Monday, December 26, 2022

Forty Years of Personal Computing - Code Practice Program

In the spring of 1980, as my first year at Georgia Tech was coming to a close, my parents were coming to pick me up and take me home to West Virginia in June. My brother Ben suggested during the trip, we could take the FCC exams for Amateur Extra class. At the time, we were both General class, and all exams were administered at official FCC offices, such as Atlanta.

I wondered how I might practice for these exams. Although I was a member of the Georgia Tech club station W4AQL, getting access was a hassle. I hadn't been on the air since I'd come back to Tech in January. I was afraid I had forgotten Morse code entirely. My solution was to write a code practice program for my SWTPc 6800 Computer System, using the Kreepie Peepie

The program is small, just over 200 bytes long. It creates a five character code group by choosing characters randomly. Then it sends the group, complete with a word space. A check of the serial port determines if someone has input a character, indicating a request to stop. If not, it continues to generate and send the next five character group.

The characters themselves are sounded out by toggling a bit on the Kreepie Peepie. A wait loop of roughly a millisecond is used. This gave a pitch of about 1 kHz. The wait cycles are repeated a sufficient number of times to sound out the element (BAUDV). Silent periods used the same timing element, without toggling the Kreepie Peepie. This made it easy to adjust speed by changing the BAUDV location. I had this programmed to 64, which came out right about 20 wpm.

I'd typically run the program for a few minutes, copying as many of the five character groups as I could. It was enough. I passed the FCC 20 wpm exam.

Thursday, December 8, 2022

Upgrading the KK1L Antenna Switch Load Resistors

My KK1L board, showing the old resistors on the
board, and the new ones on the desk.
While testing the KK1L Antenna Switch after mounting to the Single Point Ground (SPG), I noticed that some of the load resistors were no longer 50 ohms.

These resistors provide a load impedance whenever a antenna jack is not selected by either port A or Port B.  This protects the antenna from static build-up as well as dissipating any coupled RF energy.

I originally used 50 ohm, 1/2 watt resistors I had on hand. Clearly they were not up to the task. On the ports for the shunt-fed tower, and the 80/40/20m trap dipole, these resistors were completely open. Both showed signs of overheating. 

Most likely, these resistors succumbed to dissipating too much couple RF energy. I needed bigger ones. KK1L had 50 ohm, 50 watt resistors in his Mouser parts list, so that's what I ordered. These sorts of things come in handy, so I bought ten. Due to supply chain issues, they were back-ordered for months. But they finally arrived.

Replacing these parts is a pain. First, I had to remove the KK1L box from the SPG panel. Next, I had to remove the board from the aluminum box. Before I did that, I made sure to mark the locations the mounting screws for each resistor. To remove the board, I had to unsolder the eight connections to the SO-239 center conductors and bend them out of the way. Then came twenty-some nuts holding the board in place. 

With the board separated from the box, I drilled the holes for the resistor mounting screws. I used a numbered drill bit the same size as the hole in the board to give me the largest tolerance. With the holes drilled and de-burred, my attention turned to the board.

Board with new resistors
Next step was to remove the existing resistors. Where they connect to the relays was easy, but the connections to the ground plane were harder. The ground plane tended to carry the heat away, making it hard to melt the solder without damaging the board. Getting these holes cleared of solder took a lot of effort. 

Once done, the new resistors mount cleanly on the underside of the board. I oriented the resistors so the ceramic patch was toward the aluminum box. This patch does not appear electrically conductive. 

After the resistors are soldered, the process is reversed to re-install the board in the aluminum box. In addition to the existing twenty-some nuts, there are also six new #4 screws and nuts to mount the resistors securely. With that in place, re-soldering the eight connections to the SO-239 center conductors completes the job. 

I mounted the KK1L box to the SPG, and then re-tested all the switching combinations. This was to ensure I had connected the switching lines correctly.

I hope the new resistors are up to the job. I'll have to check on them in a few months to make sure.