Adventures on the road to HB!
Homebrew Radios in the age of the Internet
By Bill Meara, N2CQR
Send me an e-mail
"I listened to the magic that only comes from a radio that you built yourself." In that one sentence (posted to an Internet e-mail group), Mike, VE2GFU, nicely described the feeling that can arise in the midst of a room full of solder smoke... and the reward that awaits those who endeavor to build their own radio receivers. In an age of mass produced, homogenized, high price commercial equipment, there is still magic to be found in the production and use of simple homebrew radios. I recently put together my first superhetrodyne communications receiver - I had so much fun with it that I thought my fellow amateurs might be interested in the project.
I was a frustrated teenage radio builder....
When I put my first homebrew low power transmitter on the air a few years ago, I thought I'd maximized my ham radio satisfaction. I gleefully reported to other stations that "RIG HERE IS HOMEBREW". For a while, I really thought that my fun meter was pegged! But everytime I looked at the commercial receiver that sat alongside my QRP transmitter, I knew in my heart of hearts that I still had some work to do. The truth was that only half my station was homebrew. Until I built my own receiver, I would not be able to enjoy the warm glow of satisfaction that comes from running a completely homebrew station. As a kid, I'd always looked with wonder and envy at the exotic homebrew stations in the DX column of QST magazine. I wanted to do what those intrepid foreigners had done! I decided to finish the job. I decided to build a receiver.
"Receivers are Difficult!!!"
I approached the project with some trepidation. Since my earliest days in the hobby I'd heard that "receivers are difficult." There seemed to be a deeply believed and long-standing bit of conventional wisdom that said that most hams could sucessfully build transmitters, but receivers were somehow beyond our capabilities. During radio club meetings, old timers would share tales of homebrew adventures from days-gone-by. They told of tube transmitters built on chassis fashioned from purloined street signs. There were a lot of great stories, but they were all about transmitters. When I'd ask about receivers, the old timers would look a bit sheepish as they admitted that their receivers were all commercial.
Receivers are difficult. I knew from personal experience that there was some truth in this axiom. As a teenager I had tried to barge into the ranks of the homebrewers with an audacious attempt at reproducing a varactor diode-controlled superhet receiver I'd seen in one of the ham magazines. I never got it to work. As I approached this recent receiver project, I think a desire for vindication - and a desire to finish the job I started in 1974 - was part of my motivation.
Barebones, no frills, one step at a time
The "Barebones Superhet" presented in a July 1982 QST article by Doug DeMaw seemed to be just what I was looking for. As the title imples, it is a very simple, easy-to-understand circuit. Most of the stages were built around discrete solid state components - no mysterious IC black boxes. As a true superhet, it would be able to deliver "single signal" performance. I'd built a few direct conversion receivers from kits but had been somewhat disappointed - direct conversion receivers usually lack selectivity and tune each incoming signal at two points on the tuning dial.
Remembering my bitter defeat in my earlier receiver project, I decided to take a fool-proof approach to this one. I took Doug DeMaw's very simple schematic and made it even simpler by dividing it up into separate stages. I would build each stage one at a time, each on a separate printed circuit (PC) board. For my receiver there would be separate boards for the Radio Frequency Mixer, the Variable Crystal Oscillator (VXO), the intermediate frequency (IF) amplifier, one board for the Product detector/beat frequency oscillator (BFO) and one audio amplifier board. I would test each stage before going on to the next.
Parts acquisition in the age of the Internet
As a teenage wanna-be radio maker, parts acqusition had been a major problem. I'm happy to report that the Internet and Express mail services have largely eliminated the tortuous "waiting for the mailman" vigils that many of us endured back in the dark ages. I kicked off my project with a brief session involving several parts catalogs, my computer and a credit card. A few short days later, the boxes started coming in and actual construction was about to begin.
While the catalog houses provided many of the parts, my junk box, hamfests and fellow hams were the sources for many of the components. I think that this diversity of parts sources adds to the character of the final product. When I look at my receiver, I can see parts that came from my old friend (now SK) Pericle, HI8P. There are components in there that were sent to me by Tom, W1HET and several other ham friends. There is a reduction drive from an old Swan 240 and a grommet from a deceased Heathkit Luchbox. The LM386 audio amplifier chip (a concession to modernity!) came out of a Kanga Kits Direct conversion receiver; I didn't have an eight pin socket for it, so I scrounged through my junk box, found a 16 pin socket and cut it in half. Like I said, this approach to parts acquisition gives the radio some character!
Lunch time PC board design
My "one stage at a time" approach resulted in some special challenges and opportunities. I had to design the PC board patterns myself. For hams accostomed to using ready-made PC boards, or simply reproducing patterns made by others, this might seem like an intimidating task, but since I was dealing with only one stage on each board, it turned out to be easy and rewarding. I was using boards that fit very conveniently in the front pocket of my shirts. I made PC board design a lunch-hour project. I would go to work with my schematic and a couple of index cards in pocket. I'd cut the cards down to PC board size and used them to plan the layout of the boards. I usually had to do two or three "drafts" before I was satisfied, but I found that I was able to do about one board per lunch hour. Doing the layout myself definitely added to the "I did it myself" feeling at the end of the project.
I set a goal of completing one board per week - most of the design and planning would take place during the lunch hours, most of the construction took place early on Saturday and Sunday mornings.
My arsenal of test gear is far from laboratory grade! I have a little (ancient) Eico 435 oscilliscope and an old Heathkit signal generator. I bought the scope for 25 dollars on the Internet. The generator was a 15 dollar hamfest purchase. The 'scope will only read up to about 5 Mhz, but since the IF of my receiver would be 3.579 Mhz, I knew it would be very useful.
Testing the stages was a lot of fun. The VXO and BFO were easy to test - I just listened for the signal on a Radio Shack general coverage receiver. For the IF AMP I used the signal generator to put some 3.579 Mhz energy into board and used the 'scope to make sure it was amplifying.
One of the most difficult parts of HF superhet construction is the IF filter. Doug DeMaw's circuit employed a three crystal ladder filter. Doug described it as simple and easy, but to me it looked a bit intimidating. One of the benefits of homebrewing is that you can really "have it your way". Wishing to avoid a frustrating battle with a complicated filter, I searched through the QRP/Homebrew literature for a simpler approach to IF filtering. I found what I was looking for in another article by DeMaw. In this cicuit he used one crystal with a resistor to ground. I decided to use this simple filter and put off construction of the more sophisticated (and narrow) three crystal circuit until later.
So I redesigned the RF mixer board to accomodate my simplified filter. I wasn't quite sure if this little foray into electrical engineering would be successful (my degree is in economics!) so the testing of this stage was tinged with some anxiety. I set the signal generator for the low end of the 20 meter band. I got the VXO oscillating and put the scope on the output of my simple filter. Slowly I tuned the generator across the 20 meter CW band. All of a sudden, at one very specific point, a big 3.579 Mhz signal popped onto the 'scope screen! Eureka! My mixer was mixing and my filter was filtering!
Holy cow! It really works!
After about a month and a half of this, I had assembled an impressive looking collection of small circuit boards. I couldn't resist putting them all together on the workbench to see if this thing would really receive. Armed with a set of alligator clip test leads I connected inputs to outputs. It was early in the morning and 20 wasn't really open yet, but it was Saturday and I figured there were some folks out there trying to coax the ionosphere into action. As I was checking the test leads, I started to hear - almost imperceptiably at first - CW. At first I thought the sound was coming from my Drake 2-B, but a quick check showed the Drake was completely off. My little creation was actually receiving radio signals!
As late afternoon rolled around I decided to see how my still incomplete device would handle SSB. As luck would have it, my crystal let me tune around 14.200 Mhz. There I found the very melodious tones of EA3OT. Echo Alfa Three Oooold Timer, with his "six over six over six" antenna system filled my shack with beautiful phone signals. My relatively wide, one crystal filter was ideal for reception of Mike's fine signal. There really was something quite magical about looking at my little collection of boards and realizing that they were receiving signals from far-away Barcelona. I was experiencing "the magic that only comes from a receiver that you built yourself..."
Enclosure (sort of)
Now it was time to start putting the radio in a proper enclosure. A few years ago, Paul Carr, N4PC, * (*Described in several editions of the CQ magazine during 1993) built a 40 meter solid state rig on a wooden base. Disliking metal work, I immediately appreciated the wisdom of this approach. Realizing that I'd probably want to add additional circuits later on, I decided to make the chassis about twice the size I really needed. A visit to Home Depot yielded a suitable (16"X11") piece of pine. I also picked up some very light sheet metal that I thought would help with the front panel.
I had three large double sided PC boards in the junk box. The circuit boards were attached to these PC "base" boards with some Radio Shack spacers and 4X40 screws. The Base boards were bolted to the wooden base.
I used the sheet metal to fashon an L shaed front panel. The material was not quite rigid enough, so (in keeping with a very old ham tradition of stealing radio materials frm the kitchen) I put a little "cookie baking sheet" between two layers of the Home Depot sheet metal. The L shaped panel was afixed to the pine base. A smaller L shaped piece of scrap aluminum was attached to the back side of the pine chassis - this would serve as the mounting point for the antenna and power connectors.
My creation was starting to look like a radio. Better yet, it resembled one of those impressive homebuilt rigs that I used to see in the DX column of QST. I felt I was getting close to membership in the that elite group of intrepid hams who had actually "rolled their own." I was starting to feel a kinship with all of those intrepid, creative wackos who build things in their basements or garages. I felt part of the same homebrew tradition that dates from in the early days of ham radio. Just like the guys who build small airplanes in their backyard shops, just like those guys in California's Homebrew Computer Club, I was approaching the point when I could begin sentences with the proud phrase, "I built..."
But of course, I was not done yet. Not by a long shot. When you are homebrewing, you have to be patient. You have to start out realizing that you are definitely not involved in "plug and play" radio! Very few homebrew receivers will work properly the first time you fire them up. The radio needs to be properly aligned. Amplifiers and oscillators need to be tamed. But I think this is one of the most satisfying part of the homebrew experience. It is during this phase that you really get the sensation that you are molding your creation to satisfy your requirements. You are physically molding it by deciding where you want the control knobs and external connectors. And (even better) you are molding it electronically by deciding how you want to to sound. It is during this phase that you really put electronic theory to work.
I had a few very common problems. My audio amplifier would scream like a banshee if I turned the gain up. My Variable crystal oscillator was kind of sluggish - it sometimes wouldn't start up right away when I applied power. A preacher from the 22 Meter broadcast band urged me to repent every time fired up my new radio. And worst of all, 80 meter CW signals from the venerable W1AW jumped right over my receiver's front end filters, landing right in my 3.579 Mhz IF frequency. These signals not only appeared to be mocking my technical abilities, but they also seemed to be making fun of my code speed! Like I said, this was definitely not plug and play!
In my effort to fix these problems, modern technology provided me with resource that was completely unavailable during my earlier (1974) battle with a superhet: the Internet. The 'net puts the radio builder in almost instantaneous contact with a worldwide network of entusiastic solder melters. I found the rec.radio.amateur.homebrew USENET group to be an excellent source of information, advice and moral support.
The internet can turn your homebrew project into a multinational enterprise! Hams from around the world chimed in with helpful hints. It was a lot of fun to encorporate suggestions from distant Australia into my little HB receiver. And it was very reassuring to know that all those far-flung Elmers were available if I got into a real jam.
I was particularly gratified when I got some e-mailed words of encouragement from the guy who had designed the receiver I was building, Doug DeMaw, W1FB. Doug's son had spotted one of my pleas for help in one of the USENET groups and had relayed my message to his father. Doug sent me a very nice and encouraging note. I was saddened to learn that shortly after our exchange he became a silent Key.
Solutions to most of my problems came very quickly - and I learned something with each of them.
The screaming banshee audio amp turned out to be the result of a simple circuit error - I'd failed to ground one of the bypass caps on the LM386 AF amp chip (the only IC in the rig).
The Variable Crystal oscillator was made more obedient by playing a bit with the values of the two capacitors that madeup the feedback network in the Colpitts oscilator.
The preacher and W1AW required a little more effort. I decided that I needed a bit more filtering at the front end of the radio. I could have easily just thrown in one or two more tuned circuits between the antenna and the mixer, but I was concerned that losses in these circuits would adversely affect receiver sensitivity. Roy Lewllan, W7EL, had advised me (via the net) to perform a simple check of receiver sensitivity: I was told to listen to the receiver output while connecting and disconnecting the antenna. If connecting the antenna resulted in a noticeable increase in the noise output of the receiver, there would be no need for additional front end amplification. My receiver was not really doing well on this test, so I was concerned that adding more tuned circuits at the front end would worsen the sensitivity problem. It seemed to me that a stage of RF amplification that included a couple of tuned circuits might help me banish the unwanted preachers and code practice sessions without further degradation of receiver sensitivity.
Doug Demaw's QRP Notebook pointed to a simple, grounded gate FET amplifier with tuned circuits at the input and output. I quickly put this stage together on its own small PC board and put it between my antenna connection and the Mixer board. The amp was obviously amplifying, but it seemed to be getting carried away. Whenever I'd tune both the input and output circuits to peak, the amp would begin to oscillate. I turned to the Internet and aske for advice. Help quickly came from afar. A fellow named PK Singh sent me an email with the solution: I had to "tap down" on the toroidal coils in the two tuned circuits. This deliberately introduced impedence mismatches that effectively reduced the stage gain and thus stopped the howling. (A side benefit was a noticeable increase in tuned circuit Q - a big help in my battle with the 22 Meter station!). With the tapped down amp in the circuit, my receiver passed Roy Lewellan's noise test with flying colors and I was no longer the subject of harrassment from 22 meters and W1AW. Viva el Internet!
Coffee can frequency readout
My frequency readout scheme needed some work. The tuning capacitor I was using had a little venier reduction drive built into the cap. This made for very smooth tuning, but it made it impossible to work out any kind of frequency readout on the front panel. I had to peer over the panel and look at the variable capacitor to determine where I was in the band. In an age of multidecimal numeric digital readout, I was clearly behind the times! And my neck was starting to bother me!
To upgrade, I found a junkbox 365 pf variable cap with no built in reduction drive. This was about twice the capacitance that I needed, so I simply plucked out about half of the rotor plates. I also found a Johnson 6:1 reduction drive in a junker Swan 240 transceiver. With a piece of scrap aluminum, I engineered a little mount for the capacitor. The Johnson drive allows for the attachment of a frequency readout dial. I found that the top of a coffee can (the metal part you always throw away) was ideally sized for my front panel. Soon I had the modified cap, reduction drive and coffee can readout dial mounted on the front panel. A triangular piece of electrical tape provided a sharp looking pointer. A few pieces of masking tape on the coffee can top served as frequency markers. I realize that my "coffee can readout" will seem incredibly crude to those accostomed to glowing numerals, but I get a real kick out of it every time I spin that little homebrew mechanism.
In a certain sense I was done. I was able to pair my new receiver with my QRP transmitter and was easily able to make QSO's. I was working European stations regularly with 3 watts out. But my simplified crystal filter was a too wide for serious CW work. I could hear several CW signals simultaeously and - worse yet - I could hear the "other sideband" on the stronger signals. So I hadn't really achieved the coveted "single signal reception" status that is - after all - one of the main reasons for going the superhet route.
There are a number of excellent article out there on the design of CW crystal ladder filters. Unfortunately the building of these filters requires the use of some special test gear to determine the electrical charecteristics of the particular crystals that will be used.
Wishing to avoid the construction of test gear that would be more complicated than my radio, I decided to simplify filter construction. I bought a bag of 50 3.579 Mhz TV color burst crystals from Dan's Small parts. I then built a simple Colpitts oscillator circuit on a Radio shack breadboard. I tuned my Drake 2-B receiver to 3.579 Mhz and started plugging crustals into my breadboard oscillator. I screened out those rocks that were signficantly off frequecncy, then I went through the pile again, judging by ear (using the tone from the Drake 2-B) to select three crystals that were very close in frequency. (I know that a frequency counter would have made this easier, but I don't have one so I had to "make do".)
I simply pugged these crystals into the filter circuit described in Doug DeMaw's 1982 article. Essentially I was "hoping for the best", hoping that the characteristics of my rocks would not be significantly different from those employed by Doug DeMaw.
It all worked out very well. The new filter significantly sharpened my receiver's selectivity. I could no longer hear strong signals at two points on the dial. Single signal reception had been acheived!
My filter proved to be far to sharp for confortable SSB reception, so I worked out a little switching arrangment that allowed me to switch between my original (wide) filter and the new, sharp CW filter. MISSION ACCOMPLISHED
I found that my technical skills improved dramatically during the course of this project. I even notices a marked impovment in manual dexterity. By the time the receiver was finished, I was much more confident about putting together my own circuits. In order to be truly "100 percent homebrew", I needed to whip up a power supply for my station and a sidetone oscillator for my transmitter. These projects were quickly completed and I was soon on the air with a 100 percent HB station.
Homebrew is good for you! It really doesn't matter what band or mode
you build for, a homebrew radio will provide a kind of satisfaction
from store-bought units. A project like this will improve your skills,
expand your knowledge and will put you in league with all of those
inventors who have turned piles of parts and wires into devices that
extract signals from the ether.
Return to Bill's homepage