Meteor Scatter - Beginner's Luck

[IanL]

Frustrated by the continuing run of cloudy weather (it only seems to clear when there is a full Moon), I decided to dip my toe in to Meteor Scatter detection.  Thanks to the BAA RA page, the S@N articles and a lot of online research, three or four days in and I think I'm getting somewhere.  I took a trip to B&Q and got all the parts for the DIY three element Yagi antenna and built it in a few hours over two days at a cost of less than 20 quid:

The ''Mast' is just a bit of leftover wood from the observatory build that I wedged in to a spare metal fence post spike. I'm looking to attach it to the house in due course as there is a good open aspect to the SouthEast and should be easy to get it about four metres or so above the ground once I have a longer coax cable. I am using a NooElec NESDR SMArt SDR dongle, about 22 quid from Amazon (https://www.amazon.co.uk/dp/B01HA642SW/ref=psdc_430550031_t4_B00P2UOU72).  It's running off the observatory PC when I am not using it for imaging.

I haven't made a huge effort on the electrical noise front, I have a Nevada 12v/8A supply, the PC and a PowerLine WiFi extender in there about four feet below the dongle.

The biggest challenge was getting the SDR Dongle talking to Spectrum Lab as I was too quick off the mark Googling and following other people's instructions. Turned out that I just needed to install the ExtIO DLL that is available from the manufacturer's web site instead of looking for more arcane solutions.

I got things up and running using the MetScat Starter configuration from the BAA web site.  A bit of refactoring of the conditional actions scripts to give all the variables meaningful names made it easier to figure out what was going on (including a eliminating few redundant variables and conditional actions that don't actually do anything at all - we call it 'Cargo Cult' programming in the trade  ).  I've picked up some links to other useful scripts and configs so I'll be experimenting with the setup once I am happy with the hardware.

Anyway, after running for a few nights and days, I got some really nice captures last night:

Following an earlier post by The Admiral on here, I decided to do a bit of maths (and some diagrams) to figure out what I should be able to detect from my location.  I want to site the antenna at a height of 4m (about 15m above sea level all in), with the highest point from there in my preferred direction being 32m asl at a distance of 3.8km, after which it's sea until Kent. My rough calculations suggest I should be able to get down to under 0.5 degrees above the horizon, and the fact I am getting some reasonable responses from a metre and a half in a similar direction suggests I am in the ballpark.

Using the Yagi antenna, I think it should give me good sensitivity over somewhere between 60 (outer red lines) and 90 degrees (outer blue lines) once far enough above the ground. The optimal direction (centre red line) seems to be roughly towards Lyon / Tunis as this puts the radar illuminated area right in the middle of the antenna's best reception angle.

I calculated and roughly marked out the upper (yellow) zone where meteors at 120km would be illuminated by the Graves radar, and the lower (green) zone at 80km. If I can get reception down to one degree above the horizon, then I can detect in the whole zone subject to sufficient SNR.  (imagine a volume extending up from the green zone to the yellow zone and hinted at by the paler green lines). I also marked out the ISS area (purple) at 408km, where it would be illuminated by the radar and within the 90 degree angle of the antenna. Should certainly be detectable inside the red 60 degree lines.

Just to check, I did the same plot but assuming I could only see down to 5 degrees above the horizon:

As you can see, the ISS is still readily detectable, but the radar-illuminated meteor zone with line of sight to the antenna has shrunk drastically, getting worse the lower the meteor goes. I don't know if line of sight is the absolute cut-off, but I'm assuming it is.

Anyway, any thoughts or suggestions gratefully received!

[SteveNickolls]

Hi,

Thanks for posting your success with the equipment. I totally agree that outages for maintenance apart you can detect meteors this way 24/7 unaffected in the main by weather though one pick up interference from poorly suppressed vehicles and the freak atmospheric disturbance.

In the (what now feels the dim) past The Admiral and myself detected many meteor trails this way and initially expected to be able to show whether a particular meteor was approaching or receding going on the 'hockey stick' trail for the longer detections. We ended up realising the picture is more complicated and indeed that the capabilities of the system may not be sufficient to do this. Mostly I recall we detected decreasing frequency events, very few increasing ones. We did try and get a reply from Dr D Morgan on the issue. His pdf on the subject of the Detection & Analysis of Meteors by RADAR is well worth reading-https://www.britastro.org/radio/projects/Detection_of_meteors_by_RADAR.pdf

As regards detecting the ISS the widest detection I have ever recorded was on 8.11.2015 at around 11.56 AM detecting the ISS first off the SW Eire coast and following it comparing the passage on ISS Tracker until it faded at a point just West of Rome. At the time of the ISS pass it had an altitude of 249-251 miles and velocity around 17,173 mph. It did open up the question hoe GRAVES could provide a detection for something so NW of its location. I have wondered if GRAVES had by coincidence detected another satellite at the same time as the ISS pass over.

My equipment, including the S@N article antenna are still working fine and due for a major overhaul this spring. My antenna is horizontally polarised and you need to keep an eye on the first an third elements slipping out their connections, pigeons too I've found like pipes to land on!

Best Regards,
Steve

 

[SilverAstro]

2 hours ago, IanL said:

Anyway, any thoughts or suggestions gratefully received!

A few random thoughts ! :-

At vhf (143Hz +/-  in this case) the radio horizon is not your optical horizon, ie. 'line-of-sight' means different things in those two situations, so a simple geometric plot will not define what you can hear.

Very approx you can use a 4/3earth model for propagation at vhf (under quiescent condx, - tropo ducting, Es and similar are a whole other kettle of fish !)

Obstructions at a distance can be an aid to extending your horizon by diffraction, they are not always an obstacle.

There used to be freeware path-profile software to plot your path to a distant Tx, (or reflector)  but it is years since I dabbled so not to hand, sorry, but some delving into google may trawl up something.

It is all a very long time ago that I did meteor scatter (voice via !) if anything else bubbles up in my brain I'll be back

Very interesting, good luck 

 

[The Admiral]

2 hours ago, IanL said:

My rough calculations suggest I should be able to get down to under 0.5 degrees above the horizon

Well done for working through the morass to get some results .

If my shaky understanding of antenna theory isn't too far off the mark, things don't quite work out that way. There is a big difference in antenna response plots for free-field conditions compared to when thy are actually sited near the ground. Below is a calculated response in the vertical plane, for a Sky@Night antenna mounted 6m above the ground with its elements vertical.

As you see, the peak response is about 4° above the horizontal (even though the antenna is directed horizontally). These peaks in response are due to the constructive and destructive interference between the incoming direct radiation and that which will have been reflected from the ground some way distant. Canting the antenna downwards or upwards won't significantly change that as it's the effect of the wave interference, rather than the antenna design per se. Of course, this is really for an idealised arrangement, so one might expect things to be somewhat perturbed from this in real life. Perhaps other scatters also take place which allow more distant reflections to be observed. May be?

Either way, the higher the antenna the better. I have mine in my loft space and it works fine. Or it used to, when I was active in this field.

Ian

[IanL]

14 hours ago, SteveNickolls said:

Hi,

Thanks for posting your success with the equipment. I totally agree that outages for maintenance apart you can detect meteors this way 24/7 unaffected in the main by weather though one pick up interference from poorly suppressed vehicles and the freak atmospheric disturbance.

In the (what now feels the dim) past The Admiral and myself detected many meteor trails this way and initially expected to be able to show whether a particular meteor was approaching or receding going on the 'hockey stick' trail for the longer detections. We ended up realising the picture is more complicated and indeed that the capabilities of the system may not be sufficient to do this. Mostly I recall we detected decreasing frequency events, very few increasing ones. We did try and get a reply from Dr D Morgan on the issue. His pdf on the subject of the Detection & Analysis of Meteors by RADAR is well worth reading-https://www.britastro.org/radio/projects/Detection_of_meteors_by_RADAR.pdf

As regards detecting the ISS the widest detection I have ever recorded was on 8.11.2015 at around 11.56 AM detecting the ISS first off the SW Eire coast and following it comparing the passage on ISS Tracker until it faded at a point just West of Rome. At the time of the ISS pass it had an altitude of 249-251 miles and velocity around 17,173 mph. It did open up the question hoe GRAVES could provide a detection for something so NW of its location. I have wondered if GRAVES had by coincidence detected another satellite at the same time as the ISS pass over.

My equipment, including the S@N article antenna are still working fine and due for a major overhaul this spring. My antenna is horizontally polarised and you need to keep an eye on the first an third elements slipping out their connections, pigeons too I've found like pipes to land on!

Best Regards,
Steve

 

Thanks - I had looked at the paper but bears re-reading a few times - I do need to learn a lot more about radio as his setup indicates approaching meteors with lower frequencies and vice-versa, so the opposite of what common sense would indicate. I don't (yet) understand what the stuff about "sideband demodulation" means (anyone who can explain in words of one syllable please chip in - I'm not completely technically illiterate, but this is a domain outside my current experience) or whether it is applicable to my setup.

It is interesting that you got an uneven distribution of events and I'll see if I can replicate it in due course. Regardless of the directionality vs frequency shift confusion, I do appreciate that there are two elements to the Doppler shift - one for meteor / trail relative to the radar transmitter, and one relative to the receiver. Naively I has assumed that the situation would be simpler for receivers North of Graves as generally speaking a meteor approaching the the transmitter would also be approaching the receiver (or at worst moving parallel to the receiver), and vice versa. The situation to the South of Graves should be more complex as the meteor could approach one and recede from the other?  Any further info you can link to or share on this would be appreciated as it would save me re-inventing the wheel.

Your ISS detection appears to suggest that the detectable area might be a lot wider than one assumes from the available data about Graves? Did your antenna have a wider angle as I'd have guessed a directional antenna pointing SE would limit detection to the NW even if there was a radar signal going that way?

13 hours ago, SilverAstro said:

A few random thoughts ! :-

At vhf (143Hz +/-  in this case) the radio horizon is not your optical horizon, ie. 'line-of-sight' means different things in those two situations, so a simple geometric plot will not define what you can hear.

Very approx you can use a 4/3earth model for propagation at vhf (under quiescent condx, - tropo ducting, Es and similar are a whole other kettle of fish !)

Obstructions at a distance can be an aid to extending your horizon by diffraction, they are not always an obstacle.

There used to be freeware path-profile software to plot your path to a distant Tx, (or reflector)  but it is years since I dabbled so not to hand, sorry, but some delving into google may trawl up something.

It is all a very long time ago that I did meteor scatter (voice via !) if anything else bubbles up in my brain I'll be back

Very interesting, good luck 

 

Thanks, a few more terms to Google. I was aware that VHF propogation is somewhat beyond the horizon but went for simplicity in the first instance. It is helpful that you have confirmed other effects come in to play - I thought there must be a lot more to it given people much further North were able to perform detections when they have pretty much no LOS.

13 hours ago, The Admiral said:

Well done for working through the morass to get some results .

If my shaky understanding of antenna theory isn't too far off the mark, things don't quite work out that way. There is a big difference in antenna response plots for free-field conditions compared to when thy are actually sited near the ground. Below is a calculated response in the vertical plane, for a Sky@Night antenna mounted 6m above the ground with its elements vertical.

As you see, the peak response is about 4° above the horizontal (even though the antenna is directed horizontally). These peaks in response are due to the constructive and destructive interference between the incoming direct radiation and that which will have been reflected from the ground some way distant. Canting the antenna downwards or upwards won't significantly change that as it's the effect of the wave interference, rather than the antenna design per se. Of course, this is really for an idealised arrangement, so one might expect things to be somewhat perturbed from this in real life. Perhaps other scatters also take place which allow more distant reflections to be observed. May be?

Either way, the higher the antenna the better. I have mine in my loft space and it works fine. Or it used to, when I was active in this field.

Ian

I've been looking at a few of these diagrams, and it seems that to get an idealised response similar to the freespace model, the antenna must be at significant height above the ground?  See this for example, which slide 2 says should be applicable to any horizontal antenna:

http://www.dcarc.club/2016 N8PR 2 Meter Antenna Plots.pdf

So to get a 1 degree takeoff angle, you'd need it to be about 20 metres above the ground - a bit too ambitious for my liking, though maybe I could talk nicely to the local church - easily accessible tower with a flat roof on top of a hill with open aspects over flat ground to the Southeast

Looking at your model I guess 6 metres would be significantly better in that you get -3dB down at 2.1 degrees (half the max gain?) vs. 3.something degrees at 4 metres?  Might need a longer ladder but probably worth a try.

Thanks for the encouragement everyone.

[The Admiral]

32 minutes ago, IanL said:

I've been looking at a few of these diagrams, and it seems that to get an idealised response similar to the freespace model, the antenna must be at significant height above the ground?  See this for example, which slide 2 says should be applicable to any horizontal antenna:

http://www.dcarc.club/2016 N8PR 2 Meter Antenna Plots.pdf

So to get a 1 degree takeoff angle, you'd need it to be about 20 metres above the ground - a bit too ambitious for my liking, though maybe I could talk nicely to the local church - easily accessible tower with a flat roof on top of a hill with open aspects over flat ground to the Southeast

Looking at your model I guess 6 metres would be significantly better in that you get -3dB down at 2.1 degrees (half the max gain?) vs. 3.something degrees at 4 metres?  Might need a longer ladder but probably worth a try.

Indeed Ian, and in fact I used EZNEC to produce my plot for the S@N antenna. But I wouldn't get too hung up about it, as I know Steve's antenna is quite low and he lives in Nottingham, and still gets reasonable reception. Although mine is in the loft, and therefore 'quite high above the ground', it's response will undoubtedly be affected by nearby walls and roof timbers, to a degree I wouldn't like to guess. Attached to a chimney is probably the best that can be reasonably achieved, but even then unless you are prepared to clamber up ladders and attach it yourself, it won't be a cheap exercise. And of course, you won't be able to fiddle with it when it's in place. I take the view that the greater issue is the likely affect of obstructions that will affect your view to low down close to the horizon. Mine isn't too bad, but at least putting it in the loft allows it to 'see' over the surrounding rooftops.

Just for clarification, the blue segment in my map I calculated assuming that the signals originated from within the Graves beam of between 15° and 40°, and at a height of 100km, using simple trig. The radius R1 assumes simple line of sight at an altitude of 5°, and R2 assumes that we could see down to the horizon (i.e. tangential to the earth's local surface). Again, simple trig.

All very hypothetical of course, but it does allow one to become grounded in one's expectations. I think it's fascinating to think that when you visually observe the ISS (which of course is much higher than where meteors interact), that when it is close to the horizon it is well out over the Med!

Still, you do have meteor scatter signals, so it's a great start.

Ian

[IanL]

Yes I used your diagram as the basis for doing mine. I did a bit of trigonometry and also used scale diagrams (in Visio) to verify some of the angles and measurements. The results from first principles came out pretty much the same as yours so I must have got that bit right.

Height-wise, I can put the antenna on the SE corner of the house at somewhere between 4 and 6 metres fairly easily but I'm not going up to the chimney that's for sure.  I have managed to observe the ISS fairly easily when it is over Gibraltar from my Astro Society's observatory which has a pretty much clear horizon SE through SW.

[IanL]

12 minutes ago, The Admiral said:

Indeed Ian, and in fact I used EZNEC to produce my plot for the S@N antenna.

Do you have the antenna file available and would you be willing to share it?  Would be interesting to play with this and see what I can learn.

[SilverAstro]

1 hour ago, IanL said:

Thanks, a few more terms to Google.

given people much further North were able to perform detections when they have pretty much no LOS.

Thought you might like some hints

yes, I was thinking of Gepeto(?) way up north somewhere who was most surprised to get reception, not seen him around in a long while, cant even find the forum we frequented now !

[The Admiral]

23 minutes ago, IanL said:

The results from first principles came out pretty much the same as yours so I must have got that bit right.

Or I have .

20 minutes ago, IanL said:

Do you have the antenna file available and would you be willing to share it?

I'm not quite sure what you mean by this (it's some while since I played with EZNEC). Do you have the EZNEC program? To be honest I can't find the .EZ file now, but you would need the EZNEC program. If you have it, it's simple enough to put in the coordinates of the Yagi elements.

Ian

[IanL]

1 minute ago, The Admiral said:

I'm not quite sure what you mean by this (it's some while since I played with EZNEC). Do you have the EZNEC program? To be honest I can't find the .EZ file now, but you would need the EZNEC program. If you have it, it's simple enough to put in the coordinates of the Yagi elements.

OK yes I can get the program, I was just being lazy and hoping you'd have the data file for the antenna to get me up and running more quickly. I'll see what I can figure out.

[The Admiral]

6 minutes ago, IanL said:

OK yes I can get the program, I was just being lazy and hoping you'd have the data file for the antenna to get me up and running more quickly. I'll see what I can figure out.

There is a demo program but there is a limitation on the number of data points IIRC, otherwise it's a paid for program. I'm not sure you'd glean much more than the information you already have to be honest.

Ian

Edit. Found them!

3-el yagi S@N vert.ez

3-el yagi S@N.ez

[IanL]

OK, the demo version won't run those files due to the number of points limitation. The completely free NEC (http://www.qsl.net/4nec2/) will though, and have managed to replicate your plots already. Time to experiment with different heights to save me going up and down the ladder....

[The Admiral]

It'll be an interesting intellectual exercise, but in the real world I wouldn't worry too much. If it was me, I'd just put it as high as I could whilst maintaining access. Better to experiment and see what signals you actually receive than try for a theoretically better placement, which it may not be.

Ian

[IanL]

40 minutes ago, The Admiral said:

It'll be an interesting intellectual exercise, but in the real world I wouldn't worry too much. If it was me, I'd just put it as high as I could whilst maintaining access. Better to experiment and see what signals you actually receive than try for a theoretically better placement, which it may not be.

Ian

Practically that's all I can do, but given it is nearly dark and we have snow showers, I'm quite happy to learn all I can about the theory right now!

[The Admiral]

3 minutes ago, IanL said:

Practically that's all I can do, but given it is nearly dark and we have snow showers, I'm quite happy to learn all I can about the theory right now!

Makes sense! It is interesting to see how things vary.

Ian

[IanL]

Okay, first set of results are in. I used your original .ez files in 4Nec2. I optimised them a bit to reduce the number of segments per wire to the level recommended by the validation menu. I also changed the ground to be more representative of this part of the world using the World Atlas of Ground Conductivities (http://hamwaves.com/ground/doc/ground-conductivity-atlas.vlf&mf.pdf), so conductivity of 0.007mS/m (from the "7" on the relevant part of the UK map), plus the Dielectric Constant of about 10 which seemed reasonable (highly variable depending on moisture content and soil type).

First I tested the three element Yagi in horizontal vs. vertical orientation:

Horizontal seems to be the way to go as the maximum gain is 12.3dB vs 9.62 for vertical (this was the case across a range of heights above ground). So that's almost twice as good in theory. The downside is a slightly narrower -3dB angle in the horizontal plane, you'll get about 78 degrees for vertical vs. 56 for horizontal. Based on the mapping models previous, an approximately 60 degree spread seems fine for our purpose.

Sticking with horizontal orientation, the next thing was to check heights. As expected, the higher the better - a less lobed pattern and more gain at a lower elevation angle the higher you go:

There is a 1dB gain improvement between 1.5 metres above ground (where I am now) and 6 metres. Practically I think 4 metres is a safer working height for my purposes and the difference between 4 and 6 is not huge. More importantly, the angle of maximum gain comes down significantly by getting a few metres above ground.

Lastly I looked at elevation angles. Again as expected it doesn't make a lot of difference at 4 metres (though it is more pronounced at 1.5 metres):

There is a tiny increase in gain if you point it horizontally vs up or down by 10 degrees.  I've started looking at some other plans for Yagis with more elements and different materials. The direction of travel seems to be that the more elements, the more gain at the expense of the antenna becoming more directional, so it's a bit of a trade off. I also noticed that there is a big difference in quality of cable. Over perhaps 15 metres for my setup, as much as 3dB by spending a few more quid on better 50 ohm coax. (roughly 3dB = 2x better).

[The Admiral]

For what it's worth, I ended up using an LFA-Q Yagi from InnovAntennas (http://www.innovantennas.com/antennas-a-accesories/view/productdetails/virtuemart_product_id/371/virtuemart_category_id/1.html). Better I think than my home-made simple Yagi, and IIRC it has better rejection of noise and front-to-back ratio. It's actually squarer in outline than suggested here though, and does take up space. Had to assemble it in the loft in order to get it through the loft hatch .

Ian

[kev]

Hi. Ive been meaning to look into this for ages, and finally im not away from home.  

So as i understand this thread, im ideally sited almost.  We are nearly on top of a hill on the ese slope.  Ive a flat roof c4m high on the kitchin facing ese and after a small path theres the hedge and another 2m drop to the next houses level.  (?)

Cheers for any feedback

it would be niceto do some astronomy in not so sunny scarborough   

 

 

[The Admiral]

20 minutes ago, kev said:

So as i understand this thread, im ideally sited almost.

Potentially I would guess. You are quite far North so I would imagine the clearer the view to the horizon the better. What about the rooftops in the desired direction, and is it easy to get a line of sight above them? Looking at the RMOB site, there's a chap on the West coast of Scotland, near Glasgow, so I guess it'd be worth a try. His detection rate does look lower than others more favourably placed, but it's hard to tell as there isn't a standard set-up.

Anyone on here who does meteor scatter and who lives well North have anything to add?

Ian

[jarraman]

hi there.

I'm in Jarrow in the middle of a housing estate,10 miles east of Newcastle.

my equipment is a bought 3 element yagi attached to my chimney,say about 6 meters above ground,facing south.

using spectrum lab,with the fun cube dongle.

during meteors showers can usually peak about 80 an hour,one of my hits attached.

ken.

[IanL]

Well the first results are in. Chart showing total meteors detected by hour over the first 14 days of February. (0 is midnight at the left, 12 midday and 23 is 11PM at the right). You can clearly see the diurnal (day/night) cycle as more sporadic meteors are swept up like bugs on a car windscreen as we face the direction of Earth's travel on its orbit during the early hours of the morning. I am aware that the Sun does affect the propagation of some radio signals due to changes in the ionosphere which may also have a bearing, but I think this is a good demonstration nonetheless.

[IanL]

OK so I'm working to improve the Spectrum Lab setup and make it a bit more user friendly. One thing I'd like to do is use some of the programmable buttons to switch the frequency and display settings.  If I right-click a spare command button, I can enter the following in the dialog:

sdr.freq=143047900

Clicking the button should switch the VFO to the appropriate frequency for Graves to suit my current waterfall setup.  I can then have a different button to set the SDR VFO (frequency) to (say) 144428000 for the GB3VHF beacon to test the SDR and antenna are receiving OK.

What actually happens is that the VFO entry field at the top right in the "Freq" Tab does get updated to the desired frequency, and the waterfall/spectrum display also updates the displayed frequency scale and the radio station markers (having created a radio station file with those frequencies in - see below).  So far so good.

The problem is that the SDR does not actually re-tune until I click in the VFO entry field and hit the "Enter" key. (So for example switching from Graves to GB3VHF all appears good, but the beacon signal does not show up in the spectrum or waterfall until I hit enter in the VFO field).

Anyone have any ideas as to what else I could try to fix this?  I know that the documentation says only the Perseus and SDR.IQ devices were supported at the time of writing, but frustratingly this appears to be a case of "close but no cigar", like so much of Spectrum Lab's rather eccentric user interface and internals.

The other thing I am working on is changing the conditional actions script.  In a later presentation by Paul Hyde, he highlights some problems with the original 'cyclic' conditional actions approach to logging meteor data. Basically he found that when the conditional actions measuring the FFT data were triggered (e.g. by the Signal exceeding Noise by the specified threshold), reading the data would return the current FFT data rather than the data that triggered the action to run, leading to spurious results as the values change rapidly between triggering and being read.

The answer is to not use a cyclic script but to use the "new_spectrum" condition to trigger the CA table each time a new FFT is completed.  He gives a rough outline in his slides but I haven't yet found a complete and working example script - happy to write and share one as time allows.

Lastly, attached is a basic Radio Station file for three frequencies - the Graves radar, the GB3VHF beacon and BBC Essex (substitute a nearby FM station for your own setup by editing the file): Meteor_freqlist.txt

- The file itself should be saved in the Spectrum Lab "Spectrum\frequencies" folder.

- Then go in to "Options > System Settings > Filenames, Directories and Stream I/O".

- In the radio station list box, add ", frequencies\Meteor_freqlist.txt" (with the comma but without the quotes).

- Hit "Apply".

- Restart Spectrum Lab.

- In "Options > Spectrum display settings" you need to pick "both / Plot right" in the "Show:" dropdown to add the spectrum analyser to the waterfall view.

- Right click on the white frequency scale that should now be at the right or top of the display, and choose "Frequency Scale Options > Show Radio Station List (frequencies)".

- The appropriate blue marker should now appear at the correct place on the scale.