|
I wasn't totally happy with the screen captures I got yesterday when I posted the initial spectrum test with the new SSA 3021x spectrum analyzer so I re-ran them this morning and paid a lot more attention to detail.? I set the fundamental carrier as close to 0db as I could get it (20 meters is the only one that wouldn't quite get there by about 2dB so you'll have to do the math on that one but it is not really pretty. My 2nd V5 board should be here tomorrow (latest Wednesday) as it's left DHL Leipzig, Germany enroute their Cincinnati customs inspection point before continuing to Wichita, KS and as soon as I get it, I'll put it in the test fixture so I can run the same tests on it before building it into a new case.? Someone mentioned open frame coupling to the spectrum analyzer so I covered everything with a metal cooking pan and used short clip leads to ground the pan to the rig.? I did not notice any difference in readings with the shield over the rig vs removed so I left it in place just to be sure. This time I ran through all the bands the stock uBITX covers - 80/60/40/30/20/17/15/12 and 10 meters.? Each plot is labeled by band in the screen shots attached.? None of the pictures are edited and I tried to make sure the markers were close to the right frequency peak with marker 1 as fundamental, 2 as 2nd harmonic, 3 as third and 4 as fourth.? Only have the luxury of 4 on screen markers so the 5th harmonic and beyond if even present in the display will have to be taken for granted. As stated before in yesterday's thread, I do not have a two-tone test oscillator (yet) so no SSB spur pix until I get one.? If someone else with a decent spectrum analyzer wants to do them in my stead, please do so as it will probably be a couple weeks before Pacific Antenna ships me the kit and a couple hours after that to build it. Here are the pix from today's spectrum tests - SSA 3021X Spectrum Analyzer, 50 ohm RF sampler with variable attenuator and terminated by a Microwave Associates 50 ohm, 20 watt dummy load.? No actual power measurements were taken at this time, only checking for harmonic levels related to the fundamental carrier level in dB. Jim Sheldon, W0EB TSW Project Coordinator         
|
Hi Jim
Have you run these tests on any other radios yet. Say your k3? Something that is supposed to be top notch.
Sounds like a neat tool.
Dave
|
Jim, These graphs don't show much more than the earlier ones. 2nd harmonic on 80m and 3rd harmonic on 40m/20m need to be investigated. A low pass filter for each band could fix these but it would be better to find where the distortion is actually coming from. tim ab0wr On Mon, 25 Feb 2019 12:11:29 -0800 "Jim Sheldon" <w0eb@...> wrote: I wasn't totally happy with the screen captures I got yesterday when
I posted the initial spectrum test with the new SSA 3021x spectrum
analyzer so I re-ran them this morning and paid a lot more attention
to detail.? I set the fundamental carrier as close to 0db as I could
get it (20 meters is the only one that wouldn't quite get there by
about 2dB so you'll have to do the math on that one but it is not
really pretty.
My 2nd V5 board should be here tomorrow (latest Wednesday) as it's
left DHL Leipzig, Germany enroute their Cincinnati customs inspection
point before continuing to Wichita, KS and as soon as I get it, I'll
put it in the test fixture so I can run the same tests on it before
building it into a new case.? Someone mentioned open frame coupling
to the spectrum analyzer so I covered everything with a metal cooking
pan and used short clip leads to ground the pan to the rig.? I did
not notice any difference in readings with the shield over the rig vs
removed so I left it in place just to be sure.
This time I ran through all the bands the stock uBITX covers -
80/60/40/30/20/17/15/12 and 10 meters.? Each plot is labeled by band
in the screen shots attached.? None of the pictures are edited and I
tried to make sure the markers were close to the right frequency peak
with marker 1 as fundamental, 2 as 2nd harmonic, 3 as third and 4 as
fourth.? Only have the luxury of 4 on screen markers so the 5th
harmonic and beyond if even present in the display will have to be
taken for granted.
As stated before in yesterday's thread, I do not have a two-tone test
oscillator (yet) so no SSB spur pix until I get one.? If someone else
with a decent spectrum analyzer wants to do them in my stead, please
do so as it will probably be a couple weeks before Pacific Antenna
ships me the kit and a couple hours after that to build it.
Here are the pix from today's spectrum tests - SSA 3021X Spectrum
Analyzer, 50 ohm RF sampler with variable attenuator and terminated
by a Microwave Associates 50 ohm, 20 watt dummy load.? No actual
power measurements were taken at this time, only checking for
harmonic levels related to the fundamental carrier level in dB.
Jim Sheldon, W0EB
TSW Project Coordinator
|
Tim,
Yes, I know they don't. I ran the tests for the second time to make sure I got the test settings as close to right as possible. I didn't expect any seriously different results.
I also deliberately didn't comment on the results of the tests. The data are there and each is free to make their own conclusions.
I'm trying to remain as neutral as possible here, merely reporting MY results with MY test equipment on MY only uBITX V5 board as of right now. The second one will be here tomorrow or Wednesday and I'll run the same tests on that one and neutrally publish the results as soon as possible.
Jim
toggle quoted message
Show quoted text
On Feb 25, 2019, at 3:31 PM, Tim Gorman <tgorman2@...> wrote:
Jim,
These graphs don't show much more than the earlier ones. 2nd harmonic
on 80m and 3rd harmonic on 40m/20m need to be investigated.
A low pass filter for each band could fix these but it would be better
to find where the distortion is actually coming from.
tim ab0wr
On Mon, 25 Feb 2019 12:11:29 -0800
"Jim Sheldon" <w0eb@...> wrote:
I wasn't totally happy with the screen captures I got yesterday when
I posted the initial spectrum test with the new SSA 3021x spectrum
analyzer so I re-ran them this morning and paid a lot more attention
to detail. I set the fundamental carrier as close to 0db as I could
get it (20 meters is the only one that wouldn't quite get there by
about 2dB so you'll have to do the math on that one but it is not
really pretty.
My 2nd V5 board should be here tomorrow (latest Wednesday) as it's
left DHL Leipzig, Germany enroute their Cincinnati customs inspection
point before continuing to Wichita, KS and as soon as I get it, I'll
put it in the test fixture so I can run the same tests on it before
building it into a new case. Someone mentioned open frame coupling
to the spectrum analyzer so I covered everything with a metal cooking
pan and used short clip leads to ground the pan to the rig. I did
not notice any difference in readings with the shield over the rig vs
removed so I left it in place just to be sure.
This time I ran through all the bands the stock uBITX covers -
80/60/40/30/20/17/15/12 and 10 meters. Each plot is labeled by band
in the screen shots attached. None of the pictures are edited and I
tried to make sure the markers were close to the right frequency peak
with marker 1 as fundamental, 2 as 2nd harmonic, 3 as third and 4 as
fourth. Only have the luxury of 4 on screen markers so the 5th
harmonic and beyond if even present in the display will have to be
taken for granted.
As stated before in yesterday's thread, I do not have a two-tone test
oscillator (yet) so no SSB spur pix until I get one. If someone else
with a decent spectrum analyzer wants to do them in my stead, please
do so as it will probably be a couple weeks before Pacific Antenna
ships me the kit and a couple hours after that to build it.
Here are the pix from today's spectrum tests - SSA 3021X Spectrum
Analyzer, 50 ohm RF sampler with variable attenuator and terminated
by a Microwave Associates 50 ohm, 20 watt dummy load. No actual
power measurements were taken at this time, only checking for
harmonic levels related to the fundamental carrier level in dB.
Jim Sheldon, W0EB
TSW Project Coordinator
|
Jim,
Run the test for power out at 28.4mhz? but sweep from 12 to 30mhz.? Above 20mhz harmonics
are less an issue than spurs below 30mhz.? One of the big issues was at 10M there would be
a strong image spur at around 16mhz.
Your latest show the aharmonic issue that was noted for 80 and 40m.? The 20M spike at 28mhz
me be trying to get 10W at 14mhz.? The test there is lower power gradually till the harmonic
drops faster than the power. (that will correspond to the 1DB compression point of the amp).
Generally the amp get dirtier faster if pushed at over 10mhz and 2 to 5W.
I presume your setup has 10W at the 0dbm level.
Allison
|
On the bright side, the 45 MHz into the final mixer spurs seem to be cured on this one sample.
Are the relays same make as v4? I understand they are arranged differently on the schematic.? Maybe like v4 and v3 it might need a mod.
SSB results can be better or worse than these cw measurements.?
Curt
|
Jim, thanks for running through these results and posting them in as unvarnished a way as possible. :) It certainly helps those of us who don't have access to such a nice SA to see how things are behaving.
|
Jim is operating in CW mode, CLK0 and CLK1 shut down, so no 2x45mhz=90mhz harmonic
coming out of the IF amp to create the 16mhz spur when operating on 10m.
That spur check will have to wait for his two tone test generator to arrive.
Yes, the amount of output power he is trying to achieve will be a major factor.
You don't want to push this thing when operating on the upper bands.?
Jerry
toggle quoted message
Show quoted text
On Mon, Feb 25, 2019 at 02:09 PM, ajparent1/KB1GMX wrote:
Jim,
Run the test for power out at 28.4mhz? but sweep from 12 to 30mhz.? Above 20mhz harmonics
are less an issue than spurs below 30mhz.? One of the big issues was at 10M there would be
a strong image spur at around 16mhz.
Your latest show the aharmonic issue that was noted for 80 and 40m.? The 20M spike at 28mhz
me be trying to get 10W at 14mhz.? The test there is lower power gradually till the harmonic
drops faster than the power. (that will correspond to the 1DB compression point of the amp).
Generally the amp get dirtier faster if pushed at over 10mhz and 2 to 5W.
I presume your setup has 10W at the 0dbm level.
Allison
|
you can generate a tone at 1 khz from here:
Once you connect your pc to ubitx, measure the audio at mic input to be 50 mv peak(100 mv peak to peak). Then you can repeat the tests. - f
On Tue, Feb 26, 2019 at 10:44 AM Jerry Gaffke via Groups.Io <jgaffke= [email protected]> wrote: Jim is operating in CW mode, CLK0 and CLK1 shut down, so no 2x45mhz=90mhz harmonic
coming out of the IF amp to create the 16mhz spur when operating on 10m.
That spur check will have to wait for his two tone test generator to arrive.
Yes, the amount of output power he is trying to achieve will be a major factor.
You don't want to push this thing when operating on the upper bands.?
Jerry
On Mon, Feb 25, 2019 at 02:09 PM, ajparent1/KB1GMX wrote:
Jim,
Run the test for power out at 28.4mhz? but sweep from 12 to 30mhz.? Above 20mhz harmonics
are less an issue than spurs below 30mhz.? One of the big issues was at 10M there would be
a strong image spur at around 16mhz.
Your latest show the aharmonic issue that was noted for 80 and 40m.? The 20M spike at 28mhz
me be trying to get 10W at 14mhz.? The test there is lower power gradually till the harmonic
drops faster than the power. (that will correspond to the 1DB compression point of the amp).
Generally the amp get dirtier faster if pushed at over 10mhz and 2 to 5W.
I presume your setup has 10W at the 0dbm level.
Allison
|
toggle quoted message
Show quoted text
you can generate a tone at 1 khz from here:
Once you connect your pc to ubitx, measure the audio at mic input to be 50 mv peak(100 mv peak to peak). Then you can repeat the tests.
- f
On Tue, Feb 26, 2019 at 10:44 AM Jerry Gaffke via Groups.Io <jgaffke= [email protected]> wrote:
Jim is operating in CW mode, CLK0 and CLK1 shut down, so no 2x45mhz=90mhz harmonic
coming out of the IF amp to create the 16mhz spur when operating on 10m.
That spur check will have to wait for his two tone test generator to arrive.
Yes, the amount of output power he is trying to achieve will be a major factor.
You don't want to push this thing when operating on the upper bands.?
Jerry
On Mon, Feb 25, 2019 at 02:09 PM, ajparent1/KB1GMX wrote:
Jim,
Run the test for power out at 28.4mhz? but sweep from 12 to 30mhz.? Above 20mhz harmonics
are less an issue than spurs below 30mhz.? One of the big issues was at 10M there would be
a strong image spur at around 16mhz.
Your latest show the aharmonic issue that was noted for 80 and 40m.? The 20M spike at 28mhz
me be trying to get 10W at 14mhz.? The test there is lower power gradually till the harmonic
drops faster than the power. (that will correspond to the 1DB compression point of the amp).
Generally the amp get dirtier faster if pushed at over 10mhz and 2 to 5W.
I presume your setup has 10W at the 0dbm level.
Allison
|
Farhan,
Under "Binaural beats" on the left pane, we can generate dual
tones.
Raj
At 26-02-19, you wrote:
toggle quoted message
Show quoted text
you can generate a tone at 1 khz
from here:
Once you connect your pc to ubitx, measure the audio at mic input to be
50 mv peak(100 mv peak to peak). Then you can repeat the tests.
- f
|
Wow, test equipment is everywhere. a single tone will get us started. Of course I knew the last mixer is biased away for cw, but easy to forget. In our group build we line them up to test both cw and ssb spurious. Except I am quite occupied now away from my ubitx. Let's see how v5 does on complete chain ssb spurious. I was suspicious it might need some help also in spur compliance.
Curt
|
toggle quoted message
Show quoted text
Farhan,
Under "Binaural beats" on the left pane, we can generate dual
tones.
Raj
At 26-02-19, you wrote:
you can generate a tone at 1 khz
from here:
Once you connect your pc to ubitx, measure the audio at mic input to be
50 mv peak(100 mv peak to peak). Then you can repeat the tests.
- f
|
Jim Sheldon
_Dave_ AD0B
Raj vu2zap