Audio_Processing

After the downconverter produces a quadrature pair of baseband (audio) signals, the ASP does the actual demodulation. The most important mode is image-reject SSB demodulation by the phasing method. This is fully explained in the R2Pro receiver writeups, and my implementation follows those designs almost to the letter. What I did differently is to use opamp-based active filters instead of LC filters. I won’t show all the schematics here because such filter designs are routine.

At the input end of the ASP, I used INA128 instrumentation amplifiers to reject common-mode noise and to break up any ground loops. This is a very powerful technique, useful when you need to ship audio signals around between boxes.

Next, I included a recommended phase balance adjustment. That, in conjunction with the usual gain balance, allows you to optimize alternate sideband rejection. I put the pots on the front panel and made a little table of settings for each band. It’s simple and effective.

My receiver also has a binaural mode, another of KK7B’s contributions (QST, March, 1999, and reproduced in the ARRL Handbook). This is a fascinating kind of receiver to listen to because your ears are fed the raw quadrature signals, which produces a very cool spatial representation of all the signals near the VFO frequency. It’s fun when you want to quickly tune around the band and see what’s happening. It’s also a great way to find perfect zero beat on a carrier, like WWV. As you approach zero beat, you hear the sound swirling and bouncing between your ears, gradually slowing to a stop when you’re right on.

The volume control is a dual linear pot (RV4N) modified for quasi-log response. A nice article on potentiometer responses is available here.

After the highpass filter, there is a muting switch (FETs), and then the final stage is a summing amplifier where the sidetone is injected.

Audio Signal Processing (ASP) Box

A good sidetone is a clean sinewave at your favorite frequency, and it should switch on and off without clicks. The one I chose is based on simple a twin-T oscillator running at low voltage and low current to limit the transistor’s gan-bandwidth product. It produces pleasingly low distortion and noise.

The first thing you have to do is avoid coupling any changes in dc level to the audio output when switching between transmit and receive. If you study the schematic, you’ll see a .47 uF capacitor before the T/R switch and volume control. That’s all it takes. Don’t put the coupling cap after the switch!

The other thing you must avoid is suddenly turning the oscillator on an off, which generates nasty transients, identical to key clicks. I used an RC combination to limit the slew rate of the T/R control signal that operates the 2N7000 switching transistor. The result is a crisp and click-free sidetone with a modest parts count.

Sidetone Oscillator