Bandswitching - Bandspreading Circuit
Several years ago I built a general-coverage FET regenerative receiver, similar to a couple of popular commercial kit receivers. While it worked well, I wanted better bandspread on the lower HF ham bands, specifically the CW subbands. This article isn't a construction project, but it may give you some pointers if you decide to attempt something similar.
The receiver, which covers the CW subbands of 80, 40, 30, 20 and 17-15 meters, is built in the case of an RCA rf signal generator. I used the supplied variable capacitor, which has a built-in vernier drive, and bandswitch. The bandswitch has two poles and five positions. In the original receiver, I supported the tuning inductances between the two poles (decks) of the switch. The original variable capacitor is about 400 pF, much too large even for 80 meters. Originally I placed a fixed 100-pF capacitor in series with the variable, and paralleled the coils for each band with parallel capacitors, to obtain roughly the bandspread I wanted. Using a single fixed series capacitor didn't look as though it would work with the new arrangement. So one pole of the switch is used to select the series capacitance for each band; the other pole selects the coil, which usually has a capacitor in parallel as well. The schematic at the top of this page shows the bandswitching arrangement I used. More on bandspreading on page 2.
My first try at mounting the coils had them hanging by one lead off the bandswitch.
This obviously wasn't a good idea, so I fabricated a mounting pad from a scrap of double-sided pc board.
It's attached to the back of the bandswitch with solder lugs.
I also reversed the position of the rear deck so the terminals for capacitors and coils were oriented for shortest lead lengths. Even with the bracket, the coils still weren't secure. When I was satisfied with the tuning coverage I tacked each one in place with a drop of gel cyanoacrylate cement. The tuned circuits of a regenerative receiver operate at the listening frequency, and the smallest variation in stray capacitance caused by vibrating wires or coils will modulate the signals. Unless you're building an earthquake detector, use solid wire and securely install it and other tuned-circuit components.
Make It Solid
Regenerative receivers are sensitive to any changes to their environment. Changing the operating voltage of the FET detector changes its internal capacitance and throws off the tuning. You should regulate the voltage applied to the entire rf section of your receiver. I used a 7.5-V Zener, but an IC regulator might be better. My receiver operates from a 12-V battery, so the supply voltage stays relatively constant over the short term.
(Download the full schematic.) I used an rf choke on the dc power lead to cut down on broadcast-station leakage. If you use an internal battery you probably don't need the choke. Because the audio output is bypassed with large capacitors, I found I didn't need a choke there. A 1N4001 diode in series with the battery protects the receiver if I accidentally reverse the battery connections.
To help my tired hearing, I souped up the LM386 audio amp, using information gleaned from a past issue of Sprat, published by the GQRP Club, driven by a FET preamp. Now there's plenty of audio, even with less-sensitive stereo headphones. Some RC low-pass filtering cuts down detector and LM386 hiss.
The receiver is sensitive enough to hear band noise on a 30-foot (9-m) indoor wire, but internal noise is very low. Even when 40 meters gets crowded at night, I haven't been tempted to hook up an external audio filter.
Bandspreading was the most time consuming part of the project, and is covered on the next page.