Late last summer and early in the fall, while watching the horrible events of 9/11 unfold on a portable teevee out in the shop, I finished the prototype of the Doerle type '19 regenerative receiver. After a few modifications, it worked so well I decided to use it "as is" for now. In addition to broadcasts from numerous foreign countries, such as Turkey, Australia, England, and Serbia, as well as "numbers stations" and pirate broadcasters, this radio has heard amateur stations in Namibia, Cyprus, Japan, and Mauritius Island. Perhaps the biggest coup came the night I was tuning around and came across K1B on 20m. I switched over to the Omni D, broke the pileup with 200 watts, and worked him! Later I may rebuild the Doerle so that it is more cosmetically attractive, but then again, I may put that energy into building a regenerative superhet...
And every now and then I get lucky and catch pirate radio station Crunch Radio playing a variety of late 'twenties to mid 'thirties music (in beautiful high-level plate modulated AM), and the Doerle becomes a true time machine.
Bear in mind as you view the pictures below, that the attempt here is not to reconstruct a station as it might have suddenly appeared, like Athena from the brow of Zeus, if a ham (newly wealthy from, say, some job in a defense plant created by Lend-Lease) had decided to throw everything out and start all over with new equipment in 1941. Rather, I wanted the appearence of a station that had grown organically as the owner's interest and skill in radio construction increased over a period of a few years. In addition to being more historically convincing, this is VERY convenient as I am more or less a novice at homebrewing myself, and in the process of developing my own skills! So, the Doerle is typical of a reciever from a few years earlier. The actual design on which it is based was published in 1934, and the look as well as the chosen design of the receiver is intended to reflect a beginning of interest in shorwave radio, such as so many unemployed and otherwise restless young men developed in the days when the NRA had nothing to do with guns.
The schematic of the receiver, and a technical discussion, can be found here.
This is, of course, the front of the receiver, with its attendent coils, headphones, etc. The home made "A" battery may be seen peeking around the right hand side.
Many people "breadboard" simple regens on open wooden chassis with only a small metal panel or metal backed panel to shield the tuning capacitor. This was indeed how they did it in 1922, but these radios often suffer badly from hand capacity, the tendency of the radio to change frequency with the changing positition of the operator's hand. By 1934 they knew better, and the original Doerle kit used a one-piece metal chassis, with the front panel formed by bending it upward 90 degrees. My version uses the front panel/chassis method of construction. It is gounded into my station ground system only via the shield of the coaxial cable, and still exhibits ZERO hand capacity. This is one reason why the radio works so well on 20 and 16m, as hand capacity problems increase with frequency.
Thanks to use of a relatively large 140 pF bandset capacitor, only five coils are needed to provide continuous coverage from 315 through 15 meters. The tuning rate is a bit fast at 15m, though, and I consider the upper limit of the useful range of the receiver to be the 16m broadcast band.
The large National dial drive is a retrofit, one of the modifications made after the radio was fabricated, and it was a job grafting it in there! However, it was necessary, as the tuning rate on the bandset capacitor was much too great, even with a large TRF knob in place. What saved the situation was the fact that the TRF knob and National drive are almost the exact same diameter.
The smaller knob on the left is the regeneration control. It is a variable resistor that works by controlling the plate voltage of the detector. Capacitor control of regeneration is much more elegant and noise-free, but potentiometer control is very effective and cheaper. The knob on the right is a volume control, which wasn't used in the original design. WWCR and Radio Marti weren't around with their respective megawatts to de-sensitize your ears in 1934! In the original design, the filament voltage control was located where the volume control is now. I moved the filament rheostat to the back of the radio; if I had to do it over again, I would put it back on the front panel somewhere so it would be more convenient.
The back of the receiver. The yellow tubular capacitor on the left is the audio coupling cap. The type '19 tube was mounted horizontally in order to keep lead lengths short. This is a traditional method of mounting the '19 tube. The knobless variable resistor in the lower left hand corner is the filament rheostat; I haven't found a knob I like yet. The top of the home made 90 volt "B" battery is just visable in the lower right hand corner of the photo.
Note the use of shielded cable for the audio section. This is an important preventative measure, excluding RF from the AF portion of the receiver. Again, this was not often done in 1922, but by 1934 people had caught on. Without the shielding, low-frequency audio oscillation, known as motorboating because it sounds like an old-fashioned one-cylinder marine motor, can occur. Rather than build a radio and then see if one encounters motorboating, why not just use shielded cable to begin with?
Close-up showing the construction of the variable inductive antenna coupler. The 80m coil is plugged in; this coil actually provides coverage from about 2 to 6 MHz. The antenna coil is wound from enamel magnet wire on common PVC stock. PVC can absorb radio waves, but this is not a problem at the power levels involved in receiving radio signals (transmitting is another matter). Note the arrangement of alligator clip and multiple taps. This allows yet another method of varying the amount of coupling.
Over a year of using this radio for DXing the ham and broadcast bands has revealed that controlling the amount of regeneration with the pot gets rather tricky above the 19m broadcast band, but that the variable coupler allows smooth control of regeneration all the way up to the 16m band. This is another very important benefit of infinitely variable coupling.
The variable inductive antenna coupler also provides a method of matching the high-impedance input of the detector to my many low-impedance station antennas. Use of large outdoor antennas is just as essential to DX work with a regen as it is DX work with a superhet. Most regen builders, however, use a small antenna, often indoors, to avoid problems encountered with frequency swing caused by the physical movement of the antenna as it blows in the wind. This can be a real problem with capacitively coupled regens. Using the inductive antenna tuner and loose coupling eliminates much of the frequency swing, although it is still present on very windy days (and we have lots of those on the Texas prarie!). Even then, though, it is never objectionable enough to render the receiver useless. A better solution is to use an RF stage ahead of the detector. This stage usually provides little gain, perhaps as low as a factor of two or three, but isolates the antenna from the detector very well.
The inductive antenna coupler is simpler than adding a stage, though, and it has worked very well. I recommend it highly. It was also inexpensive to construct, as the 1/4 inch phono jacks used as shaft bearings for the dowel rods were the largest expense. Note the split on the lower dowel rod; yes, I pre-drilled the hole, but I didn't use a big enough drill bit!