Low Voltage (12VDC) Two Tube Regen Receiver Using "Space Charge" Technology

Before I go into the details, let me say that I've built about 1/2 dozen one and two tube Regens over the years (and several superhets) and this one is by far the most selective, most stable, and easiest to use Regen. The 12AL8 and 5687WB or 7044 tube combinations work well, as does the 12AL8 and 5814WA (12AU7) combination with the plate voltage reduced to around 8V. Copying CW and SSB, as well as AM SW stations is easy. I have not had a chance to try more of the MANY other dual triodes like the 12AT7, 12BH7A and they may all work as well. I'll expand this write-up as other items come to mind. There will be a low voltage (24V ? 48V ?) CW transmitter to follow.

This was a fun project and all started when I was looking for low voltage designs for a Regen Receiver and matching Transmitter. Several of the guys on the Glowbugs (GB) Reflector reminded me of "Space Charge" tubes developed in the 60s to make car radios which did not require vibrators (used a power transistor for output). Their use was short lived with the advent of fully transistorized radios. The concept was actually invented in the 1920s but I don't know how widely it was used.

Normally, tubes operate with hundred(s) of volts of plate potential or around a hundred volts for battery tube sets. To ask them to work at very low voltages, something has to change because the lower cathode to plate potential is just not enough to enable enough current flow. The other problem is that the plate curves appear to become very non linear at low voltages. One method of improving the electron flow is to change the spacing and physical design of elements inside the tube. Another is to "boil" more electrons off the cathode by increasing the wattage of the filaments through redesign. A third is to place the 1st grid at B+ potential, thereby providing a nearby "pull" on the electrons from the cathode. Since it is a "grid", many of the electrons reaching the 1st grid zip right on through. The second grid (screen) is used as the "control grid" and the B+ charged plate gathers up the electrons. There is no suppressor grid.

The disadvantage of these low voltage tubes is that they consume lots of 12V power, which for a car radio, does not matter. For example a "normal" dual triode, miniature tube, may draw 150mA of filament current ......these may draw 350mA. The "space charge" Tetrodes may draw 500mA+ filament current. In addition, the Power Tetrodes with grid 1 tied to B+, may draw 70mA on the grid (not useful power, not input controlled) and half that on the plate (input controlled and useful). While this makes them very safe to use from a voltage standpoint, it does generate heat and they do get HOT. The reduced spacing between elements increases inter-electrode capacitance which reduces their usefulness at RF frequencies but they work well at AF frequencies. While all this is required in the "POWER" sections like the AF amplifier, it's NOT necessarily required in the low power VOLTAGE amplifier sections like the RF/IF Amplifier and the Detector in a car radio, or any other radio for that matter.


Basic Regen design is pretty common, so BEFORE the schematic comes the "concept stage", the "what's available in the junkbox stage", back to the "concept stage", followed by the "mechanical/electrical layout stage", then sketch a rough schematic ......and do it all over again. Paper/Computer layout is cheap.

After researching the "space charge" tubes, it was decided to go with the 12AL8 triode/power tetrode combination. These tubes are readily available and I found some NOS for $2 or $3. The Sylvania specifications for the tube (available online) also show a handy audio circuit which I used .....after a few modifications. It uses the triode as an AF voltage amplifier feeding the power tetrode.

For the Regen and RF amp, I looked for dual triode tubes likely to "boil" electrons with higher filament currents. The first ones found are "computer tubes" the 5687 and 7044. These are said to also be very low noise types ? I don't know relative to what. The 5687s are more readily available since audio guys use them, but both types are available NOS for $2 or $4. Since there are MANY dual triode miniature tubes out there (12AT7, 8414, 12AU7, 12BH7A, 12BY7, etc), why not make the design able to use as many as possible ? The two basing diagrams most frequently seen are 9A/9AJ and 9H, so why not a switch to change between the two pin configurations ? Turns out, only the plate and cathode of one triode section require change. I used the "stable" triode for the Regen section and the switched triode for the RF amp.

The other thing to keep in mind is that the requirement for "more power" is true in the "power" sections like the AF amplifier, but it's not necessarily true in the voltage amplifier sections like the RF amp and the Detector. What you are looking for there are good low plate current/voltage characteristics (curves). Tubes like the 12AU7 and its ruggedized 5814WA work fine ....with reduced plate voltage. The advantage of these tubes, over a 5687WB or 7044, is that you save 300mA of filament current and reduce the total Regen power from 1.1A to 0.8A.

The coil and tickler were wound on a ceramic ARC-5 Tx oscillator coil form. Any circuitry associated with the Regen section should use ceramic where you can, short leads, stiff wire, good wiring practice. If you take a look at the pictures, the wire going to the "Frequency Set" and "Frequency Adjust" capacitors is #18 solid wire suspended --OFF-- the chassis. This reduces the capacitive changes caused by slight wire movement. The capacitors are VERY solidly mounted to BOTH the front panel and the chassis.




WHAT ? & WHY'D HE DO THAT (L to R) ?

Top view showing adjustments

Coil bandswitch detail

Underneath chassis detail