2007-02-18
I went to the Wyong Field Day today. I didn't pick up much that directly relates to the challenge, but I did have some thoughts this evening about a replacement IF. It may seem odd that I am still thinking about the receiver, when the challenge is purely transmitter related, but until my MOSFETs come in the mail I have nothing to work with that can deliver beyond a watt or so.
The MK484 is nice, with trivial AGC and a low component count, but it has some drawbacks. It is a black-box, you can't really build a high-performance multi-mode IF with it. For AM-only, it is quite good however.
My design work on the Q-multiplier for the current IF made me realise I could just use the Q-multiplier as the only gain-device in the IF stage, essentially making a regenerative detector. It would also be quite easy, and perhaps preferable to add another FET as a so-called "infinite impedance" detector for AM. This removes the relatively expensive MK484 and replaces it with two FETs. The AF gain would need to be increased a little, or an additional IF gain stage added. An additional IF buffer could drive multiple detectors, including an AGC detector and a real product detector with an switchable high/low offset BFO. Crystal filters could define the selectivity, or the regeneration on the first stage would.
I considered exposing the IF regeneration control and letting it be advanced into oscillation for product detection, but this would mean AGC would be hard to implement and the "BFO" would be centred in the IF passband, rather than offset to either side of it. You really can't cheat if you want a true superhet with its performance advantages.
Image rejection will be a problem with an IF in 400 kHz region. For the challenge transceiver this isn't a big deal as I can tune the front end quite tightly, but for a general purpose shortwave receiver a higher IF would be preferable.
I need some die-cast boxes for this project. The only ones I have in the junk box aren't really the right size. I figure I'll put the receiver largely in one box, the transmitter in two boxes and the whole works inside a larger instrument case with all the outside-world interfaces and a PSU as required by the challenge.
Leave a comment on this article.
Parent article: "2007 80m Homebrew Challenge".