a commercial 0.5 to 40MHz broadband PA and all works fine -
apologies to anyone who saw the carrier at 503.0000000kHz for a few
minutes at 1430z. I couldn't modulate at that point.
Antenna measures 28 ohms loss resistance at a first quick look, but
I'll measure more accurately when the weather is a bit warmer / drier!
If that is the figure, then efficiency will be in the region of -25
to -23dB and with the PA wound up to maximum rate dOP of 200 Watts
should get within a few dB of the power limit. Loss resistance of
that antenna follows a steady progression with frequency :
73kHz 200R
137kHz 100R
502kHz 28R
1900kHz 18R
That is, of course, total loss resistance including the wet tree and
roof tile lossy-near field effects.
Next stage is to complete the IQ converter direct from soundcard
output. I'm trying to come up with a simple design that won't need
much tweaking to get a decent sideband isolation. Last week tried a
design with a pair of NE602 chips (similar to FIgure 3 in
http://www.g4jnt.com/IQConverters.htm but it needed too much phase
and amplitude adjustment to get the unwanted isolation to an
acceptable -40 to -50dBc, so not really suitable for publishing. Now
going to try one based around FST3126 bus switches - near enough a
Softrock in reverse.
For transmitting on the band I do already have a suitable upconverter,
Fig 1 in the above reference, but that is over complicated for a
plug-and-play construction project and uses obsolete chips.
For soundcard data modes, with an audio bandwidth of no more than a
few hundred Hz, a simple CR I/Q network will generate the quadrature
drive to the image reject converter. For wider B/W, a more complex
polyphse or allpass network is needed.
Of course, if only the datamode software authors could only be
persuaded to modify their code for I/Q stereo outputs, it would make
linear upconverters direct to RF a whole lot easier! Come on
writers, its only a few lines more code!
So, nearly there with linear Tx, and an NoV that permits any mode
within constraints of "reasonable bandwidth"
Andy G4JNT
www.g4jnt.com
2009/1/16 brian hodgson <g3ykb@yahoo.co.uk>:
> Hi Graham -
>
> Make that 4! I've modified my TS850 to generate SSB at 500 kHz and use a
> homebrew linear PA to produce 30 watts peak RF.
>
> Have been off air for a while during a shack rebuild and broken aerial but
> am back in business.
>
> Brian - G3YKB
>
> --- On Thu, 15/1/09, Graham <g8fzk@g8fzk.fsnet.co.uk> wrote:
>
>
>
> G0NBD Refurbished Military and Marine
>
> M0BMU His own design Linear class E system
>
> G4WGT Classic linear HF/MF home made transverter
>
>
>
|
