[Lowfer] IF Bandwidth Freeware

[Lyle Koehler]

While playing with this interesting software as an enhancement for my Icom
PCR1000 receiver, I decided to connect the output of my balanced loop preamp
directly to the mixer in the homebrew downconverter, and see if it was
sensitive enough to use as an "almost direct conversion" receiver. QRN is
pretty bad today, so I picked a relatively loud NDB (FS, Sioux Falls, SD,
245 kHz) to listen for. And I used the usual weak-signal trick for copying
NDBs: rather than listening in AM mode on the carrier frequency, I tuned to
246 kHz with a narrow CW filter to look for the keying sideband. With the IF
value in the ifdsp.ini file set to 12500, low-side local oscillator
injection, and a divide by 16 circuit in the oscillator chain, I needed a
"crystal" frequency of 3736 kHz. Nothing like that in my junkbox, so I fed
the output of a signal generator into the input of the 74HC4060 divider.
Fired up the IFDSP software, and sure enough, there was FS. Decreasing the
bandwidth to 16 Hz and carefully tweaking the IF offset to peak the signal,
FS was bouncing the S meter display up to about 40 dB (fast AGC mode), and
the noise level was about 5 dB on the meter scale. Hmmm; another signal in
there. UKL; Coffey County, Kansas, also on 245 kHz. That's one I hadn't
logged before! UKL was also moving the meter, but only about 5 dB above the
noise level.

By the way, my downconverter circuit is slightly different from the one
shown on Vittorio's web site. I'm using a 74HC4053 analog switch and a
Mouser TL002 audio transformer to make a single-balanced mixer. The main
reason for the difference was to provide a higher input impedance.

The image rejection of this receiver isn't too good, but could be improved
by using a higher Q loop and/or a tuned RF stage in front of the mixer. It
would also help a little to choose a higher IF frequency, closer to the 22
kHz limit imposed by the sound card sampling rate. But the performance of
the simple system is really pretty good. The same thing could be
accomplished with DL4YHF's Spectrum Lab software, and maybe it could even
display a spectrogram of a LowFER QRSS signal. (I can't run Argo and IFDSP
at the same time because they would fight over the sound card, and using
Argo alone places an upper limit of about 3 kHz on the IF frequency -- not
too good for image rejection.) The nice thing about IFDSP, though, is that
it is very easy to set up and use for receiver experiments.

The next step would be to hook up the ZL1BPU DDS exciter as the local
oscillator and get rid of the signal generator. Unfortunately it's too nice
outside to spend all day playing in the shack...

Lyle, K0LR