|
Andy G0FTD 3rd Earl of Whitstable MBE BBC ITV etc etc etc etc etc
Just for info, I am having difficulty getting the Python software to run the app.
Same errors as other posters have mentioned.
Upgrade Python on my Linux Mint and I appear to have all the correct additional modules.
Tried the same on my Raspberry Pi, same errors.
Other Python apps appear to work fine.
Not complaining (I think it's a joy to see all the enthusiasm and "overnight" software pop up), so this message is just to let others who
are struggling to know that they are not the only ones ;-)
73 de Andy
|
Hi Andy, could you tell me what error message you are seeing? What command did you run to try and start the software? Since you mention Linux, I assume you are using a git checkout of the source. The repository has some setup information for "pip", which should make it easier to install the prerequisites, which I was sent from a nice ham. I don't have much experience with it, but at a quick try, I was able to use the commands "pip install ." and then just type "NanoVNASaver" in the base directory to run the tool. Other than that, there is a nanovna-saver.py which starts the __main__.py in the NanoVNASaver subdirectory. Running "python nanovna-saver.py" should be able to load this... I've never previously distributed Python apps to other people, so this is all new for me as well :-) I'll accept any help offered in getting things running smoothly for new users. -- Rune / 5Q5R On Thu, 5 Sep 2019 at 13:58, Andy G0FTD 3rd Earl of Whitstable MBE BBC ITV etc etc etc etc etc via Groups.Io <punkbiscuit@...> wrote:
Just for info, I am having difficulty getting the Python software to run
the app.
Same errors as other posters have mentioned.
Upgrade Python on my Linux Mint and I appear to have all the correct
additional modules.
Tried the same on my Raspberry Pi, same errors.
Other Python apps appear to work fine.
Not complaining (I think it's a joy to see all the enthusiasm and
"overnight" software pop up), so this message is just to let others who
are struggling to know that they are not the only ones ;-)
73 de Andy
|
Thanks, Rune. I didn't realize there was an alpha setting in the color-selection window. Good to know!
Playing some more with the 0.0.5 release (and really enjoying it), here are some more comments...
First, a possible bug: I cannot import and display an S1P file as my reference plot (note: this file opens in other apps I have that read S1P files), and when I try this my currently-displayed sweep is then overlayed on itself as the reference sweep. I am attaching the S1P file and a screenshot of the console window. (The number of points and frequency span that the NanoVNA has been set to are the same as the sweep I've made (1-60 MHz, 101 points). Perhaps your import routine is getting confused by the comment lines at the start of the file?)
Additional items for the wish list...
1. If importing a file as reference or sweep, it would be nice to display the file name(s) in the window, too. (I will often use screen-captures of my VNA data to illustrate a point, and having the file names visible in the screen captures makes the data traceable back to the original files, and thus verifiable, should there be a question, months or years later).
2. For the mouse-cursor, also calculate the equivalent series and parallel components. I'm attaching a screenshot of an S1P viewer I often use that includes this feature -- the data is on the right (note that the series reactive component is shown as a negative inductance. This is a bug, in my opinion. It should be shown as a positive capacitance).
3. Somewhere show (or be able to show) the number of points in the sweep.
4. The ability to place multiple independent cursors on either the reference and/or the current sweep, and display info for all cursors, would be nice. Perhaps four or five cursors, total, each of which could be placed on either sweep.
5. By the way -- can the reference sweep be a different number of points than the displayed sweep (just wondering what will happen if I want to compare a NanoVNA sweep with a capture from my 8753C, whose default is 201 points)?
Again, great work!
- Jeff, k6jca
|
Andy G0FTD 3rd Earl of Whitstable MBE BBC ITV etc etc etc etc etc
Hi Rune,
Here's the error messages
$ python3.7 nanovna-saver.py
Traceback (most recent call last):
File "nanovna-saver.py", line 19, in <module>
from PyQt5 import QtWidgets
ModuleNotFoundError: No module named 'PyQt5'
$ python3.7 NanoVNASaver.py
Traceback (most recent call last):
File "NanoVNASaver.py", line 22, in <module>
import numpy as np
ModuleNotFoundError: No module named 'numpy'
Strange thing is, I have the required modules installed, as I needed them for another application.
So I'm not sure why the three required modules appear to be invisible for your Nano VNA software.
Strange...
73 de Andy
|
Hi Andy, the only thing I can think of is possibly having them installed using pip for python 2? Maybe check whether they have been installed with "pip3"? Try looking at pip --version and pip3 --version -- Rune On Thu, 5 Sep 2019 at 15:37, Andy G0FTD 3rd Earl of Whitstable MBE BBC ITV etc etc etc etc etc via Groups.Io <punkbiscuit@...> wrote:
Hi Rune,
Here's the error messages
$ python3.7 nanovna-saver.py
Traceback (most recent call last):
File "nanovna-saver.py", line 19, in <module>
from PyQt5 import QtWidgets
ModuleNotFoundError: No module named 'PyQt5'
$ python3.7 NanoVNASaver.py
Traceback (most recent call last):
File "NanoVNASaver.py", line 22, in <module>
import numpy as np
ModuleNotFoundError: No module named 'numpy'
Strange thing is, I have the required modules installed, as I needed them
for another application.
So I'm not sure why the three required modules appear to be invisible for
your Nano VNA software.
Strange...
73 de Andy
|
Hi Jeff,
thanks for the further suggestions - the list is getting long :D
Wrt. the import of this particular file, I currently *only* have importing
of S-parameters in real-imaginary form normalized to 50 ohms, with
frequencies in Hz. I'm sure setting up other imports would be possible, but
it's not at the top of the list. Your file is in magnitude-angle format,
which my software doesn't understand yet.
--
Rune / 5Q5R
toggle quoted message
Show quoted text
On Thu, 5 Sep 2019 at 15:31, Jeff Anderson <jca1955@...> wrote: Thanks, Rune. I didn't realize there was an alpha setting in the
color-selection window. Good to know!
Playing some more with the 0.0.5 release (and really enjoying it), here
are some more comments...
First, a possible bug: I cannot import and display an S1P file as my
reference plot (note: this file opens in other apps I have that read S1P
files), and when I try this my currently-displayed sweep is then overlayed
on itself as the reference sweep. I am attaching the S1P file and a
screenshot of the console window. (The number of points and frequency
span that the NanoVNA has been set to are the same as the sweep I've made
(1-60 MHz, 101 points). Perhaps your import routine is getting confused by
the comment lines at the start of the file?)
Additional items for the wish list...
1. If importing a file as reference or sweep, it would be nice to display
the file name(s) in the window, too. (I will often use screen-captures of
my VNA data to illustrate a point, and having the file names visible in the
screen captures makes the data traceable back to the original files, and
thus verifiable, should there be a question, months or years later).
2. For the mouse-cursor, also calculate the equivalent series and
parallel components. I'm attaching a screenshot of an S1P viewer I often
use that includes this feature -- the data is on the right (note that the
series reactive component is shown as a negative inductance. This is a
bug, in my opinion. It should be shown as a positive capacitance).
3. Somewhere show (or be able to show) the number of points in the sweep.
4. The ability to place multiple independent cursors on either the
reference and/or the current sweep, and display info for all cursors, would
be nice. Perhaps four or five cursors, total, each of which could be
placed on either sweep.
5. By the way -- can the reference sweep be a different number of points
than the displayed sweep (just wondering what will happen if I want to
compare a NanoVNA sweep with a capture from my 8753C, whose default is 201
points)?
Again, great work!
- Jeff, k6jca
|
Andy G0FTD 3rd Earl of Whitstable MBE BBC ITV etc etc etc etc etc
Hi Rune,
Python version here is 3.7.4 (same as yours if I remember correctly).
I recently installed the extra modules for Python 3 for another app, because I too had
a default of Python 2, so I knew there had to be two versions installed.
So to the best of my knowledge, all versions and modules should be correct.
Maybe I'm missing something, so don't rely on me 100% ;-)
I'll experiment in the background over the coming weeks.
It's not a big priority for me, since the reason I purchased a Mini VNA was because I wanted
to dump the PC.
But a friend has asked if it would work with PC so I thought I'd experiment on his behalf.
I hope that makes sense.
Keep up the good work.
73 de Andy
|
Andy G0FTD 3rd Earl of Whitstable MBE BBC ITV etc etc etc etc etc
Just for info, I used the command to list modules, still doesn't show them.
So I used the pip installer to try again..
Nope, still doesn't work.
I thought these languages like Python were suppose to make things portable ? ;-)
pip3 install PyQt5
Collecting PyQt5
Downloading (62.1MB)
100% |¨€¨€¨€¨€¨€¨€¨€¨€¨€¨€¨€¨€¨€¨€¨€¨€¨€¨€¨€¨€¨€¨€¨€¨€¨€¨€¨€¨€¨€¨€¨€¨€| 62.1MB 23kB/s
Collecting PyQt5_sip<13,>=4.19.14 (from PyQt5)
Downloading (67kB)
100% |¨€¨€¨€¨€¨€¨€¨€¨€¨€¨€¨€¨€¨€¨€¨€¨€¨€¨€¨€¨€¨€¨€¨€¨€¨€¨€¨€¨€¨€¨€¨€¨€| 71kB 2.2MB/s
Installing collected packages: PyQt5-sip, PyQt5
Successfully installed PyQt5 PyQt5-sip
$ pip3 install numpy
Collecting numpy
Downloading (20.1MB)
100% |¨€¨€¨€¨€¨€¨€¨€¨€¨€¨€¨€¨€¨€¨€¨€¨€¨€¨€¨€¨€¨€¨€¨€¨€¨€¨€¨€¨€¨€¨€¨€¨€| 20.2MB 75kB/s
Installing collected packages: numpy
Successfully installed numpy-1.17.1
$ pip3 install pyserial
Collecting pyserial
Downloading (193kB)
100% |¨€¨€¨€¨€¨€¨€¨€¨€¨€¨€¨€¨€¨€¨€¨€¨€¨€¨€¨€¨€¨€¨€¨€¨€¨€¨€¨€¨€¨€¨€¨€¨€| 194kB 947kB/s
Installing collected packages: pyserial
Successfully installed pyserial-3.4
$ python3.7 nanovna-saver.py
Traceback (most recent call last):
File "nanovna-saver.py", line 19, in <module>
from PyQt5 import QtWidgets
ModuleNotFoundError: No module named 'PyQt5'
$ python3.7 NanoVNASaver.py
Traceback (most recent call last):
File "NanoVNASaver.py", line 22, in <module>
import numpy as np
ModuleNotFoundError: No module named 'numpy'
73 de Andy
|
On Thu, Sep 5, 2019 at 06:49 AM, Rune Broberg wrote:
thanks for the further suggestions - the list is getting long :D
Yes -- didn't mean for it to get so long. They are only suggestions, offered in case you hadn't considered them. But it's your software. Whatever you do, or don't do, is fine with me.
Wrt. the import of this particular file, I currently *only* have importing
of S-parameters in real-imaginary form normalized to 50 ohms, with
frequencies in Hz. I'm sure setting up other imports would be possible, but
it's not at the top of the list. Your file is in magnitude-angle format,
which my software doesn't understand yet.
Understood re: mag, phase. Thanks. I've just tried loading a different file that's in real-imaginary format, and it is still failing. The only difference I see between it and an s1p file I exported from your program is the inclusion of the comment statements in the former s1p file and the larger amount of space (tabs?) between data entries. Anyway -- for your reference. Failing file is attached. - Jeff, k6jca
|
That one was in MHz ;-) The software is quite fuzzy at the moment. But
thanks for the example files, they should allow me to work on the format
parsing.
--
Rune
toggle quoted message
Show quoted text
On Thu, 5 Sep 2019 at 16:35, Jeff Anderson <jca1955@...> wrote: On Thu, Sep 5, 2019 at 06:49 AM, Rune Broberg wrote:
thanks for the further suggestions - the list is getting long :D
Yes -- didn't mean for it to get so long. They are only suggestions,
offered in case you hadn't considered them. But it's your software.
Whatever you do, or don't do, is fine with me.
Wrt. the import of this particular file, I currently *only* have
importing
of S-parameters in real-imaginary form normalized to 50 ohms, with
frequencies in Hz. I'm sure setting up other imports would be possible, but
it's not at the top of the list. Your file is in magnitude-angle format,
which my software doesn't understand yet.
Understood re: mag, phase. Thanks.
I've just tried loading a different file that's in real-imaginary format,
and it is still failing. The only difference I see between it and an s1p
file I exported from your program is the inclusion of the comment
statements in the former s1p file and the larger amount of space (tabs?)
between data entries.
Anyway -- for your reference. Failing file is attached.
- Jeff, k6jca
|
Hi Rune,
thanks for your excellent program. When running I get some warnings - see screenshot enclosed.
Furthermore I can not activate the mouse cursor in s11 window. Left or right click in this window does not do anything.
Ernst
|
Sorry, forgot the screenshots:
|
Hello Ernst,
the warnings are normal - sadly the data from the NanoVNA sometimes arrives
corrupted, and in those cases, the software re-tries reading the data. I
haven't yet figured out *why* it arrives corrupted: If anyone can figure it
out, please do let me know.
Cursor control is currently only possible on the charts that have a
frequency axis.
Thanks for the praise, and for trying out the software :-)
--
Rune / 5Q5R
toggle quoted message
Show quoted text
On Thu, 5 Sep 2019 at 16:58, dk1vi <dk1vi@...> wrote: Sorry, forgot the screenshots:
|
S11 Return Loss has frequency axis but does not work here.
|
Thank you Rune for this information. I will try this again. Bob
|
In a completely unsurprising twist, 0.0.6 is now out: It adds the option of having more or fewer charts displayed at once, as well as selecting which ones are displayed. The title bar of the application now displays the source(s) of sweep and reference, and the number of points in each. Markers now display what I think is called a reactive equivalent, in nH or nF. The project remains open to contributions from anyone, particularly those skilled in Python (who might teach me a thing or two! ;-) I hope you all get some enjoyment out of it :-) -- Rune / 5Q5R On Thu, 5 Sep 2019 at 17:02, Rune Broberg via Groups.Io <mihtjel= [email protected]> wrote: Hello Ernst,
the warnings are normal - sadly the data from the NanoVNA sometimes arrives
corrupted, and in those cases, the software re-tries reading the data. I
haven't yet figured out *why* it arrives corrupted: If anyone can figure it
out, please do let me know.
Cursor control is currently only possible on the charts that have a
frequency axis.
Thanks for the praise, and for trying out the software :-)
--
Rune / 5Q5R
On Thu, 5 Sep 2019 at 16:58, dk1vi <dk1vi@...> wrote:
Sorry, forgot the screenshots:
|
Small bug: when sweeping from >500M to 900M the mouse cursor does not work. It does work when sweeping from 50k to 900M.
|
Thanks for the bug report! I pushed a fix for it to the master branch: I
had erroneously disregarded the starting frequency when determining where
to place the marker. A sweep with a span smaller than the starting
frequency would likely cause the cursor to stop working.
--
Rune
toggle quoted message
Show quoted text
On Thu, 5 Sep 2019 at 22:17, Wim <wim_huyghe@...> wrote: Small bug: when sweeping from >500M to 900M the mouse cursor does not
work. It does work when sweeping from 50k to 900M.
|
Hi Rune,
Great! Some comments on v0.0.6:
1. S21 is a Transmission measurement. As such, Smith Charts and SWR (which depend upon reflections) are meaningless, so you can eliminate "S21 Smith Chart" and "S21 VSWR" plots. (I would also recommend changing "VSWR" to "SWR", which is the more general term)
2. Glad to see S21 Phase plot added. Two notes:
A. In the S21 fields for the two markers and for the cursor, include S21 "phase" in addition to "gain", which is already there.
B. When viewing S21 phase, it can be useful to "unwrap" it if it goes beyond -180 or +180 degrees (e.g. for determining Group Delay). To illustrate this concept, I've attached a screen capture of a Matlab S2P file display-app with the phase unwrapped (plot in the lower left-hand corner). You can see that the phase goes down to -600 degrees. Note, too, that the phase displayed at the two cursors is NOT unwrapped (not sure why it isn't).
Best regards,
- Jeff, k6jca
|
And as usual, something came to mind just after I clicked "Reply"...
3. It is nice to have a displayed field into which one can enter notes regarding what is being measured. Refer to the box in the right-hand corner in the S2P file-display image attached to my previous reply.
- Jeff
|
