2017 Eclipse HF Frequency Measurement Experiment
The objective of this exercise was to see if I could capture some data of changes in the ionospheric electron density caused by the motion of the shadow of an eclipse from Doppler shifts on HF ray paths propagating through the eclipse region, I do not have a reference signal which is locked to an accurate reference such as a GPS signal. So these measurements would include all the know reason or frequency drift in my transceiver, sound card etc. So the exercise was to see if I could take some mean full measurements with minimal test equipment.
This data was not to break any new ground but to simply try and collect data myself which people before have discovered and documented long ago.
The url below describes this experiment in full.
Collecting The Data On The Night Of The Eclipse
I decided to monitor the 5 MHz WWV signal from Fort Collins on the evening of the solar eclipse as this was the only WWV signal I could copy at the time.
The two graphs below show the data I collected, the first shows the frequency of the WWV carrier with an approx 1 KHz offset over the 10 hour period. This was to show in change in frequency due to doppler and NOT to measure the carrier frequency. The second shows the level of the signal during this period. The actual eclipse commenced on one side of the USA at about 16:00 GMT and ended on the other side of the USA at approx 20:15 GMT.
I had NOT done a dummy run the day before so I could normalise the data I was seeing however these are the things that immediately jumped out of the data I captured on the night.
There was a personal eureka moment for me, like taking a series of images of the sun then playing them back to discover the sun is rotating, yes we have known that for a long time but itís different when you confirm this for yourself. The eureka moment was seeing the doppler shift as a result of the sun rising or the grey line transversing my location, I was not expecting to see this.
I was amazed how relatively stable the TS2000 was over the 10 hours. Keep in mind we are talking about parts of Hz without using any external frequency locking such as GPS.
This was the real stand out, I could clearly see the sun rising in my data, that is the grey line was causing a Doppler effect on the frequency.
So my mind started to ask some questions:
I had NO idea what the data should look during a normal period, so I could normalize the data so I could see if there was any effect from the eclipse across the USA,
I would have expected the doppler frequency change as the grey line approached to cease as the sun rose and possibly reverse as the sun travelled to the west in the sky
This meant I had to collect more data the next day, so I had a chance of understanding the data I had.
Collecting The Data With No Eclipse
The following two graphs is the data I got last night, without and eclipse occurring during the capture window.
∑ The frequency stability during the evening is more stable, without the minor drift trends.
∑ The shift in frequency due to the grey line is what I was now expecting now that I extend the capture window past sunrise. The signal was getting quite weak after sunrise as you can see from the second graph.
Will do some more thinking, but I guess I could post the data on the nominated web site to make up the global picture. Any comments before I do this would be most welcome.
The other observation I made that occurred as the grey line passed my QTH (Sunrise), was the width of the tone in Spectrum Labs display widen, almost to the point of seeing two tones. I think this was a result of receiving the carrier without any doppler effect then there was a signal with and without doppler making trace broader. Same bad examples are shown below.
|Carrier Without Doppler||Carrier With Doppler During Sunrise|
Richard, VK6HRC, lent me his GPS disciplined 10 MHz oscillator. Using this as my reference, I setup the TS-2000 to listen to the 10 MHz GPS disciplined oscillator in CW mode with a CW side tone of 1 KHz. As you can see from the graphs below the TS-2000 is approx 5 Hz off in frequency and it drifted from approx 1005.3 Hz to 1004.9 Hz in the CW side tone when tune to the 10 MHz carrier. That is it drifted 0.4 of a Hz over the 5 hours in the data below. I can hear you asking, why is there a 0.3 Hz negative pulse a relatively regular intervals. The answer is simple, as the TS-2000 fan came on due to temperature rise within it's case, the cool air passed over the TXCO and shifted the TS-2000 reference oscillator.
Spectrum Lab - Spectrum Display - TS-2000
Drift of TS-2000 Over 5 Hours
Having data on my TS-2000 I now decided to test the IC-7300.
As you can see from the graphs the IC-7300 was off freq by approx 0.45 of a Hz and drifted about 0.12 of a Hz over the five hour period. No fan operation in the IC-7300. The 0.45 Hz freq correction can be done in software if desired by a simple menu configuration in few seconds.
|Spectrum Lab - Spectrum Display - IC-7300||Drift of IC-7300 Over 5 Hours|
My question to you is, can you see the effects on the frequency resulting from the eclipse ?
Chris James VK6JI
Back to my Astronomy Home Page