THE SIDESWIPER HOMEBREW MORSE KEY PROJECT
Whilst surfing the internet for information on Morse keys I came across references to an early key design called a Sideswiper, there seems to be some confusion over the true Sideswiper design some guys assume that any key that has a sideways action is a Sideswiper, not so, the true Sideswiper only has one make and break contact, it may have two sets of contacts but they are either fixed or wired into a single make and break circuit. Also the sending of the code on these keys is continuous, i.e.the key arm moves from one side to the other as it makes up the code lingering on each side for a longer or shorter time depending whether the character is going to be a dot or a dash, the key self centers in the middle position where it breaks or open circuits. The design is old only having a short lifetime as it was superceded by the semi auto Bug key design, I think its almost certain that this Sideswiper design brought about the Bug key as when you make a Sideswiper it automatically tries to be a bug by vibrating the contacts as it sends a problem that has to be designed out.
So having found a few pictures of early designs and even a few later designs as they are still made I decided a Sideswiper would be my next project, I wanted a nice free action incorporating a couple of ball races and spring loaded gap adjustment as I was certain there would be lots of adjustments made before the project would be completed,  I would design as I go, overcoming problems on the way.
The basic action as I said would be a vertical ball race design seen on the right, the bearing holder forming a box made from brass pieces silver soldered into the required size and then drilled to take the bearing and the two bearing clamps. The main arm would be fairly short as this would be drilled to take the handle shaft and contacts one end, and the tension adjustments the other, at this stage I didn't have a real idea how the tension part would work so I decided to make it flexible enough to try out different ideas, and concentrating on the contact bit first which I thought would be fairly straightforward which it proved to be. 
Here is the first mock up on the left, the spring loaded gap adjustments in the form of a bar has been made and the handle and shaft fitted all of which caused no problems, at this stage I decided on a torsion spring to provide the tension as it is simple and self centering one end silver soldered to a shaft which was clamped into the rear of the main arm the other end clamped into a block, And it was a total failure, the action was so hard it was useless, so I decided on a smaller torsion spring as the next step
Here is the next stage with a much finer torsion spring the key did start to work after a fashion though it still felt a bit stiff and also it didn't seem to hold its settings well, it was impossible to clamp the far end of the torsion spring as if you did it limited the range of contact spacing owing to the arc of the arm trying to stretch the spring, all in all it wasn't satisfactory and I didn't feel happy with it, time for a rethink, I decided to keep the shaft but fit a couple of coil springs instead so back to the workshop
Here on the left you can see the key on test fitted with the coil springs and temporary adjusters, these coil springs caused all sorts of problems and it was at this stage I could imagine the original designers of the semi auto bug key had arrived at, as the key became very lively and the contact arm would bounce on the contacts sending very rapid dots, at the end of the arm you can see a damper which was a copy of a Vibroplex one which I fitted to try and overcome the contact bounce, but it was hopeless, back to the workshop for another idea....
Here on the right is the key under test again, this time fitted with a couple of spring loaded shells clamped around a shaft fitted into the main key arm, to be honest the engineering to produce these shells was beyond my skills, they had to be so accurate that any slight variances in the pair caused the action to be biased one way or the other, although when I did get it right the key performed very well indeed, but although they were so difficult to adjust they had given me a great idea, why not abandon the shaft and fit a block with two spring loaded flaps, I had seen this idea used on one side of a Vibroplex key shaft to produce the dashes, so again back to the workshop!!
And here at last is the final layout, the shaft has been removed and the end of the main arm now sits precisely  between two spring loaded flaps, these are loosely held at the far end and have a spring fit to a brass block which is bolted to the wooden base, they are very easy to adjust, the tension being adjusted for either swing of the main arm and  completely self centering, and after some polishing of the center holes the key now has a remarkably light yet crisp action and holds its settings indefinitely, I'm very pleased with it, time now to start finalizing the project below.

Sn 018

HOME