[R-390] Tuning SSB

[Jacques Fortin]

Larry, Giovanni,
If I can put my 2 cents comment on this:

If we are really speaking about of the "noise sensitivity" of a BFO-injected ENVELOPE demodulator compared to a PRODUCT demodulator, it is relatively easy to understand what is the product demodulator advantage relating to the passband noise.
An envelope demodulator is a rectifier (diode) loaded by a filtering capacitor and a resistor, both in parallel.
One major drawback of the envelope demodulators is known as DIAGONAL CLIPPING.
When the noise components adds vectorially to the SSB signal components, plus the injected BFO, the rectified signal envelope can raise instantaneously way over the amplitude of the BFO + SSB components alone.
The filtering capacitor that follows the diode is then charged over the "noiseless envelope" amplitude and the time took to discharge the capacitor and catch up the envelope amplitude again is quite perceptible.
In a product demodulator, no diagonal clipping occurs, as this is not a rectifier, but an analog multiplier.
The demodulated noise products are just added to the desired SSB output information and despite that they are perceptible, their subjective effect is way less disturbing than the diagonal clipping of the envelope demodulator.

What I understand from the Larry comments since the beginning of this thread is that a R-390A SSB demodulation is better, relating to the bandpass noise, by using a product demodulator (like the Lee circuit and variants) than the original envelope demodulator + BFO.

73, Jacques, VE2JFE in Montreal

Hi Gianni,  In reference to your question as to why there will be less static type noise coming out of the product detector than an envelope detector, I've not been able to find an explanation from an engineering source, but Don Stoner writes in his book 'New Sideband Handbook' from 1958 on page 191 in the Product Detectors section '.... there will be less interference since output can only occur when a signal beats with the bfo.'

VE2JFE: Not sure that this is the right explanation....

I believe that with the envelope detector with bfo injection there is no real limiting effect on what will pass through it, so all the noise on the IF output goes through.  Whereas with the product detector, having the bfo signal on the control grid has a very limiting effect on what passes through it.  So in a product detector, only some of the noise on the signal from the IF output will pass through.

VE2JFE: Not really, all the noise products will be there, but their effect is not the same.

This matches what I see from the testing that I have done.

VE2JFE: I agree with Larry on that.

Regards, Larry