[Elecraft] [K3] CW Speed & bandwidth (was Tokyo Hy-Power HL-1.2KFX amp interfacing (to ALC or not to ALC?))

[Vic Rosenthal 4X6GP/K2VCO]

But even if the RX bandwidth is wide enough that there is negligible 
distortion, you can't distinguish dits that are 5 ms long if the rise 
time is 2 ms.

73,
Vic, 4X6GP/K2VCO
Rehovot, Israel
http://www.qsl.net/k2vco/

On 30 Mar 2016 07:17, Jim Brown wrote:
> On Tue,3/29/2016 1:03 PM, Vic Rosenthal 4X6GP/K2VCO wrote:
>> The relationship of bandwidth and speed is that for given rise/fall
>> times, increasing speed will reach a point when the code elements blur
>> together.
>
> Yes, but that's RX bandwidth, VERY different from occupied bandwidth.
> The issue is RX bandwidth is the TIME (phase) response -- every change
> in amplitude is accompanied by a change in the phase response, that that
> phase distortion (remember, phase is time) blurs the code elements
> together.
>
>> So you need more bandwidth for super high-speed CW because you need to
>> shorten the rise and fall times.
>
> You need more RX bandwidth, for the reason described above.
>
>> But the occupied bandwidth doesn't change as you vary the speed.
>
> Right. Occupied bandwidth is your "footprint" on the band -- how much
> your signal is spread out.
>
> 73, Jim K9YC