Continuing education

MCU projects

Amateur radio

Broadcast radio

High voltage



1920-1930's style broadcast 1 tube regenerator or tube "Bloomberg" radio

I had been limited to working with semiconductors for many years after graduating from university until one early spring morning a few years ago. I was walking home after running errands in the city when, I noticed an old fashioned radio set in the gutter, similar to the type we had had when I was a child. A tube receiver ... again! These things are becoming rarer than hen's teeth.

A light bulb went off in my head as I took a stroll down memory lane at the speed of lightning. I would have grabbed it immediately had the city sanitation truck not been so close to hauling it away. I only had time to pull a few tubes out of their sockets. As I continued walking home, I couldn't help thinking of experiments I would use them in. Even though I had studied them in school and knew how they worked, I actually never tried anything with them. The following is a short story of how I cut my teeth with TUBES. But should you, my visitor, try to replicate my experiments here, be forewarned. Tube technology has built-in adrenaline boosters should you accidentally touch something you shouldn't. It uses potentially lethal voltages so please be careful.

What hit me immediately was an idea of trying to make something really old, like 1920-1930's old ... a regenerative receiver or regen for short. Back then, regens were very popular; simple to build and provided very good results. One piece of the glassware salvaged was a triode-pentode 8JV8, meant for various IF and video work. In other words rf-grade component. It's datasheet lies here. The receiver circuit that emerged is pretty much a classic.

Signals from the antenna (a 2-meter piece of wire) are fed to the input tank via C3. Antenna coil LA was wound on a piece of 19mm diameter plastic tubing (in this case it came from a junked clothes hanger). There's a tap at 25 turns from the cold end. Naturally, my coil winder was employed to make it. The triode section works as a regenerative detector, being biased by C6 and R8. It's grid and cathode share the same DC voltage and regeneration is controlled by varying the plate voltage. Detected audio, taken from the plate, is first run through the low pass filter R4 and C4 and then fed to the pentode's grid. The pentode part is biased by C8 and R7. R6 and C9 provide power to the second grid. Finally, T1 matches the low impedance of the small speaker to the fairly high output impedance of the tube. For simplicity's sake there is no volume control or power switch.

An additional note: I was asked if it is OK to use an electrolytic cap for C5 as the reader couldn't find a suitable larger non-electrolytic. My experience was, it is not OK as all electrolytics I tried leaked enough to bring the triode plate voltage way down. R3 has be large to separate the plate from the biasing circuit. If there is an additional leakage through it, it lowers the plate voltage below the value where the stage no longer works. Also, electrolytics tends to be far more non-linear and that affects the smoothness of the regen control.

Constructively, I chose the open-board approach for my rig, as seen here on the left. In fact, many radios of that era were assembled by individual craftsman in that way - it allowed for easy access to everything under the hood. Also, changes in the circuitry were a breeze to implement. The front panel and horizontal piece were cut from a sheet of masonite I happened to have on hand. A round piece of wood was glued onto one end of the coil former which was then fastened underneath of the "motherboard" with a tiny screw from top side.

A close-up of my circuit layout. In order to have a large ground plane, I ended up gluing ordinary kitchen aluminum foil to the bottom side of the masonite sheet. The component legs going to nodes, other than ground, make connections through PCB islands superglued to the foil.

All radios of that era had large, intricate air variable capacitors and big bakelite knobs to turn them. Modern radios have varactor diodes to tune tanks, but in 20's and 30's these were nonexistent - this being a throwback set, it naturally needed a capacitor like that. Once virtually universally available, they have become increasingly difficult to find. Because of my somewhat frantic state, as a result of having to compete with the Department of Sanitation, I couldn't get the capacitor out of the set the tube came from. Luckily for me, I had bought two air caps from Ocean State Electronicss a few years back. Sadly, Ocean State Electronicss is no more, they have since gone out of business. The capacitor used here has two sections of vanes, both connected in parallel.

Top view.

A ready to go glowbug. A linear high voltage power supply is on the right.

Using this little receiver is different than listening to modern superhets. Turn the regen control knob up a little, connect the power and wait for a few minutes for the tube to warm up. When slight static is heard from the loudspeaker, it's time to tune it. Find a station (there are still a few AM stations left), then advance the regen control until it's heard loud and clear. You may need to re-adjust the tuning as increasing the regen level tends to couple more of the tube's internal capacitance into the antenna tank circuit, thus detuning it. Here in New York City, with a good quality "noise-free" power supply, this little circuit gives quite an account of itself. The best station heard is Bloomberg Radio, 1130 KHz. It's heard even without the antenna although it's then faint and staticy. Connect the 2-meter piece of wire and it's as clear as with modern rigs. Tune in to hear fresh economic news (mostly disappointing) along with many other topics. After playing with this little radio for a while, I decided to call it 1-tube Bloomberg, a befitting name IMHO.

If you turn the regen level up too much, a little click and whistle are heard from the speaker. This means the first stage has burst into sustained oscillations, taking the "re" out and leaving only the "generative" part . It now radiates too, creating interference to others so PLEASE DON'T DO IT!

You have to admit, there has to be something magical about tubes. You'll never get this warm feeling when working with modern transistors and IC's. :-)

Hope you had fun following this tube project, de Brian.