Softrock Lite II Softrock Lite II: Local Oscillator
Band: 20m

The Local Oscillator stage implements a basic Colpitts Crystal Oscillator with a buffer stage to increase the signal level. The oscillator produces a signal that is at the crystal's specified fundamental frequency of 18.73 MHz.

In reality, for each frequency the crystal circuit will oscillate at a slightly lower frequency (~ - 1 kHz), due to the capacitive divider (C4/C5) pulling the crystal down somewhat. The effect is more pronounced for the higher bands.

The output of the Colpitts Oscillator is fed to U1, which has the effect of dividing the oscillator's fundamental frequency (18.73 MHz) by 4 AND (most importantly) shifting the phase of its 2 output signals (each 4.6825 MHz), such that they are in quadrature (90 degrees apart in phase)

Sub-harmonic Sampling

Alan, G4ZFQ points out that on the 30m and 20m and three of the IF option receivers, the Local Oscillator produces a signal that is 4/3 times the desired center frequency as opposed to the 4x the center frequency output for the lower band models.

"Subharmonic" works like this:

• The LO for this 20m RX outputs a 18.73 MHz signal that goes to the dividers /4, resulting in a 4.6825 MHz square wave ( rich in odd harmonics) being fed to the mixer.
• At the mixer, a strong 3rd harmonic is present on the clock inputs, along with the fundamental of 4.6825 MHz. The 4.6825 fundamental multiplied by 3 yields the third harmonic of 14.0475 MHZ.
• The Bandpass filter (BPF) performs the essential function of severely attenuating any signals centered around the 4.6825 MHz fundamental frequency and first harmonmic, but allows 20m signals centering around the third harmonic (14.0475) of the 4.6825 MHz LO output.
• The result is that the mixer is dealing with signals in the passband, centering on 14.0475 MHz, as though the dividers were passing a fundamental frequency of 14.0475 to the mixer. BPFs are all that stop Softrocks from working on unwanted frequencies!
• (Thanks to Mike Collins, KF4BQ, for the following:)
• A common analogy is a strobe light on a fan....you will see the same thing when strobing for one revolution, also for two revolutions, also for three, ... The mixer is sensitive to the fundamental and all harmonics of the mixer sampling freq, but the transformer (with 180deg phasing) and the hookup of the mixer cancels out even harmonics of the freqquency, leaving only the fundamental and odd harmonics. Just as you would see with a strobe light the signal will be reduced (over the same time period). The reduction for the 3rd harmonic is 1/3 the voltage on the holding cap. That give a overall 20log(1/3) = -9.54dB loss in the signal compare to receiving the fundamental. This is the drawback to sub-sampling; i.e. a reduction in sensitivity. It does allow the mixer to switch at a much lower rate which is why it is used.
• The front end filter rejects the fundamental and also the 5th and higher harmonic products so you are left with only the third. The harmonic rich LO is often misstated as why the receiver works with harmonics. If the LO was a sine wave it would still work fine (maybe more phase noise, but basically the same).