Beginner’s Transmitter Building Saga

A few months ago I got a shortwave radio with side band reception and a digital readout and it opened up a very interesting realm of the radio waves: the open waters that radio pirates hijack and sail upon. With their antennas raised high like the mast of a ship, they risk their butts to give you…a radio show!

I learned that some pirates make their own transmitters. And I found that its about the only way to transmit besides an official transmitter that is expensive. They also need to be kind of unlocked to use the pirate frequencies that are out of the legal transmitting band allocated to licenced ham operators. With my interests raised, another obsession began (I think I have a hobby of trying different hobbies. Its either this or I get obsessed easily.) The goal was to make a transmitter. So I found multiple plans for radios but the work required in learning electronics terminology and parts was too great. I use an interest-to-energy-required ratio. I knew nothing of electronics besides wiring up car stereos. “I dont know any of this and I don’t want to research a bunch of stuff because I’m too lazy. How do I know it will even work?”

Then I found that had plans for a simple transmitter. And its simplicity made the interest-to-energy-investment ratio favor the project. A friend had most of the parts laying around to make it: the LM386 integrated circuit, the 2222 transistor. What a coincidence, its my destiny! I just needed a crystal and a variable capacitor. I was also given a small bread board with wires. I am indebted to my generous friend!

label parts

3 stages

First I made the LM386 audio amplifier to power a small speaker and it e-f-f-i-n-g worked! And I smelled something hot—the LM386. It felt HOT too. So the next day I learned from a coworker the electricity just flows as fast as it can unless you throttle it. It went through the LM386 too fast and made it dangerously hot. You can start a fire with this stuff. Electronics: manly toys. It flowed through the speaker too fast so I put resistors between the amp and the speaker and that slowed down the current and worked. I held everything together with 2 hands surprisingly. Just enough to make contact and see it work.

lm386 circle circuit

Then I made the transmitter. Its coil was ramshackle, made from the enamel coated wire of an old RC car motor. You just coil it around the correct number of turns on a form the correct diameter. And I ordered the 2 other vital parts, a crystal and a variable capacitor. And IT effing worked! Eff yeah! It made morse code beeps on the side band mode of my receiver! So 2 for 2.

xmtr circle circuit

Then I combined the amp and the transmitter and I had music coming out. Its easy to combine them, its just one wire from the amp to the transmitter! That makes sense! It was kind of low audio but it had promise. What if it worked better with all the parts soldered up with a pretty looking coil? So I ordered parts online. I made sure to buy the capacitors that the schematics recommended for the amp because I wanted good audio quality.

LM386 datasheet. An integrated circuit has a circuit of transistors and resistors inside of it, but at a microscopic level. This LM386 integrated circuit has a sheet that describes what it does. And I was able to generally understand it! It wasn’t all obfuscated with geek jargon. Its a product and they WANT people to be able to understand what it can do so they’ll buy it. This particular IC is worth producing because lots of people want a little amplifier for their projects. Look at the size of my metal can transistor. Its a little bigger than a pencil eraser. Can you imagine they fit transistors in there?

Then I drew out the wiring. I chose to use a socket for the transistor and crystal because I figured they might get hurt and need replacing. Its a good thing because they go bad later. I got the idea for a socket from a portable AM radio from the 60s. I wound the filter, its not that hard and kind of fun. The filter makes it so you transmit over a small area of the spectrum. Otherwise you will transmit all over the spectrum and you might transmit where somebody needs it like ambulance drivers and their dispatch. You can cause an ambulance driver to drive off a cliff without filters! Just kidding about the cliff. Winding is menial. I believe humans were created to do menial tasks like grooming, carving, and preparing food. Knitting is great! But I digress.


close xmtr filter

Lo and behold the official thing all soldered up. It looked great, just like makerf’s. But did it work? Yeah it did. I don’t remember being all “eff yeah” about this one. I think it was more “yeah, you better effing work.” I spent a lot of hours on this. About 2 full days worth of drawing, forming wire, and soldering. I adapted a 12v battery pack that plugs in to turn it on (unplug to turn it off). A switch and LED indicator light would be nice. The great thing about making it official was the ease of use. I didn’t have to hold everything together with tape and clothespins. I could turn it on and leave it now. My hands were free to turn the capacitor for tuning.

official pic

A friend loaned me an oscilloscope. The stuff you only see on TV. I was a bad-ass laboratory doctor now baby. I still want to use it in the garage with the door open and the screen facing out so the neighbors can see how much of a laboratory doctor I am….Turning the capacitor made a difference. The sine wave wasn’t very siney. I wish I made this when I was taking trigonometry. Prior, I found no use for trigonometry whatsoever. Good thing I took it twice so it stuck a bit more than the other useless crap. Too bad chemistry didn’t stick after taking it twice. Yes I’m bitter after obtaining a college degree and feeling the lack of good jobs available.


So I would “tune” it (twist the cap and watch the waves change). Nothing looked very “clean”. When I did this I also had the receiver on side band mode and the morse code beep would change tone as I adjust the cap. I don’t know what “clean” means for sure but I think it means a smooth wave with no sharp jagged points on it. Both the crystal and the transistor would get HOT sometimes. But it kept on working so I figured no harm done. Sometimes I didn’t have 50ohms worth of resistors between the antenna’s + and – output which is a no-no. Transmitting antennas have this same 50ohm property. I guess this throttles it, otherwise too much energy will go through the transistor and get it hot like it did my LM386? I don’t know. Use a non-conductor to twist the capacitor. I use a carved wooden stick. A metal screwdriver will change things when it gets near the transmitter. Just my finger will change the sound of the beep it makes as I move it around the transmitter. Pretty cool.

Here’s the first sine wave hooked up to ground and antenna out:
first sine wave

Here’s another with the oscilloscope probe touching the positive antenna out lead: (you can watch as a song with base really change the amplitude as the song plays)
latest sine wave

Tuning observations: I twist it until it doesn’t work. Then I slowly twist the other way and a nice small sine wave shows up. The audio sound is weak. Then continue twisting slowly until it jumps and gets jagged and the sound is louder and fuzzy. A little more and the sound is as accurate as it gets. Any further and the treble is lost. And if it is overmodulated I turn down the volume on my audio device.

While turning the capacitor I accidently flicked the crystal and the transmitter stopped working. There’s an actual piece of a quartz crystal inside a container. Did you know that if you compress quartz it creates a small electrical current? The conspiracy theorists aren’t so crazy now are they. In reverse, if you add electric curent to a crystal it will vibrate. Oscillate is the way the laboratory doctors describe it. And depending on the thickness of the crystal it oscillates a certain number of times per minute. So 6.925mHz oscillates a different number of times per minute than 6.935mHz. Anyhow, thanks to the hfunderground forum members I learned that I probably needed a physically bigger crystal. Its called an FT-243 style and mine looks like its 100yrs old and was exumed by Indiana Jones from a cave. I had originally used a small crystal meant for computers. It would do well for morse code. But its not cut out for continual use apparently. Or being flicked while hot and oscillating.

So it worked once again. Meanwhilst I installed a computer processor fan to the transmitter case to keep things cool. Otherwise transistor gets hot when the transmitter is on. The fan works great but if I wire up the fan to the transmitter power supply it makes a clicking sound in the received transmission! So I hooked up a 9v battery to the fan so it has its own separate power source for optimal fidelity.

pic with fan

Speaking of optimal fidelity, I didn’t like how the sound is pretty quiet compared to other stations. Its probably my coil because its not perfect. (I’m just joking, I think the coil doesn’t really have to be that perfect because my ramshackle coil worked ok). If I turn up the sound more than half way on my .mp3 player it gets all nasty sounding and its called overmodulating. You’ve heard it before. It kind of sounds like when you turn the volume up on a sound system past the point that your speakers can handle it and it gets distorted. But it happens when the volume is not very high. Well, my LM386 amp puts out a small 0.5Watts of power. So would it sound better if I had 0.75W? Well, I removed my LM386 and temporarily used a 20W amp in its place. I chose the left channel, put the black wire to ground and the red to the transmitter. Yup. But you see, I started with the volume all the way down and I slowly turned it up. But the sound was still low and nothing bad was happening so I turned it up more and then it started sounding pretty good but I smelled burning. And it stopped transmitting. And my 20W amp stopped too (it has built in protection and it still works). My transistor smelled like the culprit, but wasn’t that warm cause the fan was on it. Anyhow I was up an running with a spare I had just in case this happened. Good thing I put a socket on there. The thing is more powerful now! I think my original transistor got weakened from the heat I put it through so I’m glad I broke it and got it out of the way. However the sound is still quiet, it just transmits sound farther. Like out of the room…into the yard. Eff yeah! Then I plugged my computer sound output into the transmitter and it was louder! Not perfect but better than the .mp3. So after tests, .mp3 and discman are quieter than computer and laptop outputs. Hmmm.

How the transmitter works. 3 parts:
Audio amplifier/modulator: The .mp3 player goes into the amplifier. This makes it strong enough to send to the transmitter.
The transmitter: works by magic. Seriously, I don’t know how it works. I know the recipe for it is on . The crystal makes the frequency. The current changes direction in the coil really fast…I’m pretty sure…
Antenna: I know that the antenna works by induction. If you have electricity in the air, a wire will pick some of it up. So you have the coil with the oscillating energy with music data added to it. Wrap a wire around the coil and it will pick up the power. Put the filter between this and the antenna. Send it on its way through the filter and out to your transmitting antenna, into the atmosphere where it bounces off of special layers of ions back to the Earth a hundred and more miles away. I heard Border Hunter radio the other day and they are in Europe! You know he uses one of these babies…not really. His has more magic.

If its working and then quits, I think the only variables that break are the LM386, the crystal, and the transistor. Or heat has melted the solder between joints. My interest grew as I made this. This growing interest fuels itself so I am willing to expend more energy into working on it. Its even more important than my internet addiction. I’m talking about general time wasting on the internet that just wastes the time. Really, how important is a blow dart gun video? (note to self: google how to increase night vision) Now I have an actual book about antennas that I’m actually reading. I can’t transmit without a good antenna otherwise I might burn out a transistor. Wow, I just realized I can put that notch in my belt. There are those that have blown transistors and then there’s everybody else. I plan to get a ham radio license and test this baby out on the air. If anything, the morse code will transmit farther than the audio. I learned morse code from an iPhone app. Use the internet for good, not for time wasting blow dart gun videos.

From start to finish my project took me 4 months and I have a part time job and an internet addiction to fight.

I will share the total price (shipping included):
small crystal $19
magnet wire $16
audio jack $6
radio parts from mouser $19
toriods and antenna jacks (not used yet) $15
more transistors $8.50 (shipping was about $6)
big crystal $12.50
crystal socket $13

Total: $109 (Dr. Phil’s audience gasps and whispers)
This doesn’t include perf board $6 and LM386 a couple dollars. And the effing time and energy I put into it $200 (more gasps)

It was worth it! :) I am proud of it. 8)

UPDATE June, 2014

Well, with my newly acquired ham license, I finally tried my transmitter with 2 antennas: an adjustable inverted V and an inverted L with a ground radial. I tried both antennas through an antenna tuner. As well as the inverted V without a tuner (because it was resonant on the frequency I wanted and therefore theoretically had a good impedance and thus no need for a tuner).  Well, it hardly transmitted out of my yard.

I think the reason could be:

-that my circuit wires are too long (they rout down to the socket with the transistor and back up)

-that my coil could be tuned better

So if I were to invest the time and effort to make this go farther I would first try to shorten the circuit wires (solder the transistor directly and don’t use the socket). And second, find out the resonance of my coil. I think a noise bridge could be used for this but I’m not sure.

Automating Loop Design

Spring is here finally (tra-la! tra-la!) and it’s time to get back into the yard to repair Winter’s damage and install new antennas. In particular, I want to possibly upgrade the Resonant Loop I am using for 90% of my DX time. In the past, this has generally involved intensive and error-prone hand calculations, and I was looking for a way to automate the process, or at least portions of the process. Thus, I wrote a bit of code to help streamline the design and evaluation, and hopefully help to root out the best solution.

The loops I use are usually square loops built on PVC pipe frames. These loops will almost always have a pair of identical windings separated by a center gap; this gap dramatically reduces the winding parasitic capacitance an helps to extend the tuning range. The code I wrote assumes a center gap of 3″ (which is typical for construction on 3/4″ PVC pipe) and a winding pitch (gap between adjacent turns) of 1/2″, but these values may be easily changed for exploration into other design areas.

One of the more difficult problems is computing the loop inductance as a function of the number of turns and loop side length, so the first part of the code does this for me. The code generates an array of inductance values which represents the computed inductance for a loop with turns counts of 2, 4, 6, 8, 10, and 12 turns, and side lengths of 0.25 – 1.125 meters, and then displays the results graphically. These curves may be of use to anyone who is contemplating building a loop.

Inductance Curves

The curve rendered in red represents a side length of 0.25M, and the black curve represents 1.125M. These curves represent the solution array of inductance values. One of these values might be appropriate for my loop design, but which one? To select an appropriate inductance value, the code needs to know about the intended loop tuning range, and also the capacitor used to tune the loop. From that info, it can compute the winding inductance which is required.
If you scan horizontally along any horizontal grid line (along any given value of inductance) you will note that there are several turns-side length value pairs which could generate that inductance. The question becomes: “Which is the best or most appropriate for the design goal?” Which begs the question “What is the design goal?”. Resonant loop design is the art of compromising performance in one area against several other areas. In this code, the design goal is kept simple: from the turns-side length pairs which generate the proper inductance ( +/- 20% ) we are going to select the solution which has the greatest product of turns and loop area; we are selecting for maximum Numerical Aperture.

The code then iterates through the array of inductance values, and determines if a value meets the inductance requirement dictated by our desired tuning range. If it finds a solution, it computes the Numerical Aperture ( number of turns times the side length squared), and tracks the solution which represents the maximum Numerical Aperture. When done, it presents the best solution in a dialog box, or…


The code is telling me that I should make a loop which has 2 turns, with a side length of 44″, which is a point along the red curve in the graph. This is interesting, since my current loop uses 4 turns with a side length of 20″, which is point along the Ch. 2 curve in the graph. Whatever. Knowing the number of turns and side length, the code finishes off by generating a graph of important loop parameters over the loop tuning range. These are:
- The loop Tuning Curve (capacitance vs. frequency)
- Effective Height He – the length of a vertical wire having the same output as this loop
- Figure of Merit with respect to Pattern (FOMP) – the ability of the loop to generate a pattern null

Params Curve

And that’s about it. If you are interested in the code used to generate these graphs, you can see that here:

This contains greater comment detail than is presented here.


Caught this 3/8/14
While waiting for ANTI Milano to sign on at 1457z 15515kHz, this popped up out of nowhere – NOT caught in progress:
African american male “Morning”? Then mentions of “the metropolis of San Francisco, CA”, “Macon County Alabama”, “Macon County Executive Committee”, “Alabama National Democratic Party”, “Last 3 presidential elections”, “isn’t it strange”, “27 years ago”. Then ANTI Milano.
Any one have a clue?

Radio Impala

17540kHz carrier up at 1657z, sign on 1700z. Till 1755z. 2/9/14.
Details in Clandestine section of HFU.
Heres the first few minutes and the part in English (the shorter of the two).
Hope that you like.

OFF TOPIC but I enjoy it….



Hello all,
Off the clandestine topic so I apologize but I had the itch to post some pics of weather faxes that I’ve received in the past month or so.
Being not very good with computers, I use a Universal M-8000 to decode (no giggling please). So these are pics of the monitor in my shack.
The ones labeled Alaska are from NOJ in Kodiak; Aus are from VMC/VMW Charleville & Wiluna, Australia; Chile is CBV Valparaiso Playa Ancha, Chile; and UK is GYA Northwoods, UK.
Hope you enjoy,