Somebody (mathematically inclined) help me with this. Wide/narrow strips and fixed spaces between each "element" and character....drag a probe across this combination at various speeds.

Would all work out and sound correct?

I've developed a mind lock on this and just one step from building the damn thing to find out!

From a mathematical/theoretical perspective, I would say that it would work well. In practice, there is a fair amount of room for error (but then this is the case with any hand-sent CW)

It's an interesting idea and a "gee whiz that's neat" type of project that, thanks to this thread, I'm now considering doing. Although, I don't know that it would be particularly useful or beneficial. The video linked to by HB9FXW was a good example.

Let's explore some math:

First, to review the CW spacing and put it here for quick reference:

dit = 1x

dah = 3x

space between dits/dahs = 1x

space between letters = 3x

space between words = 7x

'CQ' = 27x

'DE' = 11x

'CQ CQ DE' = 79x

Also, we need to determine a measurement for 'x' our element length. The smaller the number, the more precise we have to be when cutting/building/layout because any error will be a larger percentage of the element and consequently be more noticeable in the finish product. The larger the number, the larger and more cumbersome the finished product will be. For example, if a dit is supposed to be 1 inch wide but is instead 1 1/8 inches wide, it will be much less noticeable than if it is supposed to be 1/8 inch wide and is instead 2/8 (or 1/4) inch wide in which case it will be twice as long as a normal dit. Also, the larger the physical width of an element the faster you will have to move the probe for a given speed.

If the width of a dit (or x) is 1 inch then 'CQ' would be 27 inches long 'DE' would be 11 inches long and 'CQ CQ DE' would be 79 inches long with proper spacing.

If the width of a dit (or x) is 1/8 inch then 'CQ' would be 3.375 (3 3/8) inches long 'DE' would be 1.375 (1 3/8) inches long and 'CQ CQ DE' would be 9.875 (9 7/8) inches long with proper spacing.

The next thing to consider would be 'rate' and 'rate control' - in effect, the sending speed. Using to the 'PARIS ' standard, 5WPM is equivalent to 240 milliseconds per dit and 13WPM is equivalent to 92.31 milliseconds per dit.

Using the 1 inch dit board above then would require you to traverse with the probe at a rate of 1 inch every 240 milliseconds (or roughly 4.166 inches per second) to transmit 5WPM. For 13WPM on that same board would require you to move the probe at a rate of 1 inch every 92.31 milliseconds (or roughly 10.833 inches per second).

Using the 1/8 inch dit board above for the same speeds would be 1/8 inch every 240 milliseconds (or roughly 0.5208 inches per second or 1.04 inches every 2 seconds) for 5WPM and 1/8 inch every 92.31 milliseconds per dit (or roughly 1.354 inches per second) for 13WPM.

Maintaining the rate at a constant pace and dragging the probe in a straight line would likely be the main challenges during use. Although, given how perfect or not most of us send as it is, I would imagine that the operator should be able to listen to the keying sidetone on the rig and adjust well enough for communication to occur.

Note: My apologies to the OP as this thread has drifted away from the wheel concept to that of a straight board. In an effort to bring it back, a circle with diameter of roughly 3.1449 inches with the elements around the edge would yield the same result as the 'CQ CQ DE' board above with the 1/8 inch dit size. In that case, 5WPM would equate to roughly 3.1645 RPM (about 18.9603 seconds per revolution) and 13WPM would be roughly 8.2276 RPM (about 7.2925 seconds per revolution)

Additional length would need to be added/calculated to include callsign or other information, this would also change the diameter of the wheel for that design and while the speed/rate on a straight board would remain the same, the RPM of a wheel would also need to be adjusted.