[Elecraft] K2 keying mod details and circuit theory

[W B Reese]

Dear Wayne

I would like to ask what is the rated top speed for CW with the new mod?

I am a CFO operator and am reluctant to give up my potential keying speed 
of 70 WPM which I read "somewhere" was the rated top speed of the K2.  You 
understand, I can't do that now, but I have "high hopes" for someday and 
it's important to me to retain that feature and perhaps even extend it.

If I may risk getting off topic and a bit controversial, it is my belief 
that when the rise/fall times equal more than 1/4 of the dot length the CW 
becomes harder to read, and when they equal 1/3 of the dot length an expert 
is required.  Easiest code for novice users is 1/5 ratio, and below this, 
no obvious improvement can be obtained.  I apologize if I'm upsetting 
anyone with these three statements, but they are my belief.

Next, I'll risk ire by saying that the rise/fall time can be a pure "S" 
sinusoidal wave having minimum bandwidth, and this is most desirable, but 
if it is not, the benefits of rapid rise/fall time still apply to the 1/5, 
1/4, and 1/3 ratio mentioned in the paragraphs above.

I have quite a few CFO friends who love the K2 for it's beautiful and rapid 
keying!

Please tell me this mod doesn't reduce the K2's potential speed.

Thanks!

TR WB6TMY
K2/100/KAT100 S/N 0838

At 10:56 AM 2/21/2004 -0800, you wrote:

>The complete keying mod document will be posted next week. But here's the
>executive summary: The original classic R/C exponential keying envelope is
>converted to an optimized sigmoidal (S-shaped) envelope. This is done by
>reconfiguring U10A as a 2-pole filer whose step response has the desired
>rise/fall characteristics. But we were also careful to not make the rise and
>fall times too long, which could result in "soft" keying. (We plan to write an
>application note for those who want to custom-tailor the rise and fall times,
>putting them somewhere in between the original and new values. Only one or two
>component will need to be changed to achieve this.)
>
>73,
>Wayne
>N6KR