More Medium Wave
Posted: Fri Oct 25, 2024 3:55 am
I've been doing some experiments on an "outphased" AM design. It uses a simple CMOS synthesiser which generates a sinewave at the carrier frequency (digitally generated), then the "outphasing" is achieved by driving a couple of comparators with the sinewave into one input of each comparator, and the modulation and inverted modulation into the others. The output from the comparators is fed to a couple of bistables, which give a wavetrain of square waves, with variable mark / space ratios according to the modulation.... The two logic outputs are amplified by FET amplifiers, and each is fed into an inductor. The inductors meet at their output ends, and there's a capacitor to ground (for resonance). The mixed signals at the combination point equate to AM! There's then a lowpass filter to keep everything clean 'n' green, and the output feeds the antenna directly.
At the moment, the prototype is delivering 60 Watts carrier / 240 Watts peak, and is capable of 125% positive-going mod (just by tweaking the comparator stage). The recovered audio sounds superb - I was concerned that there might be distortion, but these fears proved to be unfounded. Initial listening tests suggest very low percentage distortion, and I'll carry out some actual measurements later.
The beauty of the design is that it can be fully modulated with just a couple of volts of audio. There are no big chokes needed (or modulation transformers), and efficiency is over 90%! It's also reasonably frequency-agile (similar purity over about 220kHz of the MW band). The only minor snag is that the output inductors need to be slightly variable, but I'm considering some kind of variometer to provide this adjustment...
Cicuits to follow shortly.
At the moment, the prototype is delivering 60 Watts carrier / 240 Watts peak, and is capable of 125% positive-going mod (just by tweaking the comparator stage). The recovered audio sounds superb - I was concerned that there might be distortion, but these fears proved to be unfounded. Initial listening tests suggest very low percentage distortion, and I'll carry out some actual measurements later.
The beauty of the design is that it can be fully modulated with just a couple of volts of audio. There are no big chokes needed (or modulation transformers), and efficiency is over 90%! It's also reasonably frequency-agile (similar purity over about 220kHz of the MW band). The only minor snag is that the output inductors need to be slightly variable, but I'm considering some kind of variometer to provide this adjustment...
Cicuits to follow shortly.