CW with PTT Questions

[Bob Henderson]

Hi

I am running Thetis 2.10.3.11 on my HL2.  Much of my operating is CW for which I find the combination to perform well.  I currently key HL2 via Thetis in Semi Break-in via the MIDI interface. I monitor my sending via the Thetis software generated sidetone, delivered to my ears via an ASIO interface to address soundboard latency.

For timing reasons I would like to use PTT switching of T/R on CW rather than Semi Break-in with preset delay but I have so far failed to achieve this.

If I disable Semi-Break-in instead enabling MOX I hear the sidetone but no RF is generated.  I am wondering whether I might have missed something in settings or whether this might be a known limitation or bug?

I realise I can operate PTT and key the HL2 directly via its front panel connector but when doing so Thetis software sidetone no longer works.

I am wondering whether anyone might have been successful in operating CW with T/R under PTT control and the software sidetone continuing to function?

Many thanks.

73 Bob, 5B4AGN

[Bob Henderson]

It may be this is of interest only to me but perhaps with further explanation it may be of interest to others.

The issue I hope to resolve has its root in first CW element truncation due to the physical constraints of electromechanical T/R switching.  This is a problem which compounds as sending speed increases.  I have attached 3 scope screen grabs to illustrate the issue.  Hopefully this is allowed.  In each case a series of dits was sent at around 40wpm.

Screen A illustrates the effect of first element truncation with external keying via MIDI using Semi Break-in.

Screen B illustrates a similar effect when using CWX.  Please ignore the other issue with CWX which is its excess element weighting, which does not appear to be adjustable using the DSP; CW; Weight control.

Screen C illustrates avoidance of first element truncation when the R/T transition adequately precedes transmission of the first character element.  I achieved this by entering a keyboard space prior to the first character in the CWX memory. 

First element truncation can cause my callsign to be truncated such that instead of 5B4AGN I become HB4AGN.  I am sure you can imagine the confusion this causes.  A secondary effect is that I must operate with a long T/R delay to avoid repetitive corruption of keyed characters.  This long delay means I will frequently not hear the first character element of a station responding.

If I key via HL2 front panel jack and use its PTT I can avoid first element truncation but Thetis software sidetone is absent.

If I key via the MIDI interface I must use Semi Break-in to key the TX but I suffer first element truncation.  Thetis software sidetone functions if PTT is used instead of Semi Break-in but no RF is output.

CWX is capable of activating PTT before key down, as proven by insertion of a leading space character.

The following observation may be significant.

When I key the HL2 in Semi Break-in with duplex enabled I see my transmitted signal on the pan adapter at around -22dBm and RF is output.  When I instead key HL2 with PTT enabled via MOX I also see my signal on the pan adapter but at around -65dBm.  Not surprisingly no meaningful RF is output.

It appears as if for some reason the PA is not enabled when using PTT or MOX to facilitate T/R transitions for MIDI CW.  This seems to suggest a solution may reside somewhere within Thetis.

73 Bob, 5B4AGN

[Bob Henderson]

With sincere thanks to Reid my immediate problem has been resolved.  Not the issue affecting the inoperability of PTT with MIDI CW but the first element truncation when using Semi Break-in.

Reid suggested adjustment of Setup; Options, HL2 Options, TX Latency might reduce truncation and indeed it does.

The default appears to be 20mS which on increase to 28mS makes a very useful improvement as revealed in the attached screen grab D with dits at 40wpm.  Increasing TX Latency beyond 28mS did not provide further improvement.

As can be seen the character of the first element continues to be distorted but not significantly truncated.  It would take a better ear than mine to be concerned about this. 

73 Bob, 5B4AGN

["Christoph v. Wüllen"]

Normally I use an external keyer with PTT-out and CW-out,
and in this keyer I specify a PTT lead-in time of 150 msec.

You will notice that the first element (dit or dah) comes with
a slight delay, but truncation problems are gone.

You can connect the PTT-out and CW-out from the keyer to
the 3.5mm KEY/PTT jack of the HL2.

The keyer generates a square wave side tone, this is
passed through a narrow one-OP-amp AF filter which makes it
a nice sidetone with "rounded" edges, and with the
help of a plain vanilla AF transformer this is then
"injected" into the cord of my headphone.

This gives me a low latency side tone as well, generated
in hardware and working independent of the SDR/computer
audio gear.

Depending on the SDR software, the keyer may also generate
PTT and CW MIDI messages that are then processed by
the SDR program, but still the side tone is from the
keyer (this is actually my preferred setup).

I worked for a long time on getting a good CW setup, and
this turned out to be the best solution (for me) if it
comes to CW-with-SDR.

If you have a radio with PIN diode T/R switching the
game may be different.

> --
> You received this message because you are subscribed to the Google Groups "Hermes-Lite" group.
> To unsubscribe from this group and stop receiving emails from it, send an email to hermes-lite...@googlegroups.com.
> To view this discussion visit https://groups.google.com/d/msgid/hermes-lite/b89f5371-8ba6-4b22-924f-68a3db6a66een%40googlegroups.com.
> <Screen B.jpg><Screen C.jpg><Screen A.jpg>

["Christoph v. Wüllen"]

The latency problem has long been discussed in this group.

piHPSDR sets it to 40 msec without any user intervention (you
cannot change this value unless you change the source code and
recompile). SDR programs which still use a default of 20 msec
and rely on the user increasing this are not really
user-friendly, so I was surprised to hear that this
is still a problem with bona fide SDR software.

> --
> You received this message because you are subscribed to the Google Groups "Hermes-Lite" group.
> To unsubscribe from this group and stop receiving emails from it, send an email to hermes-lite...@googlegroups.com.

> To view this discussion visit https://groups.google.com/d/msgid/hermes-lite/b6dafd80-87db-4c04-9ada-97f0069f0a18n%40googlegroups.com.
> <Screen D.jpg>

[Bob Henderson]

Hi Christoph

Thank you for your note.

I used an independent hardware generated sidetone for some time but now prefer the Thetis generated software sidetone, as I find this works perfectly well so long as the sound board latency issue is addressed through use of an ASIO interface.  It turns out that first element truncation due to TX latency just needed an adjustment.  This was easy to do once Reid pointed me to it.  I favour MIDI keying rather than key the HL2 directly as HL2 and PC are not always collocated.

As a mostly CW operator, I find the HL2 together with the associated version of Thetis provided by Reid MI0BOT to be very capable and a pleasure to use.  While I have several other SDR and other rigs of traditional architecture they are not seeing much use these days.

73 Bob, 5B4AGN