[Elecraft] Re: Updated Version of Elecraft Keying Mod

Sat Apr 10 10:55:07 2004

Part of the problem here may be the descriptive language being used.

There are FOUR events per baud: rise time starting, rise time ending,
fall time starting, fall time ending.

In the original K2 keying envelope rise time starting and fall time
starting were VERY sharp. Rise time ending and fall time ending were
VERY rounded.

BOTH rise time starting AND fall time starting events produced broad
key clicks, easily discernible separately by sending long dashes with
a hand key, IF one was capable of hearing key clicks at all.

Wayne's mod removed the sharp edges from rise time starting and fall
time starting. The clicks went away.

At the same time, some of the roundness of the rise time ending and
fall time ending events was taken away and these two events now were
more like the starting events.

The before and after sets of keying waveshapes are broadly documented,
display results from some number of rigs, and are not in question.

Pretty good court case for what fixed what.

There is a danger here of losing sight of the forest among the
technical trees. Or as a favorite professor used to say, becoming so
theoretical as to lose one's grip on reality.

I use modeling as much as anyone, and work hard toward attaining
mastery of the involved theory. But there is a point when only what
the real wires are doing over real dirt counts for anything, theory
notwithstanding.

73, Guy.

----- Original Message ----- 
From: "John, KI6WX" <[email protected]>
To: "Guy Olinger, K2AV" <[email protected]>;
<[email protected]>
Sent: Saturday, April 10, 2004 1:52 AM
Subject: Re: [Elecraft] Re: Updated Version of Elecraft Keying Mod

> If the edges are causing the key clicks and the rising edge is the
problem
> with the K2, then we would expect that the second derivative of the
keying
> envelope would show a larger peak at the beginning of the rising
edge than
> at the beginning of the falling edge.  The opposite is true.  The
second
> derivative at the beginning of the falling edge is 23% larger than
at the
> beginning of the rising edge.
>
> So we have a falltime that is 50% faster than the risetime, and we
have a
> change at the beginning of the falltime that is 23% larger than the
change
> at the beginning of the risetime.  From this, I would conclude that
the
> trailing edge of the original K2 keying waveform is the primary
source of
> the key clicks.
> -John
>  KI6WX
>
> ----- Original Message ----- 
> From: "Guy Olinger, K2AV" <[email protected]>
> To: "John, KI6WX" <[email protected]>; <[email protected]>
> Sent: Friday, April 09, 2004 12:55 AM
> Subject: Re: [Elecraft] Re: Updated Version of Elecraft Keying Mod
>
>
> > It was NOT the rise time. It was the leading edges, the sharp
little
> > edge or corner at the BEGINNING of the rise time, and the
BEGINNING of
> > the fall time, that were causing the clicks.
> >
> > The problem was with the SHAPE of the rise time, not the amount or
> > length of rise time.
> >
> > 73, Guy.
> >
> >
> > ----- Original Message ----- 
> > From: "John, KI6WX" <[email protected]>
> > To: <[email protected]>
> > Cc: "Guy Olinger, K2AV" <[email protected]>
> > Sent: Friday, April 09, 2004 1:09 AM
> > Subject: Re: [Elecraft] Re: Updated Version of Elecraft Keying Mod
> >
> >
> > >
> > > ----- Original Message ----- 
> > > From: "Guy Olinger, K2AV" <[email protected]>
> > >
> > >
> > > > The leading edges were the killer in the K2. That's basically
what
> > > > Wayne fixed.
> > > >
> > > I have trouble believing this statement.  It is well known from
> > circuit
> > > design engineering that a faster risetime or falltime creates a
> > wider
> > > bandwidth signal.  The risetime of the original K2 keying
circuit
> > was 3
> > > milliseconds and the falltime was 2 milliseconds.  So how does a
3
> > > millisecond risetime create more key clicks than a 2 millisecond
> > falltime?
> > >
> > > -John
> > >  KI6WX
>
>
>