Crystal Radios Of The 2009 Contest Entrants
Mike Tuggle, BCB Class and Below BCB Class
The inspiration for this set came from the flurry of one-tube regenerative sets that recently appeared on Dave Schmarder's RadioBoard site. Here, a MOSFET replaces the tube. Distinctive, but not unique, features are source-derived positive feedback link and use of the forward drop across a silicon diode to establish a current-stable gate-to-source bias. The diode bias works particularly well on LW. I use the variable bias for BCB listening. Once bias is set, it is good for the entire band.
The set features no tuning dial scale. Tuning is done by the "count the heterodynes" method with the set running just into oscillation. A spotter radio was rarely used, and then it was used mainly to compare possible DX stations' programs to my local C2CAM outlet in order to see if they were C2CAM. Running in oscillation doesn't make for good audio quality, but it is great for ferreting out weak DX signals.
The ferrite rod coil and variable capacitor were salvaged from a junked Pioneer tuner and used pretty much as-is. Along the way a few tweaks were needed to get regeneration, most notably very light antenna coupling. And, the question arose, what to do with gate 2 on the MOSFET? I tried grounding it, and applying positive bias from a voltage divider, but the best results were obtained by simply tying it to gate 1.
The MOSFET is a Siemens "1216" obtained from an old rusty UHF tuner out of a Zenith TV set - after some prying and de-soldering. I can find no info on this device but suspect it may be an NTE455 equivalent. Provenance aside, this MOSFET is orders of magnitude more sensitive than a BF960 which was marginally more sensitive than the 3N211 that I started out with. The moral of the story is don't be too proud to scrounge!
With the extra gang on the tuning capacitor it took little effort to rig it up for long wave. Unlike medium wave, LW reception is much better when the antenna is connected through to the ground. Extra polystyrene capacitors are used in shunt to get down to 200 kHz. It has now become my primary LW set.
Output is to a Bogen T725 transformer wired as a 'universal match' auto-transformer. The 10 and 20 k-ohm taps worked best on the MOSFET output. The 40 k-ohm tap is just too much, and the set stops working. It appears the DC resistance drops the drain voltage to a point where the MOSFET ceases to operate. The 300, 600 and 1500 ohm outputs to the phones are valuable in peaking audibility which changes with program content (voice, music, code tones). Many thanks to Richard O'Neill for the Bogen.
Robert Weaver, BCB Class
This year's contest receiver is a superhet based on a 6ME8 beam deflection tube. I had previously built a reflexed direct conversion receiver using this tube, which worked very well. Vlad Novichkov, over on ARF, suggested a superhet version. Ideas flew back and forth, and the result was a double reflexed superhet. One tube acts as oscillator, mixer, IF amp, and AF amp. Audio detection is done using a 1n34A diode. Originally, I didn't expect this to be anything more than a novelty, if it worked at all. However, after spending a bit of time refining the circuit, this is definitely a first rate DX receiver. It has enough gain that a loop antenna is sufficient, and headphones are unnecessary; it can drive an ordinary 8 ohm loudspeaker with ease. In fact, as a challenge, I decided not to use headphones at all in this contest; all of the 157 stations logged were heard and ID'd by listening to the speaker only.
I've had some interest in using a beam deflection tube for a while, and when I built the direct conversion predecessor to this receiver, I didn't have a lot of technical information to go on. I wasn't even sure that there would be enough gain between deflectors and plates for it to oscillate. I threw together a quick and dirty circuit to test the oscillator, and found that it would indeed oscillate; and then went on to build the first receiver. Since that time, some very good technical articles were found which were a tremendous help in designing the superhet version.
Two were articles from QST (Mar. 1960 & Sep. 1963), and the other was a technical article from RCA Review (June 1960) which introduced the 7360 beam deflection tube. Highlights from these articles indicate, a higher conversion gain than is attainable with the best pentagrid converters, extreme linearity, extremely high rejection (60db) of common mode signals. I won't go into a long drawn out discussion of the circuit. For that, you can look at the original discussion on ARF that led to the current receiver.
Here is a description on my site. However, I would like to mention a couple of the unique characteristics of the beam deflection tube that make it such an interesting candidate for radio homebrewing.
Beam deflection tubes were designed to be used as balanced mixers/modulators. The most interesting feature is that any signal applied to the control grid will be amplified and appear as an in-phase (common mode) signal at the two plates. On the other hand, any signal applied between the deflector plates will be amplified and appear as an out of phase (differential mode) signal at the two plates. The sum and difference signals created by mixing the control grid signal and the deflector signal are also differential signals at the plates. This allows the different signals to be separated and redirected as necessary. This unique characteristic allows reflexing signals, at the same frequency, back through the tube without it breaking into oscillation. That is the fundamental principle behind the circuit used in this receiver.
In operation, the volume control is turned all the way up, except when tuning local stations; the RF regen control is normally left at the 80% postition. The IF regen/plate balance control is adjusted to the fully balanced position (no audio squeals, and very little background noise), then readjusted slightly until a typical regen hiss is heard. This is the sweet spot for optimum sensitivity and selectivity. It must be readjusted a bit from time to time while tuning across the band. There are two spots on the dial which are especially sensitive to oscillation. One is around 685 kHz (which I believe is due to intermodulation of local stations); the other is at 910 kHz, the second harmonic of the IF. However, by backing off on the IF regen, these present no serious problem, and I was able to log stations at both of these frequencies.
The loop antenna provided enough signal, and it's directionality helped tune out local stations while I logged stations on adjacent frequencies without the need for wavetraps.
Overall, I'm very pleased with the results, and expect to do more experimentation with beam deflection tubes. For anyone looking to try something different in a receiver circuit, I recommend having a look at beam deflection tubes. They are readily available, quite cheap and have a lot of potential.