Loading ...
Sorry, an error occurred while loading the content.

23027RE: [regenrx] Re: OOPS!

Expand Messages
  • davidpnewkirk
    Mar 1, 2015
      Brian wrote:

      > Just as a thought experiment, how many low gain RF stages ahead of 

      > a regen detector would it take to provide a useful improvement in 

      > adjacent channel strong signal rejection? A kind of  I.F. strip up front, 

      > let’s say for 40m, and ignoring tracking problems for the moment.

      Long ago and far away, hams and even professionals equated "increased image rejection" with "adding more RF stages," because adding more tuned circuits was always assumed to require amplifiers between them, or at least between coupled pairs of them. So it is that when we survey the ham literature of the day, "more RF selectivity" always meant "add RF amplifier stages." And so it is that receivers like the HRO, built to answer the challenge of images only 910 kHz away as a result of its 455-kHz-ish IF, had two tuned RF stage--and overloaded badly on strong signals, because any unnecessary gain ahead of its iffy pentode mixer was a bad idea. If the gain added by intermediate amplification doesn't stop us by creating fake signals through intermodulation distortion, the noise added by the intermediate stages will.

      The answer to your question is that even at 160 meters, regular inductor-capacitor tuned circuits aren't selective enough to begin to knock down even adjacent AM signals--forget about a CW signal a few hundred hertz away--even if you insert many tens to hundreds of such circuits in the line, the gain, noise, and distortion of any intermediate amplification notwithstanding. The physics of the pass-band-width achievable with real-life inductors and capacitors at 1.8 MHz and higher make it so.

      RF Q-multiplication is also not the answer, as it concentrates all the negativity of overloading and noise into one stage--not to mention that, unlike the desirable selectivity of loss introduced by passive filtering, Q multiplication can introduce only selectivity of _gain_, requiring us to insert attenuation downstream to begin to realize its benefits. RF Q multiplication is a non-starter, Goodman's Miser's Dream experiments notwithstanding.

      On the other hand, _crystal_ filters capable of letting pass only a narrow slice of a band are quite realizable, and some hammy companies have marketed them in the past for key DX subbands--see http://www.inrad.net/home.php?cat=146 . A quick search found the non-hammy company Filtronetics ( http://www.filtro.net/content/front_end_filter.html ) currently selling custom such filters. But we can expect that buying a signal-width _IF_ filter and putting it behind a strong frequency changer (see my posting about using a diode DBM, LO, and post-mixer amplifier as a basic for easily achievable excellent basic-radio performance) would be a better use of our time and money, as doing so allows us to scan that well- and simply-implemented narrow receiving window across an entire band or bands at will. Long live the superhet!

      Best regards,


      amateur radio W9BRD

    • Show all 35 messages in this topic