[WT-support] Support for WinKeyer version 23?

[Peter Stuge]

Laurent HAAS - F6FVY wrote:
> >> I think there are some advantages to doing things the way Win-Test
> >> it, one character at a time, though I can't list them.
> >
> > I would really like to hear what advantages that can be.
> 
> - Callsign and various variables insertion "on the fly".
> - Scripts called synchronously with keying
> - wtKeyer:Insert possibility
> - Various headphones actions ($RxRy) synchronous with keying
> - Various TX switching synchronous with keying

These are all one and the same point; that Win-Test wants to be able
to do other operations in sync with keying. I think that's a fair   
reason for sending only single characters, but more to the point I  
think it means that the communications protocol between hardware and 
software is not really well suited for the task. Unfortunately this  
is difficult to fix since the products come from different vendors
and development of both is proprietary and closed.


> - and the most important : It just works like this ;-)

This is just a lazy programmer comment. Sorry, doesn't work for me. :)


SM5AJV wrote:
> As I said earlier, I "think" this is causing the PTT line to drop.

I believe you are right about this Ingo. Even though all components
involved would have no problem staying synchronous with each other I
believe that the (perhaps lack of) protocol design; several layers on
top of USB instead of using USB features well and getting a hard
timing guarantee in the millisecond range. Serial ports on top of USB
are nothing but the lowest common denominator of data communication
and that is not really useful.


> If anyone could advice my a work-around I would be happy.

It seems clear to me that at least Win-Test will not change policy on
how the software drives the device.

The device producers could certainly change their product to work
around the software behavior, but I doubt that they will do so at
all, and at least probably not quickly enough to be useful.

I would suggest adding a glitch filter to the PTT signal. Maybe a
low-pass filter is enough, maybe something more complicated would be
required. Possibly doing signal analysis of keying to determine
current transmission speed, to dynamically set filter parameters..

Not the answer you were hoping for, I know that.


//Peter