On 22/10/2020 10:21, Christer Grännsjö via groups.io wrote:

IC-7100 FT8 transmit timing problem.
When transmitting the timing (DT) is off by up to 3 sec. (varies). Receiving is no problem.
Not an RFI issue. Both stations running low power into dummy load.
Both computers are on exact time. Change of PC does not help. Same problem with both WSJt-X and JTDX.
IC-7100 TX delay is set to off.
Suggestions?
Christer SM4FXR

Hi Christer,

what you describe is not really possible, the outgoing message timing is determined on in the software, only small extra delays due to latency, e.g. in a rig;s Tx DSP, should be added on. Only a matter of a few tens of milliseconds at most.

Do you mean the Tx starts 3 seconds late rather than the DT reported by those receiving you being off by up to 3 seconds? If so then how is your PC and rig connected for audio and PTT?

73
Bill
G4WJS.

Thanks to all that have responded.

As of now the problem is gone without me changing anything. We'll see if it comes back. The bad thing is that one does not know unless someone tells you or you monitor your transmissions on a separate receiver.

I realized that delayed TX start does not affect the timing (DT). I made a simple test connecting audio between two computers. Started TX on one of them and the decoding on the other was ok with DT 0.2. Broke up the transmit audio path and reconnected it in the middle of the TX sequence. Decoded ok and DT didn't change.
So the problem must be caused by a latency of the transmit audio in the PC or the ICOM audio codec.

73 
Christer SM4FXR

Are you using VOX? If so, could the problem be that the VOX is causing the delayed transmission? Try adjusting VOX sensitivity or try using PTT instead. 

On Thu, Oct 22, 2020 at 03:16 PM, Bill Somerville wrote:

Do you mean the Tx starts 3 seconds late rather than the DT reported by
those receiving you being off by up to 3 seconds? If so then how is your
PC and rig connected for audio and PTT?

Hello Bill (and all), just to add my input to this thread...

I'm currently running v2.2.2 on a 32-bit Win 7 PC (a Dell OptiPlex780 with 4G of RAM); the rig is an IC-7610.  My PC-to-rig interface is a G4ZLP device, which is similar to a SignaLink-USB but better...  :-)  The connection from the G4ZLP device to the rig is via the rig's ACC1 socket.  [Note to people - please don't suggest that I use the IC-7610's inbuilt audio codec instead of the G4ZLP device and the analogue audio input and output on the ACC1 socket; I have my perfectly valid reasons for preferring the analogue interface.]

The rig is keyed by the wired PTT output from the G4ZLP device onto pin 3 on the ACC1 socket; in effect the PTT signal is generated by VOX within the G4ZLP device.  I've got the WSJT-X PTT setting set to VOX but the rig's VOX is off (I'm not sure if VOX operates on the data input to the ACC1 socket anyway, when the rig is in USB-D mode).

I've tried using CAT for the WSJT-X PTT option (the CAT interface from PC to rig is via a K9JM CI-V router to the rig's Remote port) and that seems to work (possibly more) reliably but for some reason when I use CAT for PTT there's a large audio 'splat' at the start of each transmission that's very visible on the 7610's bandscope - it's maybe approaching 2kHz wide in total spaced equally about the centre - and is also audible on the transmitted audio.  There's sometimes also a 'splat' when a transmission ends.  Changing the WSJT-X Tx Delay setting to a higher figure, eg. 0.5secs, doesn't eliminate the 'splats'.

I quite often find that the rig only goes into transmit a few (typically 2 or 3) seconds after the start of a 15sec FT8 period; the delay seems to be randomly variable and is very often almost non-existent or is at least imperceptible.  On a very few occasions the rig hasn't gone into transmit at all during a 15sec period and on other occasions the changeover from Rx to Tx has happened towards the end of the 15sec Tx period, eg. at the 10 or 11sec point.  There's also often a start-of-transmission delay when I operate FT4; there doesn't appear to be any particular set of circumstances that causes the delay in either FT8 or FT4.  I haven't been able to associate the delay with anything else that's happening on the PC - there's no apparent correlation with other apps.  I haven't checked all the other WSJT-X modes but from what I can recall there don't seem to be any delays when I operate MSK144.

I'm quite prepared to believe that the seemingly randomly-delayed start of transmission is caused by something in my system rather than a software issue - the PC isn't exactly the most up-to-date hardware(!), or perhaps there's something odd happening with the internally-generated VOX within the G4ZLP device that is causing its PTT output to be delayed.  I haven't yet tried to examine the hardware PTT signal and also the Tx audio signal going into the rig using a 'scope to see if the PTT is being asserted at the correct time at the start of a 15sec transmit period, or if it's appearing late, after the onset of the audio.

I'm in the process of setting up and configuring a new PC (a brand-new Dell OptiPlex 3070 with a 9th-gen i7 CPU at 3GHz, 16G RAM and an SSD, which should overcome any PC hardware issues!); it will be interested to see if upgrading the PC to modern hardware eliminates the start-of-transmission delays that I currently experience.  I'll keep you posted...!

73
--
Martin G0HDB

On Thu, Oct 22, 2020 at 03:16 PM, Bill Somerville wrote:

Do you mean the Tx starts 3 seconds late rather than the DT reported by
those receiving you being off by up to 3 seconds? If so then how is your
PC and rig connected for audio and PTT?

Hello Bill (and all), just to add my input to this thread...

I'm currently running v2.2.2 on a 32-bit Win 7 PC (a Dell OptiPlex780 with 4G of RAM); the rig is an IC-7610.  My PC-to-rig interface is a G4ZLP device, which is similar to a SignaLink-USB but better...  :-)  The connection from the G4ZLP device to the rig is via the rig's ACC1 socket.  [Note to people - please don't suggest that I use the IC-7610's inbuilt audio codec instead of the G4ZLP device and the analogue audio input and output on the ACC1 socket; I have my perfectly valid reasons for preferring the analogue interface.]

The rig is keyed by the wired PTT output from the G4ZLP device onto pin 3 on the ACC1 socket; in effect the PTT signal is generated by VOX within the G4ZLP device.  I've got the WSJT-X PTT setting set to VOX but the rig's VOX is off (I'm not sure if VOX operates on the data input to the ACC1 socket anyway, when the rig is in USB-D mode).

I've tried using CAT for the WSJT-X PTT option (the CAT interface from PC to rig is via a K9JM CI-V router to the rig's Remote port) and that seems to work (possibly more) reliably but for some reason when I use CAT for PTT there's a large audio 'splat' at the start of each transmission that's very visible on the 7610's bandscope - it's maybe approaching 2kHz wide in total spaced equally about the centre - and is also audible on the transmitted audio.  There's sometimes also a 'splat' when a transmission ends.  Changing the WSJT-X Tx Delay setting to a higher figure, eg. 0.5secs, doesn't eliminate the 'splats'.

I quite often find that the rig only goes into transmit a few (typically 2 or 3) seconds after the start of a 15sec FT8 period; the delay seems to be randomly variable and is very often almost non-existent or is at least imperceptible.  On a very few occasions the rig hasn't gone into transmit at all during a 15sec period and on other occasions the changeover from Rx to Tx has happened towards the end of the 15sec Tx period, eg. at the 10 or 11sec point.  There's also often a start-of-transmission delay when I operate FT4; there doesn't appear to be any particular set of circumstances that causes the delay in either FT8 or FT4.  I haven't been able to associate the delay with anything else that's happening on the PC - there's no apparent correlation with other apps.  I haven't checked all the other WSJT-X modes but from what I can recall there don't seem to be any delays when I operate MSK144.

I'm quite prepared to believe that the seemingly randomly-delayed start of transmission is caused by something in my system rather than a software issue - the PC isn't exactly the most up-to-date hardware(!), or perhaps there's something odd happening with the internally-generated VOX within the G4ZLP device that is causing its PTT output to be delayed.  I haven't yet tried to examine the hardware PTT signal and also the Tx audio signal going into the rig using a 'scope to see if the PTT is being asserted at the correct time at the start of a 15sec transmit period, or if it's appearing late, after the onset of the audio.

I'm in the process of setting up and configuring a new PC (a brand-new Dell OptiPlex 3070 with a 9th-gen i7 CPU at 3GHz, 16G RAM and an SSD, which should overcome any PC hardware issues!); it will be interested to see if upgrading the PC to modern hardware eliminates the start-of-transmission delays that I currently experience.  I'll keep you posted...!

73
--
Martin G0HDB

Hi Martin,

the audio click you are hearing when using CAT PTT is probably due to the PTT not being asserted until after the audio from WSJT-X has started. WSJT-X does try and avoid that but some rigs do not provide feedback that they have switched to SEND mode.

I would guess that the delayed PTT via CAT is something to do with your K9JM CI-V router. Have you tried eliminating that for a test, taking care to remove any sensitive/buggy devices from the CI-V bus, like a PW-1 PA.

73
Bill
G4WJS.

On Sat, Oct 24, 2020 at 08:04 PM, Bill Somerville wrote:

Hi Martin,

the audio click you are hearing when using CAT PTT is probably due to the PTT not being asserted until after the audio from WSJT-X has started. WSJT-X does try and avoid that but some rigs do not provide feedback that they have switched to SEND mode.

I would guess that the delayed PTT via CAT is something to do with your K9JM CI-V router. Have you tried eliminating that for a test, taking care to remove any sensitive/buggy devices from the CI-V bus, like a PW-1 PA.

73
Bill
G4WJS.

Morning Bill, thanks for the above.

Re: the audio clicks (or splats!) at the start and occasionally the end of a transmit period, I take your point about the PTT not being asserted (within the rig?) until after WSJT-X has started to generate and send the audio, but shouldn't the Tx delay setting in WSJT-X take care of that?  My understanding has always been that the Tx delay is between WSJT-X executing its command to enable PTT and the onset of the Tx audio generated by the s/ware, so increasing the Tx delay to say 0.5secs should give ample time for the rig to receive the PTT command via whatever option is selected and action the changeover from Rx to Tx before the audio arrives.  As I mentioned previously, increasing the Tx delay setting when I use the CAT PTT option doesn't seem to affect the presence of the audio splats, although I haven't tried using a delay setting of >0.5secs.

With regard to your second point, there's a slight misunderstanding - I experience the seemingly randomly-delayed start of transmission when I DON'T use the CAT PTT option, so the presence of the K9JM router in the CAT connection between PC and rig shouldn't have any bearing.  The rig's PTT is controlled entirely by the G4ZLP datamodes interface device which appears to WIndows as an external audio codec in the same way as a SignaLink-USB.  The PTT signal generated by a VOX-type circuit within the G4ZLP device is hard-wired to the rig's PTT input on the ACC1 socket, so as soon as the audio generated by WSJT-X begins to emerge as an analogue signal from the codec within the G4ZLP device the VOX circuit should toggle the PTT line - apparently the 'VOX delay' is only of the order of 10-20msecs.

73
--
Martin G0HDB