In our two previous discussions of telecommunications in the 1632 series, we focused on radio communications uniquely available to up-timers (“Radio in the 1632 Universe,” Grantville Gazette, Volume One), and to wired communications (“So You Want to do Telecommunications in 1633,” Grantville Gazette, Volume Two). In this article we will discuss radio options available to down-timers both for transmitters and receivers. This will require a brief discussion of radio theory, which we will restrict to no more than one equation.

As discussed in Grantville Gazette, Volume One, up-timers are generously supplied with radio equipment. Various categories include CB radios, FRS handhelds, commercial FM radios for utility and police work, ham radios in a variety of bands and “down-time built from up-time parts” radios including the transmitter for the Voice of America. As well, there are custom military and diplomatic “cw” radios.

Down-timers want radios for a variety of reasons; people anywhere near Grantville want to listen to the Voice of America radio station. Starting in early 1635, its sister station, the Voice of Luther in Magdeburg will go on the air. People, especially governments and military types further from Grantville, want the magic instantaneous communications over great distances that radio gives you.

How can they do either of those things without tubes, or transistors, or the other things we think of when we say “electronics”?

The magic word is “detector.”

Take a telephone receiver. In a telephone circuit. current flows through the handset and is stronger or weaker according to the signal sent by the telephone transmitter. So the diaphragm in the receiver moves in and out, attracted to the electromagnet more or less depending on there being more or less electricity flowing through it.

If we have an antenna attached to a handset and then to ground, and we put the antenna near an AM radio transmitter, the transmitter makes its radio signal stronger or weaker, just like the telephone current is stronger or weaker in response to varying sounds. That radio energy is received by the antenna, passed through the receiver, and into the ground. We ought to be able to hear it. But we hear nothing . . . why? It’s because the radio waves are going BACK AND FORTH, positive and negative, many, many thousands of times per second. The oscillations are so fast that the diaphragm of the receiver just sits there. Nothing happens.

We need a way to DETECT the RF currents. The first detectors used in commercial radios were called “coherers.” Edward Branly developed the first workable ones. Imagine a glass tube filled with sharply cut nickel and silver shavings. Hook a battery and an earphone to the tube. The high resistance of the shavings prevents electricity from flowing through the tube, and you hear nothing. But, if we also hook the tube up to our antenna and ground, and there is a strong radio signal in the area, the electric field from the radio signal lowers the resistance of the shavings, and they conduct suddenly. You then hear a CLICK in the earphone as electricity suddenly runs in. If you rig a small hammer to tap the tube when electricity runs through it, this re-arranges the shavings, and you hear another click as the current goes away. If the radio signal is turned on and off (as it is, when you’re transmitting Morse code) then you hear a series of clicks in the ear piece each time the radio turns on. You also can listen to the clicking of the hammer against the glass tube (until it shatters.)

Tesla is famous for inventing a coherer which ROTATED instead of being hit with a hammer. This was considered a big step forward.

But coherers require strong signals, they can ONLY receive Morse code, and only that sent slowly enough to allow the coherer time to reset. For about ten years in the real world, they were the best we had. But they sucked, and everyone knew it.

What was needed was a way to keep the electricity from running through the earphone in BOTH DIRECTIONS. If you could cause the signal from the antenna to be “rectified” and only take the POSITIVE or the NEGATIVE half of the signal, the little pulses would “add up,” each one pushing the diaphragm of the earphone a little further out instead of causing it to wiggle back and forth too fast to hear.

Many people attempted to make rectifiers. The most successful was a sharp platinum wire placed JUST BARELY into a pool of weak acid. When the electricity went one way, tiny bubbles formed insulating the tip, and when the electricity went the other way, the electricity would flow normally. It was fussy, and worked well only for wizards, required constant adjustment and so forth, but it did work. Then, in 1906 Greenleaf Whittier Pickard patented a solid state rectifying detector. Pickard (whose name is almost unknown in the real time line, and who deserves far more acclaim than he gets) discovered that a variety of materials—if prepared just right—would rectify a signal. They would conduct electricity easily in one direction while having a resistance hundreds or thousands of times higher in the other direction. Additional benefits included that the crystal detector had no moving parts, no liquids and no glass tube.

The situation down-time is similar to the situation in the US around 1920. In the autumn of 1920, in Pittsburgh, PA, station KDKA went on the air just in time to broadcast the Harding-Cox presidential election returns. In addition to reporting on special events, broadcasts to farmers of crop price reports were an important public service, in the early days of radio.

In 1921 factory-made radios were very expensive. Many of them cost more than $2,000 USD (in 2006 equivalent dollars), and less affluent families could not afford to have one. However, in 1922 the U.S. Bureau of Standards publication Construction and Operation of a Simple Homemade Radio Receiving Outfit showed how almost any family having a family member handy with simple tools could make a radio. It became an immediate best seller. More than any other system, this design was responsible for bringing radio to the general public.

Similarly, in 1632 the publication How To Make and Use a Crystal Radio will give anyone with sufficient patience everything they need to know to build a simple crystal radio receiver. The parts list which must be purchased is astonishingly short.

  • A coil of very fine wire (for making the earphone), about 2 ounces.

  • A very small magnet (for making the earphone).

  • An iron nail.

  • A coil of larger wire for the tuning coil and antenna, about 1 pound.

  • A piece of lead ore (galena) the size of your little fingernail (the crystal).

Locally sourced materials would include thread, glue, a disk of parchment for the earpiece drumhead, a wooden cup, a board, and a coil form (toilet paper tube) wound from paper and glue to a size drawn in the printed instructions. (The traditional up-time forms are toilet paper rolls and oatmeal boxes. It is presumed that down-time hand-made toilet paper tubes will be fabricated.)

Home made earphones are possible, as demonstrated by thousands of hobbyists, but commercially prepared earphones especially sensitive piezo-electric sets will be available inexpensively quite early (See “Dr Phil’s Aeolian Transformers” in Grantville Gazette, Volume Six).

The circuit’s sensitivity and tuning ability is improved by the presence of a capacitor, which will require the purchase of a sheet of copper foil. Beyond that, mere careful measuring and careful assembly will almost certainly result in a working radio.

Purpose built coil forms, ceramic forms for wrapping toilet paper tubes, purpose built glass and ceramic antenna insulators, lightning arrestors, pass-through tubes for walls, and such will be available for purchase from electronics dealers soon after VOA goes on the air. But even the poorest village should be able to afford a single crystal radio, the cup forming the earpiece being passed from hand-to-hand as programs change.

Crystal sets as described have several disadvantages. The largest is that only a single person can listen at a time. Also, you can’t rig multiple receivers to a single antenna, so a village wanting to have several people listen at once would end up with a forest of copper hanging over their heads.

The solution is an amplifier to make the weak signal strong enough to hear. There are several amplifier designs available which do not require tubes or transistors—which is good, since we don’t have any to spare. They are all horrible compared to the cheapest single transistor amp, but they will allow a group of people in a quiet room to hear a radio at the same time.

The simplest amplifier is to place the earphone at the base of a trumpet. If you think about the classic Victrola record player with its flower-like horn, the weak thready sounds coming from the needle were transformed into a room-filling sound by expanding the waves in a hyperbolic horn. This can be done with any earphone and a properly built trumpet. The same company which makes record players will doubtless produce adaptors for earphones.

A more robust solution is the “telephonic amplifier.” Take an earphone, glue a telephone microphone to it, and use a powerful battery to power the output into a telephone receiver. A stack of four telephone circuits like that can amplify a crystal radio eight to ten times, the final output going into a loudspeaker or into the base of a hyperbolic horn.

The best non-electronic down-time amp is the selenium photo-voltaic amp. Take a thin foil of selenium (which you will have to find, purify and stretch into thin foils. The more light that hits it, the less its resistance to electricity. Hook a battery and loudspeaker up to the selenium foil (Okay, it’s not that simple but it can be made at home if you have selenium.) and then, use a mirror moved by your crystal earphone to focus more or less light on the selenium. A very substantial amplification is possible. Alexander Graham Bell first patented this as a “photophone.”

So, almost immediately after the Ring of Fire, every village in the area has gotten a broadsheet with instructions about building a radio, there are radio companies making deluxe sets and amplifiers, and it is the rare tavern which doesn’t AT LEAST have a “radio room” for listening to VOA and probably is trying to save up for a “Real Up Time Radio.”

That takes care of the receiving part. What about transmitting?

Sadly, transmitting radio is MUCH harder than receiving it. To receive a signal you only have to catch that tiny, tiny bit of a signal that hits your antenna. Sure, your antenna may be 100 feet long strung from house to house, but still, all you’re catching is that little itsy bit that hits your antenna. On the other hand, the transmitter had to fill all of space with that much power. Your itsy bit, your neighbors itsy bit, the bit over there by the horse corral that doesn’t HAVE an antenna in it, the space down by the creek, and so on. You’re catching a snowflake’s worth of power, but the transmitter must create the snowstorm.

Also, we have to decide what KIND of transmitter we want. Do we need a transmitter to send voice and other sounds, or is it enough to tap out Morse code and send messages? Code transmitters are much easier than voice transmitters, and we’ll talk about those first.

Consider a bell. You tap a bell with a hammer, and the bell “rings” for a while. The harder you tap, the louder the initial sound, and the longer the ringing lasts, but eventually the energy from the tap dissipates, partly by heating the bell, and partly by transferring energy to the air making sound. You can send Morse code by tapping a bell. The size of the bell controls the loudness and the pitch or note of the bell. A big bell is louder. A big bell sounds a lower note. You can raise and lower the note of a bell of a given size within limits by making the shell thicker or thinner, and by using stiffer or more flexible materials, but nothing you do is going to make a two inch bell sound like it is a three foot bell.

Radio waves, (and light) are waves with two parts. There is an “electric field” wave, and a “magnetic field” wave. They are related and create each other. When you wave a magnet over a coil of wire, it makes electricity move in the coil. If you move electricity through a coil, it makes a magnetic field. Radio waves are electric and magnetic waves creating and supporting each other as they travel through space. They oscillate up and down, back and forth, similar to sound waves in air. If you take a coil of wire (which you remember causes a magnetic field to be created when it has electricity running through it) and a capacitor—a device for storing electric fields—and hook them together, they form a “resonate circuit” which will ring, just like a bell rings when hit. If you put a pulse of electricity into the circuit, it goes round and round the coil, and makes a magnetic field, then it gets stored in the capacitor, and as the magnetic field collapses it causes another pulse going the other way which bleeds out the capacitor and charges it the other way. Bigger capacitors, and bigger coils change the resonate frequency of the circuit just as bigger walls and bigger diameters change the pitch of the bell. (If you’re looking for an analogy for the stiffness of the material, I’m really stretching an analogy beyond all limits here, but you could think of the tightness or diameter of the coil as an equivalent and I wouldn’t be upset. This is an analogy, all right? I’m not giving you equations, be happy.)

Back to the bell. If you graph the loudness of the bell after you hit it, the loudness decays away fast at first and then slower and slower . . . the graph tapers to a point, shaped sort of like a ski jump. Similarly if you hit a resonate circuit with an electric pulse, it rings with RF, tapering in a very similar shape. If you hook the circuit up to an antenna, the RF is sent out into space, just like the sound is sent from the bell. It’s very loud at first, and then quickly (in much less than a second) the radio energy is used up in heating the coil and transmitting radio waves out into space.

Just like the bell, the harder we hit the circuit, the louder it rings (until we hit it hard enough to make it melt). So, to make a Morse code transmitter, we put up an antenna, we build a big, strong resonate circuit out of a coil and a capacitor, and we connect the coil to a powerful electric power source very briefly so that the circuit “rings” . . . we do that over and over, each “tap” of electricity making a ringing, and we can send Morse code by controlling the timing of the taps. More electricity corresponds to “harder taps.”

I can hear it now. “What about the sparks? Don’t you need sparks?” That would be “yes and no.” The function of the spark gap is to cause there to be a very high resistance in the circuit which allows the capacitor to charge. When the capacitor is charged “enough,” the sparking voltage of the gap is reached, and the spark gap “sparks.” This causes there to be a lower resistance in the circuit causing the capacitor to discharge. The discharge through the conducting spark takes the form of a damped oscillation, at the frequency determined by the resonant frequency of the circuit. If you could make a switch that did the same thing WITHOUT sparking, the system would still make radio waves. The spark gap acts as a voltage dependent switch. STOPPING the spark is as important as starting it. The eventual solution was a “rotary” spark gap that broke the spark by pulling the contacts apart as a central disk rotated, and then lined them up again as the disk rotated the next stud into line with an unmoving contact.

So, the parts list for our down-time Morse code transmitter is considerably longer than our crystal radio. We will need:

  • An antenna cut for the frequency we want to transmit on.

  • A coil wound from heavy wire with careful spacing, thick enough to be self-supporting in air because we don’t want anything shorting out the coils.

  • A large capacitor to match the coil. Stacks of glass interlayered with gold foil—or better, sheets of mica interlayered with gold foil.

  • A large copper rod or copper plated iron rod driven into the ground.

  • A source of electricity—a bank of the same type of batteries we use to run the telegraph will do. We will need a large number of batteries (perhaps as many as one hundred gallon sized batteries) to make a powerful signal.

  • A “spark coil.” This is a pair of coils wound around a common center. The first coil, which is attached to the batteries, has only a few loops (four or five). The second coil, attached to the circuit, has MANY loops, dozens certainly, so that our transmitting voltage is several thousand volts.

  • A switch that turns the electricity to the coil on and off. Each pulse through the buzzer makes a pulse into the resonate circuit. A door-bell buzzer like arrangement is fine, but it must be scaled up to handle heavier currents.

  • A Morse code key that allows us to turn the electricity to the buzzer on and off as we need to transmit.

  • A rotary spark gap.

  • A small electric or other motor to rotate the spark gap.

Everything must be very heavily constructed. The high voltages are dangerous and will eventually break down almost any insulation. If the capacitor breaks down, the entire device is likely to melt. Smaller units, with the same basic design can be made semi-portable, but to achieve any distance you need large currents and high voltages. The problem is that the spark transmitter spreads its power over a broad band of radio frequencies. A typical spark transmitter has an efficiency forty to one hundred times less than a modern CW transmitter, so a one-watt portable battery operated CW transmitter made by up-timers would be equivalent to a 100 watt massive spark station in terms of how far away it can be received. Spark stations handling multiple kilowatts of signal were very common prior to the development of tubes. In addition to all that, they are loud, dangerous and give off large quantities of ozone which can damage the operators lungs.

Also, spark stations use up bandwidth. Each one sends out a very wide signal so far fewer can “fit” in a given piece of spectrum.

There is one situation where this can be an advantage. If you are wanting to jam reception of an up-time signal, a spark transmitter close to the receiver can effectively “splatter” over the band and block the transmission.

So, spark transmitters suck. But they suck LESS than pony express riders or building hundreds of miles of telegraph. They will be used, and “king spark” will have his day until down-time tubes come into production. Once that happens, just as it did up-time, spark will be legislated out of existence.

Of course, in addition to the other disadvantages, spark is only good for sending Morse code. No one is going to curl up by the fireside to listen to the evening news via Morse code. Reaching the mass audience requires transmitting voice. Spark transmitters just can’t do that; we need something else.

There are two candidates for “something else”: the Poulsen arc and the Fessenden/Alexanderson alternator.

To discuss the Poulsen arc, let’s go back to Ol’ Sparky. Remember how it works? We present a high voltage to the capacitor, it charges, eventually it is charged, the current rushes out of the capacitor through the spark and the circuit “rings.” What if instead, we treated the spark a little differently? Instead of pulsing the current into the circuit, use a very high voltage DC current. As long as the capacitor is charging, there isn’t enough voltage to spark the spark. But once the capacitor is charged, the spark goes, and the capacitor drains . . . but this lets the capacitor start charging again, which pulls voltage off the spark, and the spark stops. (Aside to the electronics types. No, I’m not going to talk about negative resistance and LC circuits.)

It is very easy to make a singing arc like that in audio frequencies, but when you try to re-design the system to run in radio frequencies, problems begin to appear. Residual ions stripped of the electrons by the high temperature of the arc “hang around” between the poles of the spark and make the stopping voltage unpredictable, as well as the starting voltage. So, as you try to increase the frequency, the spark’s start and stop “jitters” and you can’t get a reliable signal. Much above audio frequencies, it doesn’t work.

In 1902, a Dane, Vlademar Poulsen realized that he could use a magnetic field to “sweep” the ions out of the way, and if he used a hydrogen atmosphere instead of air in the gap, the ions would be light and easy to remove. Poulsen was able to get his arc up into radio frequencies.

Poulsen arcs are big, messy, complicated devices with moving parts and plumbing. They require a constant supply of high voltage DC current produced by a big generator, generally run by an electric motor, which is ITSELF run by another generator which is run by a steam engine. They require a continuous supply of hydrogen gas (or you can use vaporized kerosene, but it isn’t as good). They need large rotating graphite electrodes, water cooled copper electrodes, a bronze chamber to hold in the hydrogen around the arc, and big “sweeping” magnets around the bronze chamber to remove the offending ions. All this, to get a radio signal.

But how do you modulate it? How do you take the signal and make it talk? The most common solution was to put six carbon microphones between the arc and ground. Speaking into the mike would vary the resistance and that would change the amount of current flowing into the ground and into the antenna. If you want to play a recording instead of talking live, you will need to set a speaker in front of each microphone. Actually, that would probably be better for the announcer too, since it would allow him to be at a distance removed from the arc.

The other something else was developed in 1903 by a Canadian, Reginald Fessenden, working with an engineer from General Electric, Ernst Alexanderson. By 1916, Fessenden and Alexanderson had developed a mechanism which allowed reliable voice transmission across the Atlantic. How? Let’s go back to first principles.

Take a coil of wire. Attach the coil to a meter. Nothing happens, the coil just sits there. Now, get a magnet. Stick the magnet into the coil. As the magnet goes in, the magnetic field of the magnet pushes on the electrons in the coil and they are shoved around the wire. The meter flicks to the right a little as long as the magnet is moving in.

Now, pull the magnet out. The magnetic field is pushing on the electrons the other way and they are shoved around the wire in the opposite direction, and the meter flicks to the left. It you put the magnet just outside the coil, and wave it from side to side, the same thing happens, as you get closer to the coil, the meter flicks right, as you get further away, the meter flicks left. If the magnet sits still, no matter how big the coil, no matter how strong the magnet, nothing happens. The magnetic field, and the electrons, just sit there.

Now . . . you can make electro-magnets much stronger than any permanent magnet. So, if you replace the bar magnet with an electromagnet, you can make BIG pulses of electricity. This is what is done in an alternator. There is a spinning coil, and a static coil and as they spin the electricity pulses back and forth. Neato! If we could spin them fast enough, we would get radio waves The problem is, the LOWEST frequency radio waves that work well for voice are at 100,000 cycles per second. If you think about trying to spin a large coil of wire 100,000 times per second, you’ll realize just how hard it would be. High power router motors spin as fast as 24,000 RPM, but that’s only 4000 revolutions per second. If the router motor is 3 inches across, the outer edge is moving at 214 miles per hour, and is experiencing a pull of 513 times the force of gravity.

Take that same motor, and try to use it as an alternator, and spin it up to 100,000 revolutions per second, and the outer edge is moving 51,000 miles per hour (far above earth escape velocity, much faster than a speeding bullet) and the outer edge of the coil is being pulled with 321,000 times the force of gravity. The wire will simply fly apart long before. The fastest spinning man-made object (a carving tool similar to a dentists drill) turns at 450,000 rpm, or 7500 revolutions per second, thirteen times slower than we need for radio. The router WOULD give us reasonable power, since it handles 5 hp, around 3700 watts. But it just won’t work.

What to do?

Let’s go back to our two coils. One is a powerful electromagnet, with a DC current running through it. The other is a coil. They’re just sitting there. Nothing happens. Now, place a hunk of iron between the magnet and the coil. As the iron comes into the field, the field seen by the coil decreases, and the meter flicks left. As the iron is pulled out, the field increases, and the meter flicks right. Cool! Of course, we have to build a strong, strong mechanism to spin the iron into and out of the field, and we have to cool it, since passing in and out of a magnetic field like that will heat metal like crazy. But still, we can spin a hunk of iron instead of spinning delicate coils and wires. Cool!

(I am oversimplifying here. Don’t shoot me.) Take a big strong iron disk. Drill a series of holes around the edge and fill them with bronze. Now place the disk so that the bronze holes are lined up in the space between the magnet at the coil. Spin the disk. As the bronze window comes between the coil and the magnet, magnetic fields “get through.” As the bronze window moves away, the iron interferes with the magnetic field, and the magnetic field “does not get through.” We now have an alternator in which the coil and the magnet do not move. The disk can be built VERY strong, encased in vacuum and water cooled by pipes running through the center. Take a disk 4 feet across, put windows every half inch around the edge and you have 300 windows. Rotate the disk 330 times per second or 20,000 rpm, and you’ll get 100,000 waves per second of RF power.

This isn’t EASY, but it turns out to be within the ability of 1902 mechanical engineering. (Well, not really, they spun it one fourth that fast and used two frequency doublers. But you really don’t want to know about frequency doublers. )

Use multiple pairs of magnets and coils spaced equally around the disk to increase your power.

Modulate it the same way you did the Poulsen arc, with six microphones in series between the transmitter and ground.

The Fessenden/Alexanderson alternator was the mechanism for radiotelephone prior to the invention of tubes. One working station remains in service in 2006 in Sweden. It has been named as a world heritage site and is run on special anniversaries.

So, there you have it, radio for everyone else. At least until the research teams manage to start building tubes again. But that, as they say, is another story.


I could give you a long list of reference sites, but frankly the easiest is simply to visit Wikipedia at and search for crystal radio, Poulsen arc, and Fessenden. The explanations there are good, and their reference links are constantly updated to working sites.