Have you ever started a ham radio project and ended up learning something totally different along the way?  Well, that happened to me with my project to build an audio break out box (or audio splitter as it is sometimes called).  Besides ending up with a useful piece of equipment, I also mastered one versatile method of making printed circuit boards.  This project grew out of a desire to be able to share QSO experiences with others at Field Day.  During Field Day, contesting, general emergency communications, or almost any other ham radio activity, headphones are often a necessity because of the close proximity of operators to each other.  With stations located only a few feet apart, shouting into a microphone, or listening to CW on a loudspeaker might disrupt a nearby operator.  With a standard "y" connector two operators monitoring the same radio will often have trouble adjusting the individual volume to their headphones.

With this device, each operator hears only what he wants to hear and has the capability of adjusting the volume of his headset to his own preference.  Up to four people can monitor the same radio, but extra channels can be added by the builder.  The only other pieces of equipment needed are the headphones.  It makes it easy to share the Field Day experience, or for that matter, any other type of exciting ham operating experience with non hams who are also potential hams.  I call it BOB , which is short for Break Out Box.

The closest thing that I have found to what I wanted is shown in many recent editions of the ARRL handbook (see "Audio Breakout Box" by G4YNM, 2006 Handbook, Chapter 19.26).  In fact, I have pondered this article for years thinking of how I could tailor it to my own requirements.  G4YNM's design has five outputs.  However, four of the outputs are low power and are really intended for input to fixed volume equipment.  I wanted something for up to four users who would need varying amounts of gain based on the individual headphones in use and each individual's listening habits.  Having a bit more power would also be good if someone actually wanted to drive a small speaker. The LM386N-4 audio amplifier chip, which I ultimately chose for this project, has the advantage of being able to supply up to 1 Watt of output, and is designed to withstand up to 22 Volts on its supply.

Above is the schematic supplied with the National Semiconductor data sheet for the LM386.  It is pretty much self explanatory except for the 10 ohm resistor and .05 uF capacitor.  These are for impedance matching of the output.  The only other parts needed for the basic circuit are a potentiometer for the volume adjustment and an output filtering capacitor.  In this format the LN386 is set for a gain of 20.  Optionally adding a bypass network to pins 1 and 8 can increase the gain to 200.

I obtained the needed parts and started fabricating the circuit on a Radio Shack prototype board 276-168.  This 3" by 4" board has been a favorite of mine in the past for various other small projects.  However, when I completed the project, I discovered a few shortcomings.  First, it was difficult servicing the solder side of the board.  The four potentiometers and headphone jacks were mounted to the project box and wired to the board, and this made it difficult to remove the board from the project box.  Bending all the leads to these parts to access the bottom of the board was also difficult.  After a long period of debugging I ended up with only one of the four amplifier circuits working.  I finally had to admit that my eyes are just not as good as they used to be.  I needed another way of producing this project that was easier on my eyes.

What immediately came to mind was the work of the master radio designer Wayne Burdick, N6KR.  Since his famous Norcal 40 QRP transceiver design all the way to the current Elecraft creations, he (and his Elecraft partner Eric Swartz, WA6HHQ) have incorporated "wire-less" construction into all of these kit designs.   To incorporate this feature would entail soldering the potentiometers and jacks directly to the circuit board, and this would require special ordering of the parts.  Also I needed a circuit board with solder lands that were spaced far enough apart and were easier to see to avoid solder bridges that are now more common with my poorer eyesight.  I had a feeling that this new requirement was going to complicate the project.

I drew up the prototype PCB layout in AutoCAD LT, trying to spread out the traces and make a layout that would be easier for me to solder.  I had to make all the traces from scratch since I did not have any templates for the parts.  So the results may look a bit unconventional, but I feel that the layout makes soldering easier for me.   For the original prototyping board I laid out one complete amplifier circuit first, and then copied this three more times.  I then joined many of the common traces such as for the ground, and the voltage supply.  Finally I added the capability to add more parts to the supply, audio in, and audio out circuits.  This would allow for the addition of RF bypassing capacitors later as I researched that subject.   I felt that I had a nice layout.  The next question was whether I could convert the layout to a high quality circuit board.

And that was a challenge for me.  I had made a small circuit board once using the "iron on a laser print to a clean bare copper board" method, but the quality of the board was disappointing ¹.  I started studying different methods of PCB fabrication.  All were much more involved than what I was used to.  I finally decided to try to make a board using the positive resist method.  This method involves using a copper clad board that has been coated with a positive photo sensitive resist.  When the exposed board is developed, areas of the resist coating that have been exposed to light dissolve.  This leaves an image of the circuits on the board.  The board is them etched in a ferric chloride solution to remove the exposed copper.  The result is a circuit board.

I purchased some laser printer transparency film² to make the positive image for the exposure.  The results were disappointing.  The black image did not appear to be dense enough to be effective.  I finally resolved this by setting up the printing of the image to be only in one quadrant of the transparency.  Since the size for the circuit board is 3 by 4 inches, this allowed four images to be printed on the transparency.  Immediately after running the transparency through the printer for the first image I put it through the printer one more time.  I then rotated the hot (from the fuser) transparency 180 degrees, placed it back into the manual printer tray and printed another image (twice) onto the transparency.  For the third image I flipped the transparency over to its other side, and printed the image again twice.  For the forth image I rotated the transparency again 180 degrees and printed the last image twice.

The results of this effort were interesting.  The first double image was blurred (out of register), while the last three double images were in perfect register.  The double images also had dark enough traces to act as a good positive image for the exposure.  Fortunately the transparency material was two sided, which allowed me to flip the transparency over and print a total of four final images.  I feel the first image was blurred because the transparency was cold for the first printing.  After the transparency was warmed up the second pass for the first image was blurred because the transparency material probably changed dimensions because of the heat from the printer's fuser.  However, the subsequent printings were all done warm and that's why they were in good register.  So I now had a good image of the circuit for the next step.  (In retrospect one pass may be dense enough since the final production of the board revealed great latitude of the system (see below), but I did not test using a single pass transparency to make a PCB.)

For a contact printing frame I borrowed two panes of glass from 5x7 picture frames that were being used to display our family photos.  I placed a Datak 3 by 4 inch Premier board (14-304) over the bottom sheet of glass, then the emulsion side of the positive image over that, and finally a sheet of glass over all of this.  I used small metal binder clips³ to keep the sandwich tightly joined.  I had chosen to use the Datak Premier boards because Datak claims that they could be properly exposed using a standard 100 Watt tungsten bulb at a distance of twelve inches for 10 minutes.  This seemed a lot easier than building an ultraviolet exposure frame.  After exposing the board as recommended I placed the board in a tray of Datak (12-404) developer.  Nothing happened.  I wasn't sure what went wrong but I tried fogging the board by placing the bulb one inch from the developer tray with the board still in the solution.  Again nothing happened.  The bulb was acting more as a safelight, than as an exposure device!

I then read the instructions that were contained on and inside the packaging for the individual boards.  (Datak says to ignore these instructions.)  Since the emulsion is light sensitive it is sealed in a light proof bag.  The outside of the bag had some interesting information on it.  First, I found that the manufacturer was "Kinsten".  Also, they recommended an 8 minute exposure with a 10 to 15 Watt standard fluorescent bulb.  I tended to believe this information more than the Datak information since it was included in the original packaging.  The next night, I took a light fixture from one of my aquariums and placed it on my dining room table.  It has one 95 Watt dual tube compact fluorescent bulb that is so bright when lit that you can't look directly at it.  I shimmed it up so the bulb was about 3 inches from the top of the table.  I then made a test exposure using a new board with the positive image.  By using an opaque piece of cardboard, I was able to make a test strip on the board using exposures of 32, 16, 8, 4, 2, and 1 minute.  I then developed the board as before.  The 32 minute strip developed almost instantaneously.  It took about 20 minutes for the 1 minute strip to fully develop.  After the board was washed I inspected all the strips, and they were perfect except that the 32 minute exposure had a little bit of print through of the traces, but it was still quite usable.  This is amazing to me, as it appears that this system has a great deal of latitude in choosing an exposure.  I decided to use 4 minutes as my standard exposure and exposed another board.  I then put both boards in a warm solution of ferric chloride.  After about 30 minutes I had two very nice printed circuit boards.  Later, I used some brake fluid to remove the resist from the copper traces.

It was now time to build the circuits with the parts that I had collected.  This time the building went quickly as there was no need for installing interconnecting wiring.  The following diagrams explain the layout of the board and where the parts are placed.

The effort to make the circuit board from scratch was justified.  It took considerable time to debug the PCB process, but after this, it was actually very easy to quickly make the boards, and subsequently to install the parts.  I will definitely not hesitate to make other boards of designs in the future if the need arises.   Not only does BOB work well and sound good, but it also looks much better than using a proto board.  On my transceiver I normally run the audio volume at about 9 o'clock.  This gives plenty of volume and then each user can adjust their potentiometer on BOB for their own preference.

Installing the parts took far less time than the original attempt on a prototype board as no interconnecting wiring was needed.  The installation of the jumpers could also be avoided resulting in less construction time if the board was made double sided.  This I only recommend for a production version of the board.  For small quantities I prefer a single sided board.

The final version of the board simplified the layout and eliminated some jumpers, but the circuit design remained unchanged.  I now have a reliable way to demonstrate "hands on" amateur radio to potential hams.

During Field Day we had a visit from a Boy Scout leader from Houston who was very impressed with BOB.  He felt that two BOBs in parallel (8 listeners) would be about perfect for radio merit badge work.

1   http://www.pituch.net/p/Steve's%20Page/Radio/SW40+/SW40+.html
2   Office Depot small binder clips (3/4") (429-415)film
3   Office Depot Black & White Laser Transparencies (753-611)