[Elecraft] Feature Request: improved internal keyer and CW PTT behavior

[donovanf at starpower.net]

Hi Richard, 


The K3 and external K1EL Winkeyer is a great combination, it corrects 
most of the deficiencies of the internal K3 keyer and PTT VOX delay. 


Hopefully the next next upgrade of the K3 internal keyer will at least 
make it better, if not perfect! The biggest faults are: 
- VOX delay added to the tail of the PTT signal when in VOX mode 
- inability to use the internal keyer in CW PTT mode, it always reverts to QSK 


Thanks for your wise observations! 


73 
Frank 
W3LPL 

----- Original Message -----

From: "Richard Ferch" <ve3iay at gmail.com> 
To: donovanf at starpower.net 
Sent: Wednesday, February 15, 2017 3:30:16 PM 
Subject: Re: [Elecraft] Feature Request: improved internal keyer and CW PTT behavior 


Frank, 


In a nutshell, what you would like is PTT control when the CW comes from the computer (via the straight key jack), and VOX when the paddles are used (via the paddle jack). 


OK - but it may not be quite as simple as it seems. Not that it cannot be done, but it might not be quite so easy to program and debug. 


First off, some users might want VOX from both inputs (in particular, someone using QSK probably would). So, this looks like a new configuration option - VOX from the paddle input only, but not from the straight key input, vs. the standard option with VOX enabled on both inputs. OK so far. 


Suppose you are using the paddle and want to continue directly with a computer message. If you hit the function key before the VOX has dropped, the computer will assert PTT and the rig will stay in the transmit state, but you will want the VOX delay to be disabled for the duration of the computer message, i.e. the firmware has to remember not to impose the VOX delay, whereas it would have done so if you had not hit the function key. The state machine is getting a bit more complicated. 



Now, what happens if the operator wants to interrupt a computer message by touching the paddle? Using a Winkey, touching the paddle aborts the computer-generated CW. If you want to be able to do this, then touching the paddle will have to turn the VOX delay on and temporarily disable the straight key CW input. Meanwhile, the computer has no way of knowing what has happened, so it has not dropped PTT. If the paddle input plus VOX delay ends before the original PC message is finished, the computer will still be asserting PTT and toggling the CW line. You will want the rig to continue to ignore the straight key input until the computer releases PTT - but in the meanwhile, do you also want the rig to be ignoring the hardware PTT from the computer and switching back to receive, even while the computer is still "transmitting"? This could get messy. Maybe the easier choice would be not to allow paddle input to interrupt the computer keying, so you have to hit the Esc key before using the paddle. Simpler to implement, but not nearly as convenient as using a Winkey. 


Add to this the fact that the interaction between CW and PTT in the K3 has always been somewhat problematic (witness the CW problems when the TX Delay setting is changed from the default 8 ms); throw in the fact that somehow this has to work even when computer PTT is done via software commands rather than a hardware signal (this has historically been a significant problem area for the K3 that was addressed in F/W revision 5.46 after significant testing effort); and maybe programming and debugging this might just turn out to be harder than it looks. 


Maybe the easiest solution is just to use a Winkey after all! 


73, 
Rich VE3KI