Water Fall display distribution of users #Waterfall #question


Chuck Moore <wd4hxg@...>
 

While trying to identify an anomaly around 2300 Hz on the waterfall display I was curious about what is above
2500 Hz. The typical grab the edge of the waterfall display with the mouse and pull right did nothing of course.
But after a bit of lifting various rocks I found the control to set the lower edge frequency of the waterfall. As my
rig will allow a 4 KHz wide IF bandwidth I decided to see what was above 2500 Hz and set the lower edge of the
waterfall to 1000 Hz which pushed the upper edge of the water fall to 3500 Hz.  After a few minutes the display
shown below, revealed a number of signals above the default 2500 Hz limit. This was observed while operating
on 17 Meters (18.100 MHz USB).

 


A couple of questions come to mind. (this is not a request for software changes but just a query.)

  • Why is the waterfall limited to an  absolute bandwidth of 2500 Hz?
  • In reading the manual, the recommendation is to only use the 1500 Hz to 2000 Hz audio
    range for transmission to mitigate transmission of audio harmonics. Of course if I open
    my IF Bandwidth to 4 KHz I have just negated the reason for using the specified frequency
    segment. How much of a problem are audio harmonics?
  • Is the audio harmonic content a show stopper if the 1500 Hz to 2000 Hz rule is ignored?
  • Can the waterfall be used to gauge the effects of audio harmonics?
  • It was a bit of a surprise to see the gap from 2300Hz to about 2550 Hz. But more puzzling
    to me was the cluster of ops from about 2550 Hz to 3000 Hz. Would their operation there
    not limit the number of distant ops that can decode the ops operating above 2550 Hz?
    Why stop at 3000 Hz?

Regards

Chuck WD4HXG


Bill Somerville
 

On 06/10/2021 22:10, Chuck Moore via groups.io wrote:

While trying to identify an anomaly around 2300 Hz on the waterfall display I was curious about what is above
2500 Hz. The typical grab the edge of the waterfall display with the mouse and pull right did nothing of course.
But after a bit of lifting various rocks I found the control to set the lower edge frequency of the waterfall. As my
rig will allow a 4 KHz wide IF bandwidth I decided to see what was above 2500 Hz and set the lower edge of the
waterfall to 1000 Hz which pushed the upper edge of the water fall to 3500 Hz.  After a few minutes the display
shown below, revealed a number of signals above the default 2500 Hz limit. This was observed while operating
on 17 Meters (18.100 MHz USB).

 


A couple of questions come to mind. (this is not a request for software changes but just a query.)

  • Why is the waterfall limited to an  absolute bandwidth of 2500 Hz?
  • In reading the manual, the recommendation is to only use the 1500 Hz to 2000 Hz audio
    range for transmission to mitigate transmission of audio harmonics. Of course if I open
    my IF Bandwidth to 4 KHz I have just negated the reason for using the specified frequency
    segment. How much of a problem are audio harmonics?
  • Is the audio harmonic content a show stopper if the 1500 Hz to 2000 Hz rule is ignored?
  • Can the waterfall be used to gauge the effects of audio harmonics?
  • It was a bit of a surprise to see the gap from 2300Hz to about 2550 Hz. But more puzzling
    to me was the cluster of ops from about 2550 Hz to 3000 Hz. Would their operation there
    not limit the number of distant ops that can decode the ops operating above 2550 Hz?
    Why stop at 3000 Hz?

Regards

Chuck WD4HXG

Hi Chuck,

the waterfall bandwidth is not limited to 2500 Hz. The WSJT-X audio processing chain is currently limited to roughly a 5 kHz upper frequency, this is due to the filtering applied after an initial down sample to 12 kHz bandwidth. Drag the sides of the waterfall wider to increase the displayed (and decoded) bandwidth. Use the "Bins/Pixel" spin box on the waterfall controls to change the frequency scale factor of the waterfall. A best option is to set the waterfall width to just a little wider than the pass-band of your receiver, using those controls and the "Start frequency" spin box. Unchecking the "Flatten" option and adjusting the waterfall and 2D spectrum gain and zero offset sliders for a reasonable display is a good way to visualize the pass-band of your receiver.

If you have CAT control of your rig then WSJT-X can ensure the Tx audio tones sent to the rig are always between 1500 Hz and 2000 Hz, whatever Tx offset you choose to use. See the "Settings->Radio->Split Operating" options to enable that feature. Without CAT control we recommend that you operate with Tx audio offsets above 1500 Hz and below the low-pass limit of your rig's Tx SSB filter.

Audio harmonics should only happen if you overdrive the Tx audio input to your rig, but be aware that is very easy to do, using CAT control and Split Operating is by far the best option to control harmonics. Of course not over-driving your rig's Tx audio input is always a good idea.

The waterfall only shows received signals, it tells you nothing about your transmitted signal quality.

Many users still have SSB rigs with crystal filters, they are limited to Rx bandwidth choice that doesn't usually exceed 2800 Hz. Those choosing to operate at Tx offsets above 3000 Hz are either using SDR rigs with almost unlimited Rx bandwidth options, or they have simply tuned their rig's VFO dial up a kHz or two, forsaking lower offset decodes for more space to operate in without QRM.

One reason to stop at 3000 Hz is that other modes operate above that offset, don't forget we share the bands with many other users and modes!

73
Bill
G4WJS.


Michael Black
 

You don't say what rig you have but you're correct many rigs have about a 2400Hz bandwidth.
If you turn off Flatten, set Start=0, and increase Bins/Pixel to see the right edge, and go to a quiet spot (usually 5Khz below or above the FT8 segment) you can see your Rx bandwidth which should look something like this.  Your edges are probably not as steep as my SDR.
Inline image



#1 Waterfall is not limited to 2.5KHz -- what makes you think it is?  The rig bandwidth is the main limiting factor,.
#2 Where do see a recommendation to limit to 1500-2000?  There is mention that is what the split modes do.   A couple of us have been on the war path the last few years to clean up all the operators generating harmonics.  None of them were using split mode as that mode avoids even the 2nd harmonic which could be at a minimum of 3KHz (i.e. 1.5Khz*2=3KHz).  We're down to seeing about 1 or 2 ops generating harmonics per week now.
#3 If you use Rig Split or Fake It any harmonics you generate are unlikely to be seen.  Better to not generate any as that is some of your power being wasted.
#4 You can't see your own harmonics without another receiver.
#5 Yes -- transmitting above 2400 or so will limit who sees you.  I've worked with many operators with 2KHz or less bandwidth as they didn't understand the effect on decoding.

Mike W9MDB


On Wednesday, October 6, 2021, 04:16:59 PM CDT, Chuck Moore via groups.io <wd4hxg@...> wrote:


While trying to identify an anomaly around 2300 Hz on the waterfall display I was curious about what is above
2500 Hz. The typical grab the edge of the waterfall display with the mouse and pull right did nothing of course.
But after a bit of lifting various rocks I found the control to set the lower edge frequency of the waterfall. As my
rig will allow a 4 KHz wide IF bandwidth I decided to see what was above 2500 Hz and set the lower edge of the
waterfall to 1000 Hz which pushed the upper edge of the water fall to 3500 Hz.  After a few minutes the display
shown below, revealed a number of signals above the default 2500 Hz limit. This was observed while operating
on 17 Meters (18.100 MHz USB).

 


A couple of questions come to mind. (this is not a request for software changes but just a query.)

  • Why is the waterfall limited to an  absolute bandwidth of 2500 Hz?
  • In reading the manual, the recommendation is to only use the 1500 Hz to 2000 Hz audio
    range for transmission to mitigate transmission of audio harmonics. Of course if I open
    my IF Bandwidth to 4 KHz I have just negated the reason for using the specified frequency
    segment. How much of a problem are audio harmonics?
  • Is the audio harmonic content a show stopper if the 1500 Hz to 2000 Hz rule is ignored?
  • Can the waterfall be used to gauge the effects of audio harmonics?
  • It was a bit of a surprise to see the gap from 2300Hz to about 2550 Hz. But more puzzling
    to me was the cluster of ops from about 2550 Hz to 3000 Hz. Would their operation there
    not limit the number of distant ops that can decode the ops operating above 2550 Hz?
    Why stop at 3000 Hz?

Regards

Chuck WD4HXG




Chuck Moore <wd4hxg@...>
 

Evening Mike

#1 Waterfall is not limited to 2.5KHz -- what makes you think it is?  The rig bandwidth is the main limiting factor,.

While I can raise the low end frequency, and the high frequency end rises with a 1:1 ratio, I am
only seeing 2500 Hertz displayed with the Waterfall adjusted to display edge to edge. This is
using  '2 Bins/Pixel'. I can adjust the 'Bins per Pixel' too increase the displayed bandwidth but
being a noob with the software it seemed prudent to leave presets alone. I did raise the 'Bins
per Pixel' and noticed a corresponding rescaling with display of a larger frequency range. I also
set is back to the install default value of '2'.

With the default 2 Bins per Pixel that displayed absolute bandwidth is limited to 3000 Hertz.
The upper frequency cutoff moves up or down depending on where the start frequency for
the left end of the waterfall is set.

#2 Where do see a recommendation to limit to 1500-2000?  There is mention that is what the split modes do. 

When reading the above I interpreted it as the best practice was for the user to select a frequency
for transmission that fell between 1500 and 2000 KHz inclusively in the waterfall. What is escaping
me is how to select a transmission frequency without stepping on other ops.

 A couple of us have been on the war path the last few years to clean up all the operators generating harmonics.  None of them were using split mode as that mode avoids even the 2nd harmonic which could be at a minimum of 3KHz (i.e. 1.5Khz*2=3KHz).  We're down to seeing about 1 or 2 ops generating harmonics per week now.

It is likely I am one of the PITAs. I had not been using split, as up until now it was not clear to
me what it really did. I did however follow the advice to not drive the audio to the point that
the ALC was driven active as displayed on the rigs meter and I keep the audio set to less than
55%.

#3 If you use Rig Split or Fake It any harmonics you generate are unlikely to be seen.  Better to not generate any as that is some of your power being wasted.

#4 You can't see your own harmonics without another receiver.

Understood. Unfortunately my SDRPlay seems to overload in the near field. Guess it is time to
order a 30 dB attenuator or build one from some discrete resistors to pad the front end of the
receiver and get a good look at what is happening locally.

#5 Yes -- transmitting above 2400 or so will limit who sees you.  I've worked with many operators with 2KHz or less bandwidth as they didn't understand the effect on decoding.
I was wondering if there was some reason for operating above 2.4 KHz. I can see where it may
offer a spot where competition for spectrum is not so furious.

Thanks for your and Bill's feedback.

Chuck WD4HXG




On Wednesday, October 6, 2021, 04:16:59 PM CDT, Chuck Moore via groups.io <wd4hxg@...> wrote:


While trying to identify an anomaly around 2300 Hz on the waterfall display I was curious about what is above
2500 Hz. The typical grab the edge of the waterfall display with the mouse and pull right did nothing of course.
But after a bit of lifting various rocks I found the control to set the lower edge frequency of the waterfall. As my
rig will allow a 4 KHz wide IF bandwidth I decided to see what was above 2500 Hz and set the lower edge of the
waterfall to 1000 Hz which pushed the upper edge of the water fall to 3500 Hz.  After a few minutes the display
shown below, revealed a number of signals above the default 2500 Hz limit. This was observed while operating
on 17 Meters (18.100 MHz USB).

 
<dummyfile.0.part>

A couple of questions come to mind. (this is not a request for software changes but just a query.)

  • Why is the waterfall limited to an  absolute bandwidth of 2500 Hz?
  • In reading the manual, the recommendation is to only use the 1500 Hz to 2000 Hz audio
    range for transmission to mitigate transmission of audio harmonics. Of course if I open
    my IF Bandwidth to 4 KHz I have just negated the reason for using the specified frequency
    segment. How much of a problem are audio harmonics?
  • Is the audio harmonic content a show stopper if the 1500 Hz to 2000 Hz rule is ignored?
  • Can the waterfall be used to gauge the effects of audio harmonics?
  • It was a bit of a surprise to see the gap from 2300Hz to about 2550 Hz. But more puzzling
    to me was the cluster of ops from about 2550 Hz to 3000 Hz. Would their operation there
    not limit the number of distant ops that can decode the ops operating above 2550 Hz?
    Why stop at 3000 Hz?

Regards

Chuck WD4HXG












Jim Brown
 

On 10/6/2021 3:00 PM, Michael Black via groups.io wrote:
Better to not generate any as that is some of your power being wasted.
Audio distortion products that fall outside the sideband TX filter do NOT burn TX power. But the war on distortion products is certainly a worthy one. Learning how to properly set audio levels is also quite worthwhile, but may be too much to expect from most hams. There's a tutorial on my website that offers four methods, depending on resources. See page 3 of this link.
http://k9yc.com/USB_Interfaces.pdf

73, Jim K9YC


Bill Somerville
 

On 07/10/2021 01:43, Chuck Moore via groups.io wrote:
 A couple of us have been on the war path the last few years to clean up all the operators generating harmonics.  None of them were using split mode as that mode avoids even the 2nd harmonic which could be at a minimum of 3KHz (i.e. 1.5Khz*2=3KHz).  We're down to seeing about 1 or 2 ops generating harmonics per week now.

It is likely I am one of the PITAs. I had not been using split, as up until now it was not clear to
me what it really did. I did however follow the advice to not drive the audio to the point that
the ALC was driven active as displayed on the rigs meter and I keep the audio set to less than
55%.

Chuck,

this is a misunderstanding of the issue and such advice can easily lead to a very poor transmitted signal. Audio harmonics are generated when the audio input to the rig is too high causing clipping at the modulator input. ALC is not a measure of that issue, it is too late in the system. To understand this consider an audio tone of 3000 Hz being input to the rig, you will not get an ALC indication however high you set the audio input level, this is because the audio tone plus all the generated harmonics from gross over-driving of the rig's audio input will be attenuated by the SSB Tx filter before even reaching the PA  where ALC is measured. Then if you set your Tx offset to 500 Hz then the over-driven input will now mean you will transmit a tone at 500 Hz and harmonics at 1000 Hz, 1500 Hz, 2000 Hz, and 2500 Hz.

The lesson to be learned here is that ALC *may* be used as an indication of audio over-drive, but *only* if the setup is done with a tone around 1500 Hz. Further to that do not *ever* attempt to compensate for low output when using audio offsets near the edges of the rig's SSB Tx filter, particularly the lower bounds!

This pinned post contains recipe for setting the correct Tx audio levels: https://wsjtx.groups.io/g/main/message/13890

73
Bill
G4WJS.


Joe Subich, W4TV
 

On 2021-10-06 8:47 PM, Jim Brown wrote:
Audio distortion products that fall outside the sideband TX filter do
NOT burn TX power.
That *assumes* the transceiver still has a crystal filter in the
transmit path. Many of the newer SDR based designs lack crystal
(or other 2.7 KHz bandwidth) filtering in the transmit path and
rely simply on "audio shaping" in the microphone amplifiers.

Over driving the audio input not only creates distortion in those
circuits, it renders the simple passive audio filtering useless.

73,

... Joe, W4TV


On 2021-10-06 8:47 PM, Jim Brown wrote:
On 10/6/2021 3:00 PM, Michael Black via groups.io wrote:
Better to not generate any as that is some of your power being wasted.
Audio distortion products that fall outside the sideband TX filter do NOT burn TX power. But the war on distortion products is certainly a worthy one. Learning how to properly set audio levels is also quite worthwhile, but may be too much to expect from most hams. There's a tutorial on my website that offers four methods, depending on resources. See page 3 of this link.
http://k9yc.com/USB_Interfaces.pdf
73, Jim K9YC


Chuck Moore <wd4hxg@...>
 

Bill

I appreciate the feedback and guidance. I suspect based on the premise the
harmonic problem is due to overdriving the rigs modulator that there is hope
I am not one of the lids agitating Mike.   :-)  I generally keep the output down
to 30 to 60 watts. The rig is rated for 200 watts out and microphone gain is
reduced to avoid running higher power. Also I will go back to the IF Bandwidth
and throttle it back to a nominal 2.4 KHz with Split turned on. Hopefully that
will keep me off the radar and in good graces of other ops.

I have a couple of friends about 125 miles distant and will ask them to critque
the signal using their water fall display.

Regards

Chuck WD4HXG

On Oct 6, 2021, at 9:01 PM, Bill Somerville <g4wjs@...> wrote:


On 07/10/2021 01:43, Chuck Moore via groups.io wrote:
 A couple of us have been on the war path the last few years to clean up all the operators generating harmonics.  None of them were using split mode as that mode avoids even the 2nd harmonic which could be at a minimum of 3KHz (i.e. 1.5Khz*2=3KHz).  We're down to seeing about 1 or 2 ops generating harmonics per week now.

It is likely I am one of the PITAs. I had not been using split, as up until now it was not clear to
me what it really did. I did however follow the advice to not drive the audio to the point that
the ALC was driven active as displayed on the rigs meter and I keep the audio set to less than
55%.

Chuck,

this is a misunderstanding of the issue and such advice can easily lead to a very poor transmitted signal. Audio harmonics are generated when the audio input to the rig is too high causing clipping at the modulator input. ALC is not a measure of that issue, it is too late in the system. To understand this consider an audio tone of 3000 Hz being input to the rig, you will not get an ALC indication however high you set the audio input level, this is because the audio tone plus all the generated harmonics from gross over-driving of the rig's audio input will be attenuated by the SSB Tx filter before even reaching the PA  where ALC is measured. Then if you set your Tx offset to 500 Hz then the over-driven input will now mean you will transmit a tone at 500 Hz and harmonics at 1000 Hz, 1500 Hz, 2000 Hz, and 2500 Hz.

The lesson to be learned here is that ALC *may* be used as an indication of audio over-drive, but *only* if the setup is done with a tone around 1500 Hz. Further to that do not *ever* attempt to compensate for low output when using audio offsets near the edges of the rig's SSB Tx filter, particularly the lower bounds!

This pinned post contains recipe for setting the correct Tx audio levels: https://wsjtx.groups.io/g/main/message/13890

73
Bill
G4WJS.









Jim Brown
 

On 10/6/2021 6:43 PM, Joe Subich, W4TV wrote:
That *assumes* the transceiver still has a crystal filter in the
transmit path.  Many of the newer SDR based designs lack crystal
(or other 2.7 KHz bandwidth) filtering in the transmit path and
rely simply on "audio shaping" in the microphone amplifiers.
Thanks Joe.

73, Jim K9YC


Jim Brown
 

On 10/6/2021 7:05 PM, Chuck Moore via groups.io wrote:
I generally keep the output down
to 30 to 60 watts. The rig is rated for 200 watts out and microphone gain is reduced to avoid running higher power.
The distortion being discussed here is AUDIO distortion, generated in the audio chain between computer output and the audio stage driving the modulator. It consists of harmonics of the audio. Reducing transmit power reduces RF distortion (harmonics and intermod). Very different.

Also I will go back to the IF Bandwidth and throttle it back to a nominal 2.4 KHz with Split turned on. Hopefully that will keep me off the radar and > in good graces of other ops.
The IF Bandwidth is part of the RECEIVER, has nothing to do with the transmitted signal. In most rigs (except SDRs, as W4TV has pointed out), Transmit filters are fixed crystal filters. In a K3, it's a plug-in 2.7 or 2.8 kHz wide filter, and I use the 2.8, which a bit flatter, because it adds less incidental AM when transmitting RTTY.

73, Jim K9YC