Grape1-vlfrx-tools

127 views
Skip to first unread message

Graham c

unread,
Jan 18, 2024, 12:41:18 PM1/18/24
to hamsci-grape
Sometime ago Jonathan KC3EEY suggested the use of Paul Nicholson's fabulous vlfrx-tools set for collecting doppler data on WWV/CHU/(etc).

Always liking a challenge and have a liking for Linux and command line tools, I had a go.

I cleaned up an older Dell Optiplex 380 SFF ( small form factor ) desktop, loaded Ubuntu 22.04lts, and the latest version of vlfrx-tools http://abelian.org/vlfrx-tools/

For a bit of background, I currently have the PC set up using three sound cards, the internal soundcard being used by two instances of WSJT-X ( on 20m, on instance WSPR and the other FST4W ), an external E-MU Pre USB soundcard sampling VLF at 192kHz. and a small Plugable ( 48k sampling MONO ) brand USB soundcard capturing WWV 15MHz doppler data form a split audio feed that is also feeding my Grape1-DRF system. I have created a number of bash and gnuplot scripts to help tie the various vlfrx-tools together to create a couple of data files containing the measured data.

That older PC handles all of these tasks reliably and without complaint.

Recently, I took the time to create a simple gnuplot script to plot the doppler data into a similar form that is familiar to the Grape1-fldigi and Grape1-drf data plots. There are the are the plots from yesterdays ( 2024-01-17 ) data, top one is from my Grape1-vlfrx-tools system and the second from my Grape1-drf system. Like the Grape1-fldigi, the plot from my Grape1-vlfrx-tools is simple minded in that only the doppler value having the greatest amplitude is plotted whereas the Grape1-drf data and plots provide more detail. I have also been keeping the results of the FFT processing so that when time permits I will be able to create a similar visual plot from my Grape1-vlfrx-tools data as that created from the Grape1-drf data.  If the order of the images get out of sorts, the "green" plot is from the Grape1-drf data.

cheers, Graham ve3gtc

 


S000069_59_2024-01-17T00-00_FN25ig.png
wwv_15MHz_FN25ig_20240117.png

Jonathan

unread,
Jan 18, 2024, 1:12:00 PM1/18/24
to hamsci-grape
Hi Graham,

Absolutely excellent results!! 

This is the first time that Grape 1 data’s timing accuracy is within hundreds of nanosecond due to sample rate and alignment calibration utilizing a GPS PPS!!

The next thing to try is the compressed spectrogram feature in vtsid. You simply set the upper and lower limits in your sid.conf.

For those of you who might not be aware, vlfrx-tools is an open source toolset for signal processing and has many applications in VLF monitoring, natural radio emissions, and sferic/lightning location and analysis. 

The immediate benefit to using vlfrx-tools is that the sample rate and sample alignment are calibrated to UT time based on the GPS PPS, allowing you to get hundreds to tens of nanoseconds of accuracy of timestamp and sample alignment using a decent soundcard and GPS. This is especially important with sferic analysis and lightning location and is currently utilized by the Indian Lightning Detection Network. 

Another benefit is the data format has built in timestamping which can aid in analysis. Data can plotted like in the attachments and archived. It can even be streamed over the internet either uncompressed or vorbis compressed and analysis done on other servers, with a much smaller data footprint. It’s extremely scriptable with shell scripting and runs on both the individual station and central processing server.

This is great alternative software that can be used and the only one to provide accurate sample and timestamp alignment without having to worry about variations in the sound card clock. 

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.
To view this discussion on the web visit https://groups.google.com/d/msgid/hamsci-grape/acf491fa-6c53-4f3a-b10d-259ae7d8b6f7n%40googlegroups.com.

Jonathan

unread,
Mar 16, 2024, 4:45:20 PM3/16/24
to hamsci-grape
Graham,

Have you had a chance to use vtsid's compressed spectrogram function to create diurnal spectrograms of the data?

Thanks.

Jonathan
KC3EEY

Graham c

unread,
Mar 18, 2024, 9:44:00 AM3/18/24
to Jonathan, hamsci-grape
Good morning Jonathan,

Unfortunately no, not yet - well, no not really.  I have played a bit but I have gotten lost down the bunny hole playing around and testing many of the features of Paul's vlfrx-tools and have been spending far too much time cleaning up my rats nest of old messy cables and testing sound cards, etc.

I was hoping that I would have my vlf monitoring set up for the eclipse but that too has fallen to the wayside.

I have also gotten involved with collecting data on CHU for the group at Case Western and their eclipse project.

Seems I have, as I usually do, spread myself a bit thin ;)

Additionally, the PPS source I was using for discipling the vlfrx-tools sound card input has been repurposed.  It was seen that providing a more precise PC time source through NTP/chrony should be my first priority. My GPS clock's ( homebrew arduino based with uBlox Lea5 GNSS module providing two direct from GNSS module buffered PPS outputs ) PPS outputs have been used to discipline two PC's leaving me none for the moment to feed to vlfrx-tools.

I am hoping that within the next couple of days I can return to vlfrx-tools and set up to at least record raw data ( i.e. vtwrite ) for some time period before and after the eclipse which could then be post processed. I don't know what sample rate would be appropriate however. When I was using Spectrum lab I sampled at 48k, decimated ( 8x I think ) and used a half million samples for each fft.

With vlfrx-tools I would be using a sound card at 48k but would need to decimate / vtresample to something more reasonable for storage.  I don't see that vlfrx-tools has a decimate function unless I am mis-reading the notes and I don't know if using vtresample would be sufficient or appropriate in order to preserve the sub Hz doppler with sufficient granularity.

vlfrx-tools is a fabulous set of tools with many options. Ancora imparo - I am still learning.

 cheers, Graham ve3gtc

Jonathan

unread,
Mar 18, 2024, 10:13:12 AM3/18/24
to hamsci-grape
Hi Graham,

As long as you can set up ntp to sync to another stratum 1 source through another system, that is adequate for the system vlfrx-tools is running on. It is more important to have a PPS source for vttime to use and is the best effort that can be done in terms of the best timing accuracy for data acquisition. It should be the first priority. Without it, it's only marginally better than the timing accuracy of Grape 1 data and would be a fruitless effort without it.

You can sample at any rate 8k or higher. The "decimate" function is resampling in this context. It works the same way as it does in Spectrum Lab, so you can do the same thing. Sub-Hz Doppler observations will only be possible with a GPS PPS fed into the soundcard.

It's worthwhile to set up vlfrx-tools to generate a compressed spectrogram using vtsid as well as the data products you produced before and use vtwrite to save the raw data. It's easier to have vtsid running on live data instead of processed data as it's designed to be better used in that way.

Jonathan
KC3EEY

Dirk Schmidhofer

unread,
Mar 18, 2024, 1:07:21 PM3/18/24
to Graham c, Jonathan, hamsci-grape
Greetings, Graham, Jonathan and others,
I too will be traveling to Texas (several days early) to set up all i can for eclipse data acquisition....
These discussion  threads are truly enticing..... What types of antennas are you using?  (Perhaps with enough description that a new ham could fabricate them? (O:)
Cheers!
Dirk Schmidhofer
KK7JZF


Graham c

unread,
Mar 18, 2024, 7:13:34 PM3/18/24
to hamsci-grape
Good evening Dirk.

I am using what I refer to a as a WOF antenna.  It is little more than a Wire On (wood) Fence. Basically an unterminated short beverage antenna that runs along the back edge of my property and is about 6 feet above the ground supported by the fence and is a bit longer than 1/2 wavelength on 40m. 

Just about anything will work and a dipole will do quite nicely.  You could also look at LOG ( Loop On Ground ) or BOG ( Beverage On Ground ) antennas if you wanted to try something a bit different for receive purposes. These LOG and BOG ( and my WOF ) antennas can be low noise trading  a bit of signal strength over a dipole for a bit better SNR.

cheers, Graham ve3gtc

Reply all
Reply to author
Forward
0 new messages