|
Hi all
I updated the SSB beta testing webpage of the brave, here:? and released firmware version 1_01_001 to it. These are the changes: - As George Korper pointed out to me, the SWR protection wasn't working for SSB modes. Apparently I had overlooked that in my massive "integration", making sure all parts of the QMX firmware?play nice with the new SSB, CESSB and Calibration modules. So I have fixed that, SWR protection now works for SSB!?
- As many people pointed out, CAT PTT via the TX; command did not work. You had to use VOX. I checked and realized that I too had been using VOX on my QMX+ while making my first ever Winlink emails yesterday. So I have fixed this too, TX; works fine now.?
- There was a report of a blown PA transistor(s) during Calibration. The phase pre-distortion measurement ramps up the PA from near zero to full power in DAC steps of 50. There can be up to 60 points in the vector but for 12V supply typically about 50 points are measured. Each measurement takes about 1.6 seconds (I will explain the technical details, somewhat later, about how I achieved this measurement). Previously there was 10 milliseconds of key-up in between tests. I have now changed that so that for DAC values 1200 and above (corresponding to about 6V PA voltage), a delay is inserted which is equivalent to the DAC value. For example as you get to larger values like 2000 for the DAC, that would cause a 2 second key-down between measurements. This will prevent overheating in the PA transistors.?
73 Hans G0UPL
|
Great to see the first fixes coming in so quick. That might have been the most important ones. Thank you for the changelog! Did you provide an 1.01.002 version?
?
?
|
My bad, versioning started at zero ?
|
My bad, versioning started at zero ?
Yeah. Ya always know a C programmer...
|
On 19/03/2025 08:17, Hans Summers via groups.io wrote: My bad, versioning started at zero ?
Yeah. Ya always know a C programmer...
Those of us who wrote software in BASIC on the amazing (for the time) HP-85 computer are well used to starting programs with 10 OPTION BASE 0 so that the array index maths looked correct. :-)
|
On Wed, Mar 19, 2025 at 02:50 AM, Hans Summers wrote:
There was a report of a blown PA transistor(s) during Calibration. The phase pre-distortion measurement ramps up the PA from near zero to full power in DAC steps of 50. There can be up to 60 points in the vector but for 12V supply typically about 50 points are measured. Each measurement takes about 1.6 seconds (I will explain the technical details, somewhat later, about how I achieved this measurement). Previously there was 10 milliseconds of key-up in between tests. I have now changed that so that for DAC values 1200 and above (corresponding to about 6V PA voltage), a delay is inserted which is equivalent to the DAC value. For example as you get to larger values like 2000 for the DAC, that would cause a 2 second key-down between measurements. This will prevent overheating in the PA transistors.?
Hans,
?
For my understanding, did you mean to say there is a "2 second delay between key down?"? I read the above as the transmitter is kept on for 2 seconds.
?
73
Evan
AC9TU
|
Please, what is the setup for digimodes (Ardop, VARA)? Can I leave the output level at -0 dB and the firmware will equalize it or would I have to check (how?) the correct modulation drive level?
?
73 Jindra OK4RM
|
Please, what is the setup for digimodes (Ardop, VARA)? Can I leave the output level at -0 dB and the firmware will equalize it or would I have to check (how?) the correct modulation drive level?
Referring to the diagram at?, the equalization, AGC, compression etc applies to microphone input only. Audio from the USB interface is modulated directly. So an output level of 0dB would be correct (full volume from the PC).?
|
Maybe a detail, but it seems that the menu is not always updated properly. When you cycle trough the options by pressing down and up, sometimes options are not drawn on the terminal.
This is fixed by pressing another key down and it happens ~ 1 out of 5 presses. No big deal, but wanted to share for the bug backlog ;)
?
  Looking forward to testing! PA3MAT, Mathijs
|
Maybe a detail, but it seems that the menu is not always updated properly. When you cycle trough the options by pressing down and up, sometimes options are not drawn on the terminal.
This is fixed by pressing another key down and it happens ~ 1 out of 5 presses. No big deal, but wanted to share for the bug backlog ;)
This happens to me sometimes too. I am not sure why it happens, but it is not new, and I've seen it on QDX too.??
73 Hans G0UPL
|
|
|
A question on the phase error calculation.? The plot on 20m here is very different to the other bands:
The 17m one looks similar to the beta firmware description.
Also I'm seeing a short burst of noise on key up until some voice modulation starts, maybe this is the noise gate parameter?
?
Absolutely amazing work Hans!??
?
Glenn VK3YY.
|
And we are in 2025 ;)
Hi hi, so we are, I will fix that too! Luckily we didn't go any further into 2026, 2027 etc before SSB came to the world...?
|
A question on the phase error calculation.? The plot on 20m here is very different to the other bands:
The 17m one looks similar to the beta firmware description.
I don't have an explanation for this. You could re-run it and see if it is repeatable? Does your 20m performance seem otherwise OK? Did you do anything particular to 20m such as change the defaults parameters in the band configuration screen?? ? Also I'm seeing a short burst of noise on key up until some voice modulation starts, maybe this is the noise gate parameter?
That will need some attention from me. Maybe some buffers need initializing; the SSB subsystem needs to be cranked up gradually or something to avoid some initial burst of nonsense... I will check it.?
73 Hans G0UPL
|
Hi Hans,
?
I ran the 20m phase error test about 5 times with power down between with the same or similar plot.? The band configurations are default on the high band QMX.? 20m performance as far as power output in CW mode is fine.? RX is fine as well.? The filters look good.
I had a listen to the audio on 20m and it didn't sound as good as 17m.? I'll do some more controlled tests tomorrow.
The 20m IMD sync optimisation looks a lot different to 17m as well.
?
Cheers,
Glenn VK3YY.
?
|
There was a report of a blown PA transistor(s) during Calibration. The phase pre-distortion measurement ramps up the PA from near zero to full power in DAC steps of 50. There can be up to 60 points in the vector but for 12V supply typically about 50 points are measured. Each measurement takes about 1.6 seconds (I will explain the technical details, somewhat later, about how I achieved this measurement). Previously there was 10 milliseconds of key-up in between tests. I have now changed that so that for DAC values 1200 and above (corresponding to about 6V PA voltage), a delay is inserted which is equivalent to the DAC value. For example as you get to larger values like 2000 for the DAC, that would cause a 2 second key-down between measurements. This will prevent overheating in the PA transistors.
For my understanding, did you mean to say there is a "2 second delay between key down?"? I read the above as the transmitter is kept on for 2 seconds.
My apologies for the confusion.?
Each step takes 1.6 seconds of trnamsisison (approximately). There is now a delay between transmissions to reduce the duty cycle and thereby transistor heating. The delay is active for DAC value 1200 and up (x-axis) and is equal to the DAC value in milliseconds. So for DAC value 2000 there would be a 2 second delay between measurements. So a much reduced duty cycle.? 73 Hans G0UPL
|
Hi Hans,
First of all, congratulations on this fantastic update and the work and time you've invested in the project. I'm always amazed by the great solutions you find for all these things.
I think I've found another small bug with the drive level setting in the SSB menu.
Changing the value only takes effect after you've completely exited the configuration interface, not immediately as expected. I've reproduced this in several tests.
73 Severin DL9SW
|
Hans,
?
Thank you for the clarification. The QMX+ is a great kit!
?
73
Evan
AC9TU
|
It looks like the SWR protection for SSB went a bit past the goal post in 1_01_001: It looks like it's effective even when turned off in the menu. I have mine turned off when I am at home. I have an RF sensing SDR switch, and that presents a spit second of a mismatch that triggers the SWR protection in the QMX. Now when transmitting in SSB, it still engages the SWR protection, even though it is turned off. To make sure it's not something that can be cured by turning it on and off again, I did that, but still with the same results: SWR protection is active, even though it is deactivated in the menu.?
?
--
Karl Heinz - K5KHK
|
The plot on 20m here is very different to the other bands
?
|
Glenn S
