Have you ever wondered why some electronic keyers seem to work better than others?
I have a large (and growing) collection of electronic keyers and I've noticed that, while using the same paddle, I'm able to send flawlessly with some keyers while with others I can barely complete a sentence without an error. What's the difference?
I'm convinced that differences in timing (contact closure sensing via interrupt or polling), key press memory (dot and/or dash) and contact de-bounce are significant enough to change the way an electronic keyer feels.
This goes beyond the Mode A/Mode B Iambic keyer operation. While the majority of my paddles and virtually all of my keyers are capable of iambic operation, I seldom use the capability. Yet, I can feel the difference when I select the alternate mode on many of my keyers.
Perhaps the first electronic keyer to recognize that timing differences and "feel" were important was the Idiom Press Super CMOS II chip (Version 2) and Logikey K-1 keyer. The V settings (V0 through V9) provided a wide range of timing patterns and dash/dot memory. Of all the keyers I own and have used, the Logikey K-1 is still my favorite.
I wish that I was the first to notice this attribute of electronic keyers. However, after some research on the web I discovered comments about and evaluations of keyer timing published by Chuck Adams K7QO (formerly K5FO). Much of Chuck's material is centered around iambic keying and the timing required for iambic operation. Other material published by Marshall Emm, N1FN references Chuck's analyses and offers some insight into the precise timing required to use an iambic key at higher keying speeds. www.morsex.com/pubs/iambicmyth.pdf
I have been reviewing the source code for as many electronic keyers as I can locate including; K1EL K8 Keyer, N0XAS PicoKeyer and the K3NG Keyer in an attempt to determine exactly what in the code makes a keyer have a particular feel. As each of the keyers above use a PIC, I've programmed the code and have begun testing a breadboard version of each keyer.
I've even toyed with creating a PIC based electronic keyer tester that would control the paddle inputs and measure the output timing while looking for the sensitivity of a keyer to input timing variations. Again, Chuck Adams beat me to the punch. In one of his on-line tutorials he states,
"OK, gang. I spent half the morning working on this and let's hope that I didn't waste my time or valuable bandwidth posting this. I'll put it in a book and by that time I will have experimented with the other keyers even if I have to microprogram a 16F84 driving relays to test algorithms at critical time periods. There are still many questions to be answered and so little time." (http://www.n9vv.com/k7qo-a-b-keying.html)
While I'd probably skip the relays, and use something newer than a 16F84, Chuck has the prior art for any keyer tester design. (Not that I plan on patenting a keyer tester.)
So what do you think? If you have any comments or suggestions, send an email to:
Note to spammers and spambots: I will forward any junk mail to the proper federal and state authorities for investigation. Don't even think about it.
