0

I'm at best a hack EE but I'm attempting to design a 20 meter receiver for learning reasons. I have settled on a superhet and I'm trying to avoid looking at existing designs, the struggle helps my understanding I find. I'm not attempting to build anything world class, I have commercial gear, just looking for functional.

So for the RF filter I'm looking at an RC high-pass pointed vaguely at the LO frequency, the C is (at least partially) a voltage controlled variable capacitor. For the LO I want to build a VCO colpitt using the same or similar varicaps.

So my question is, is there enough math to find a single or two voltage swings that will keep them in tune with each other? Or am I chasing something that wont work/have unobtainable values?

foreverska
  • 147
  • 6
  • 1
    A front-end **RC** filter? Will dissipate away small desired signals, and provide too little attenuation of unwanted signals. The more difficult-to-design **LC** front-end filter should be considered, even for less-than-world-class design. You might even be able to fix-tune it (no varicap) for a small frequency span, making the varicap tracking problem moot. – glen_geek Jan 17 '20 at 17:12
  • @glen_geek That would have to be a pretty narrow tuning range, no? Like 20Khz before the stuff at the bottom of the high-pass starts to image, right? I will consider the LC front-end though, thanks. – foreverska Jan 17 '20 at 17:33
  • Presumably you have discounted a direct conversion receiver for reasons that would be worth mentioning. – Andy aka Jan 17 '20 at 17:58
  • 1
    Consider an LC-type front-end filter for 20M ham band (350kHz wide). A loaded Q of 40 passes the whole band. It wouldn't *need* a voltage-var-capacitor. A tune-once trimmer capacitor would suffice. You might add proposed frequencies of local oscillator and intermediate-frequency. Note that Andy's direct-conversion (zero I.F.) idea makes tracking extremely easy too. – glen_geek Jan 17 '20 at 18:37
  • I discounted the direct conversion receiver because I had a fundamental misunderstanding of mixer products and couldn't wrap my head around DCRs. I think I now see the flaw in my logic. A search for "negative frequency" helped out. I'm going to have to tackle one of these. – foreverska Jan 17 '20 at 18:51
  • @glen_geek Let me see if I have the concept here right. I want a bandpass LC. Then I want to choose an LO higher than intended frequency by a sufficiently large IF so that the other frequency that would image is well outside the bandpass. – foreverska Jan 17 '20 at 19:32
  • 1
    Yes, yours is a valid start-of-design. Another equally valid option is L.O. *below* intended frequency by a sufficiently large I.F....this may be a bit more attractive because a lower-frequency L.O. is easier to minimize frequency drift. – glen_geek Jan 17 '20 at 21:39
  • Mmm fair point. Thanks so much for the help. – foreverska Jan 17 '20 at 23:28

0 Answers0