Sensitivity and selectivity are issues that will invariably concern a short wave listener when he wishes to purchase a new receiver. Commercially available communications equipment will undoubtedly fulfill his expectations, but we are talking here of highly priced products. Low-cost alternatives call mainly for homebrewed radios, and in this sense, the regenerative receiver is the common choice due to its simplicity, low parts count and very acceptable performance.
The author is also a short wave enthusiast and for some time used his family's Philips MW/SW vacuum tube radio. Later, he changed to a solid-state Sony ICF-7600, a high selectivity receiver featuring ceramic filters in the IF stages. Then he would discover how much fun he could have building radios on his spare time. After testing a variety of schematics available in books and on the web, the author finally decided to make his own designs.
Back in 2003, an experimental Colpitts-type shortwave regenerative receiver was made public on the web by this servant. It was reported to tune from the 22-meter band up to the 11-meter band with a single set of coils. Operation was found to be satisfactory with the prototype built on a protoboard and a ground plane fixed underneath. Further work with the receiver using different sets of coils revealed a larger usable frequency spectrum.
Fig.1 shows the schematic diagram of an improved version of the early receiver. It will tune signals from 3500kHz up to 26MHz, roughly in three bands, each with a 1.96:1 frequency-ratio. It must be quoted that the said frequency ratio is what the author got for his prototype. It is a function of the maximum-to-minimum capacitance of the variable tuning capacitor and the stray capacitance of the circuit layout.
The Colpitts approach uses no tickler coil nor throttle capacitor for regeneration, as opposed to the Armstrong circuit. In our case, it resorts to transistor Q1's base-emitter input capacitance and NPO-type ceramic capacitor C2 to effect the correct impedance transformation and phase shift necessary for regeneration to take place. Q1 is the amplifier-detector and along with its associated circuitry forms the common-collector Colpitts regenerative amplifier, with R9 as the reaction control.
The oscillating mode is employed when copying CW or SSB, otherwise, the stage should be left very near the threshold of oscillation for maximum sensitivity and selectivity. It is best that the 100-pF variable capacitor C1 be a vernier type. It will help a lot when tuning-in the crowded SW bands. Q2 and Q3 form a high gain audio amplifier and ample volume should be expected at the output. This is why a volume control has been included in the circuit. A high-impedance crystal or ceramic piezoelectric earphone should be used for listening. Alternatively, an audio output transformer that will match low-impedance magnetic earpieces to 5kohms to 7kohms may be connected to Q3's collector through a 0.47-uF capacitor. High-impedance 2k-ohms magnetic headphones do not require a transformer. They can be connected directly to Q3's output using the said capacitor.
Components C9, R11, R12 germanium diode 1N34 (it can be any other germanium detector type), Q4, the S-meter and the high-efficiency blue LED form a pseudo "tuning-indicator" circuit which gives an idea of the available audio level at the output of Q3 after demodulation. We have then a visual relative indication of the tuning action of our circuit.
Solderless breadboards known as "protoboards" show parasitic capacitances of about 3pF between adjacent connection lines. Hence, excercising the builder some minimum of care in point-to-point connections on the layout, the protoboard will lend itself as an acceptable breadboarding material for the testing of simple receivers, such as the one dealt with here. The author has tested on a breadboard of this sort the design herein described, with an additional aluminum ground plane fixed underneath. This metallic plane is fundamental for reducing hand-capacitance effects when adjusting the variable tuning capacitor C1. In order to avoid ground loops it is advisable to connect the circuit's common ground to the plane at a single point. It is very important to use short connecting wires between components.
Shortwave frequencies are nicely tuned by the receiver using a random-wire outdoor antenna and a capacitive ground of the type described in Fig.1 above. It is not absolutely necessary to resort to a cold water pipe connection, or to use a metallic rod buried in the soil out there in the garden.
At this point we should avoid any possible confusion between the aluminum ground plane beneath the protoboard and the external capacitive ground connection via a 30cm x 30cm aluminum metal sheet thrown on the floor under the workbench. These are two different issues.
Unwanted interference produced by FM or TV stations is easily blocked connecting a small 10-uH RF choke between the antenna lead-in and the receiver. The author uses two of these chokes in series, which gives further better results. It is best for the RF choke to have low distributed capacitance, so that VHF FM or TV signal leak-in problems can be avoided.
Following, some pictures of the prototype are shown. The RF attenuator R13, the reaction or regeneration control R9, the variable tuning capacitor C1 and the S-meter are mounted on an aluminum panel fixed to the protoboard's ground plane. It is important to have a good mechanical and electrical joint between the aluminum panel and the ground plane.
Views of the completed receiver:
Pictures of My Prototype
Below a sheet of metal used as a ground system:
author: Ramon Vargas - Lima, Peru.