NEW: Spectrum plots are now available on Personal Space WX Network

126 views
Skip to first unread message

Bill Engelke

unread,
Nov 8, 2023, 2:06:45 PM11/8/23
to hamsci...@googlegroups.com
QST HamSCI:

The  https://pswsnetwork.caps.ua.edu  website has a new feature. Every time  your Grape1_DRF system uploads its data, a waterfall plot of the spectrum will be automatically generated. You can see the plots by going to the Observations tab, then clicking on the link in the File/Observation column. 

A few details:
  • When you click on the Observation link, you will see the Download page with the plot on the bottom. If you wish to see it at full resolution, right click on it and open it in a new tab.  Click on the new tab to see the full size plot.
  • You can still download the original data using the Download button.
  • We are running plots for historical data going back to May of this year (it will take a while for this to complete).
  • The new plots use a color scheme similar to the gnuradio waterfall shown with the Grape1_DRF system.
  • We will be adding a button that will let you download the plot png file.
  • If a data file is missing some data (e.g., because the Grape was down for a while), that will show as white space.
  • All plots and data (from all users) are available to all users. The data is public.
  • Example below.



-73-  Bill    AB4EJ

Charles (Chuck) Smyder

unread,
Nov 8, 2023, 2:19:38 PM11/8/23
to hamsci...@googlegroups.com
That is awesome. I’ve been a bit concerned about my HF station causing interference of receiver de-sense while running QRO. So now I can examine the plot corresponding to my operating time. 

Looking at the plots from 11/5 for my station & others, I’m surmising that the dip from 10-12UTC was the arrival of the CME this past weekend.

-Chuck

<image.png>


-73-  Bill    AB4EJ

-- 
You received this message because you are subscribed to the Google Groups "hamsci-grape" group.
To unsubscribe from this group and stop receiving emails from it, send an email to hamsci-grape...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/hamsci-grape/BN8PR15MB3185B9D20FE19043D977FBB4FAA8A%40BN8PR15MB3185.namprd15.prod.outlook.com.

Graham c

unread,
Nov 8, 2023, 2:21:53 PM11/8/23
to Bill Engelke, hamsci...@googlegroups.com
Very nice.

Any chance there will be a user "how to" cookbook on how to read and plot or otherwise access the drf files without having to wade through the Haystack documentation?

I also see that  my Grape1-drf system has been off line since 2023-11-03 late in the day. Seems to be running and collecting data, just not uploading.

I recall that was the day we had several short power outages which resulted in the failure of my Grape1-FLDIGI system which is also currently offline and will remain so but soon to be replaced with another Grape1-drf system.

I must go sort out the issues.

cheers, Graham

Bill Engelke

unread,
Nov 8, 2023, 2:25:03 PM11/8/23
to Charles (Chuck) Smyder, hamsci...@googlegroups.com
Chuck:
We recognize that we have an intermittent desensing problem with the Grape1_DRF system, which seems to be associated with periods of very good propagation. The leading theory on what is causing this is that the first one or two processing blocks ion the gnuradio grc flowgraph are mishandling the sound card data when the signal level exceeds a certain threshold. If someone who has a calibrated signal generator could help with this (by gradually increasing the input level until desensing occurs) we possibly could put or finger on this. Maybe there is too much gain somewhere.  -73- AB4EJ

From: hamsci...@googlegroups.com <hamsci...@googlegroups.com> on behalf of Charles (Chuck) Smyder <nfl...@gmail.com>
Sent: Wednesday, November 8, 2023 1:19 PM
To: hamsci...@googlegroups.com <hamsci...@googlegroups.com>
Subject: [EXTERNAL] Re: [hamsci-grape] NEW: Spectrum plots are now available on Personal Space WX Network
 

Bill Engelke

unread,
Nov 8, 2023, 2:29:01 PM11/8/23
to Graham c, hamsci...@googlegroups.com
Graham - I will contact AF8A and see if we can find someplace where I can post the code to do the plots. (One headache on this is that you have to install several python libraries for this plotter to run, and  your success with that will vary according to what type of system you are using).   -73  -  AB4EJ

From: Graham c <colonel...@gmail.com>
Sent: Wednesday, November 8, 2023 1:21 PM
To: Bill Engelke <bill.e...@ua.edu>
Cc: hamsci...@googlegroups.com <hamsci...@googlegroups.com>

Subject: [EXTERNAL] Re: [hamsci-grape] NEW: Spectrum plots are now available on Personal Space WX Network

Graham c

unread,
Nov 8, 2023, 4:49:56 PM11/8/23
to Bill Engelke, hamsci...@googlegroups.com
Fair enough.

cheers, Graham ve3gtc

Gary Mikitin, AF8A

unread,
Nov 9, 2023, 7:58:08 PM11/9/23
to hamsci-grape
Hello All - Here is a link to the new HamSCI/DRF_processing GitHub site.  This site hosts the Python code that reads and plots the Grape 1 DRF files.    https://github.com/HamSCI/DRF_processing

Thanks to Bill, AB4EJ, for the above - I'm just the messenger <grin>.

73 de Gary, AF8A

Graham c

unread,
Nov 17, 2023, 8:49:22 AM11/17/23
to hamsci-grape
Good day all,

I have been viewing the daily spectrum plots created and available through the pswsnetwork web site.

When viewed on a Windows (Win11) PC using Chrome web browser, the web page view properly.

When viewed on a Linux PC using Chrome web browser, the web page is mostly correct but is cut off along the left side of the page.  The page is cutoff right around the left hand edge of the spectrum plot images themselves. 

Firefox and Microsoft Edge on Windows ( Win11 ) are both OK and display the same as Chrome.

I have not tried any browser on Linux other than Chrome ( yet ).

cheers, Graham ve3gtc

Randy Allen

unread,
Nov 17, 2023, 9:29:00 AM11/17/23
to hamsci-grape
The same 'cutoff' occurs on Chrome on a Mac OSX system.  Just another data point.

73, N4TVC, Randy

Graham c

unread,
Feb 24, 2024, 12:23:20 PM2/24/24
to hamsci-grape
I have finally gotten around to trying to use the python app for plotting Grape1-drf data but have run into a snag.

On WIN10 / WIN11 plotspectrum_V4a.py produces an error as it needs digital_rf in order to work.

But, when I try to pip install digital_rf  I get this error:

digital_rf.h(46): fatal error C1083: Cannot open include file: 'hdf5.h': No such file
or directory

I have tried pip install h5py which proceeds without error but I still get the above error when trying to install digital_rf

I have downloaded hdf5 from https://www.hdfgroup.org/downloads/hdf5/ ,  which installed without error. I first tried installing in c:\program files  ( sometimes Windoze folder and file names with spaces cause issues so I also uninstalled from c:\program files  and re-installed in c:\ but with the same issues.

I have searched the HDF forum but have not stumbled upon the correct solution (yet).

Seems like a path issue of some sort which I just can't seem to put my finger on.

Anyone have any suggestions?

I will be trying later on a Linux machine but first wanted to try on Windoze.

cheers, Graham ve3gtc

John Gibbons

unread,
Feb 24, 2024, 1:04:57 PM2/24/24
to Graham c, hamsci-grape
It was originally written for LINUX - so go to windows at your own peril...

My suggestion would be to just go back to linux - but I HATE Windoze with a passion so my opinion 
should probably be avoided. 

John N8OBJ


Graham c

unread,
Feb 24, 2024, 4:47:41 PM2/24/24
to John Gibbons, hamsci-grape
John,

You are preaching to the choir.  My dislike of Windoze grows with every passing day; however, it is necessary evil which I must deal with.

cheers, Graham

John Gibbons

unread,
Feb 24, 2024, 4:58:49 PM2/24/24
to Graham c, hamsci-grape
Graham,

LOL!  If people understood better what an OS is supposed to provide for them MicroSux would go out of business
overnight.   Bit I digress....

soapbox mode off....

John



Graham c

unread,
Feb 24, 2024, 5:47:31 PM2/24/24
to hamsci-grape
Bill ( et al ),

I have the drf_processing python code running on ubuntu 22.04lts.

However ( there always seems to be a however ... ) I encountered an error and have raised an issue on github.

For anyone else wishing to use this code, these are the details to save you having to find it on github:

plotspectrum_V4a.py", line 126, in


bigarray = np.zeros(1474560,dtype=np.complex)

File "/numpy/init.py", line 324, in getattr
raise AttributeError(former_attrs[attr])

AttributeError: module 'numpy' has no attribute 'complex'.

np.complex was a deprecated alias for the builtin complex. To avoid this error in existing code, use complex by itself. Doing this
will not modify any behavior and is safe. If you specifically wanted the numpy scalar type, use np.complex128 here.
The aliases was originally deprecated in NumPy 1.20; for more details and guidance see the original release note at:
https://numpy.org/devdocs/release/1.20.0-notes.html#deprecations. Did you mean: 'complex_'?

line 126 changed to

bigarray = np.zeros(1474560,dtype=np.complex128)

and all appears to OK.

cheers, Graham ve3gtc



On Sat, Feb 24, 2024 at 1:04 PM John Gibbons <jc...@case.edu> wrote:

Graham c

unread,
Feb 25, 2024, 2:33:53 PM2/25/24
to hamsci-grape
After some time tweaking and wrangling with matplotlib I have an initial mod version of Bill's plotting script.

Now I can plot any day of data at any time that exists on my Grape1-drf system, for example, a plot for today 2025-02-25 made a short time ago:
S000069_59_2024-02-25T00-00_FN25ig_annotated_20250225_1910utc.png 
Which brings me to several questions regarding the data,

1) I can infer the frequency scale of the data as +/- 5 Hz in spectrum plot Y axis from the plots that can be found on the PSWS web site and from the results of using Bill's example code to plot the data. Where does the +/- 5 Hz come from?   I scanned the gnu_radio scipt and nothing stood out (unless I just didn't see it).

2) Well, I guess there is only one question for now.

I will tweak my code some more and post on github if anyone is interested. I want to better understand the digital rf data and matplot some more. I tend to fall back to gnuplot but seeing as matplot is the preference here I will adapt. I would like to make the plotting script a bit more universal so that I can specify at the least a time range to plot rather than just the whole day. 

Fussing over setting signal levels to the sound card led me here. I have made some observations on that and will post in that topic.

cheers, Graham ve3gtc



Jonathan

unread,
Feb 26, 2024, 6:12:52 AM2/26/24
to hamsci-grape
Graham,

+/- 5 Hz comes from a 10 Hz bandwidth window that shows normal doppler shift in the carrier and is set in the GNU radio graph.

Jonathan
KC3EEY

--
You received this message because you are subscribed to the Google Groups "hamsci-grape" group.
To unsubscribe from this group and stop receiving emails from it, send an email to hamsci-grape...@googlegroups.com.
Reply all
Reply to author
Forward
0 new messages