Ssdr TX-Delay Video

Steve N4LQ

This video shows the effect of TX-Delay on the transmitted signal https://www.youtube.com/watch?v=_B2S5pew0pg&feature=youtu.be

Paul Christensen, W9AC

It's more serious than shown on the video.  See photos below.  When the 6700 Tx Delay is set to the default 0 ms and CW Delay is 0 for full QSK, we now get the first waveform that shows hot-switching of the trailing edge.  This may be in the RF or DSP domain, but it's still a hot-switch timing error. 

The second image shows CW delay at 17 ms and TX Delay staying the same at 0.  It did not look like this in prior SSDR releases.

 

Paul Christensen, W9AC

Below with 17 ms of CW Delay added.  Blue trace in both screens is the key line.  The onset of the rising pulse is the key closure.  10ms horizontal divisions.

Paul, W9AC

Paul Christensen, W9AC

Next, for those thinking we need to have TX Delay higher than 0 ms, I set it to 15ms.  The keyed waveform is delayed by the proper amount of 15ms, but now "dit shortening" occurs and the trailing edge of the waveform is still cut. 

The problem is not so much the dit shortening as that is discussed in the manual when Tx Delay is used.  The main problem is the truncated trailing edge that is causing severe key clicks.

Paul, W9AC 

Theoff. 

Al_NN4ZZ

Paul,
Thanks for the traces, they really help.  

I'm trying to extrapolate for my settings using an amplifier TX Delay = 15 and CW Delay = 200 for semi-breakin.  It would seem that would avoid the hot switching problem and the element shortening problem also.  Do you agree?

Regards, Al / NN4ZZ  
al (at) nn4zz (dot) com


 

Ken - NM9P

Excellent, Paul.  
I was getting ready to request that someone with a scope could make those exact tests!  
Thanks.
Can you run an additional few to document the changes made if TX Delay and CW Delay are both altered, as Al is suggesting?  What are the timing break-points.
FOr now, I have been calling it a "timing inversion" where the cut-off time is not compensated with the TX Delay adjustment, especially if the CW Delay is shorter than the TX Delay.

Thanks again,

Ken - NM9P

Ken - NM9P

Also...
Is there any difference in effect with the CW speed control setting at the three different "threshold" speeds that have been reported....?

Ken - NM9P

Paul Christensen, W9AC

Ken,

I tried to, but there are too many moving pieces between CW Speed, TX Delay, and CW Delay.  Also keep in mind there are three CW waveform "break points" that occur at 5-19 WPM, 20-29 WPM, and 30+ WPM.  The Delays affect these break points differently.

The matrix just gets too big for the limited time I have. 

Paul, W9AC

Paul Christensen, W9AC

Al,

Those settings would normally be fine.  Using any any modern amp with either PIN or vacuum relay switching -- and/or perhaps a reed relay on the input, a TX Delay setting of 0 should work well.

Based on what I'm seeing with v1.3 and my 6700, I would absolutely validate amp timing with a scope until we know more. 

Paul, W9AC

Steve-N5AC

We reproduced and found the issue here are we are working to correct it.  We will release the fix in the upcoming 1.3.x point release as soon as we have it ready. 

np2g

looking forward to this correction

click click for now

Ken - NM9P

Wonderful! Thanks.

Brian - AE0P

Steve,

These timing issues are of a concern to many of us, but the "work-a-round" of proper settings are not well defined and clear for the variety of models, cw speed, amp delays, cw delay... etc.  
   
1.  Please list for us the definitive, proper settings for release 1.3 so we can feel assured and confident at we are free of clicks, hot switching etc.  

2. When will point release 1.3x be available that corrects these problems?

Tnx es 73,

Brian, AE0P

Paul Christensen, W9AC

I'm not Steve, but I've taken many measurements.   For now, I would ensure that CW Delay is set to 15 ms. to avoid key clicks and "hot switch" effects with SSDR V1.3. 

Paul, W9AC