During our Easter 2004 stay at our holiday QTH we at the worst of times discovered that my KAM had gone bad. Stone dead, no schematics and a quick check showed an exploded transistor.
Later repair shows that the exploded transistor was part of a voltage inverter that was powered before the voltage regulator. Likely we had gotten 32V backwards into the KAM during the last 23cm contest before our holiday trip. The PA we use on 23cm is powered using 32V. Now it's repaired again, two transistors and a 4069 needed to be replaced. That could not have been repared on site due to lack of spare parts in Stora Blåsjön.
So, what do you do when you are "out in the wild" and do not have a CW-key, will you take something else and turn it into a electronic keyer.
But what about switches for controlling it? Sure enough we did not bring any bencher or begali manipulator. This would have to be the hard-core stuff. Well, my father looked around a bit and soon enough found what he needed. Some pieces of wood, a few screws, plugs, washers and most important. A blade for a saw that could be cut into half to make two paddles and a double pole light switch that sure enough would never be used for its original purpose. From the switch the contacts could be salvaged and be used as great contacts.
After a few hours of bending, drilling and screwing the manipulator was ready. It only needed minor adjustments to make both sides give the same tension and movement distance. A few turns of tape arond the saw-blade gave the contacts their adjustment.
The software had been pretty simple to write. Already before we decided to build a electronic keyer I had worked out most parts of it in my head. All the original drawings are either thrown away, or left in Stora Blåsjön, but it basically consists of a state-machine that you move around in due to the different paddles beeing used and a timer loop expiring. If I find the original drawings I migth scan them just to give you that off-road feeling.
One thing that did worry me a bit and my father a lot was if the contacts would produce a lot of bounces. With the choosen design this was not a problem from the start, and a little extra code made it even more safe.
Considering the time used to construct this project (3 hours * 2 persons) we are quite proud of it. We often compare this to the BBC TV-show "Rough Science". That is a great show so if you have missed it, go search for a channel that brodacast it! I don't see that anyone could ever reproduce this as is, but use it as an inspiration. The source code for the PIC 16F676 is available here if you want it.