Novice attempts at meteor detection

[The Admiral]

Very much a novice with radio detection of meteors, but managed to get set up in time for the Geminid shower. Lashed out on a Funcube Dongle Pro + (but still a good bit cheaper than a dedicated HF receiver), and installed SDR Console v2 onto my PC. I initially tried SDR# but wasn't happy with the way it was installing, but Console seems to be a good alternative, and has I gather additional features over the SDR#. As a youngster I dabbled in short wave radio so I didn't find it particularly daunting. One nice feature is that it has an audio frequency analyser with waterfall display. It's not as sophisticated or as flexible as Spectrum Lab, but at least it's a start. I built a 4-element Yagi out of copper tubing and plastic plumbing bits and installed it in my loft space. It's supposedly centred in the 2m ham band (@ 145MHz), but it should have a reasonable enough bandwidth to receive the Graves transmissions OK. It partially shoots through the gable end wall, but I'm happy with its performance.

During  a couple of days at the weekend of the Geminid peak I was getting pings every few minutes during the evening, some of them lasting for several seconds. I've included a few screen grabs. Overall I've been quite pleased with what the set-up has given me.

A few questions have come to mind though, and I'll be interested in your comments.

1. I don't think that I have observed many, if any, events where the frequency has started low and gone up. I haven't got to grips with the directions of approach of the Geminids, but would this be expected?
2. I've not seen any mention of the impact on monitoring of the way the Graves transmitter scans the sky. Given that it jumps segments every 3.2 seconds, presumably one would expect to see some evidence of this where the echo lasts longer than this?

Here are a few (well, rather a lot actually!) of screen grabs that I managed to get over some short listening spells during the Geminid peak. The LH window is the signal trace in a waterfall display moving downwards, and RH window is the audio frequency on a waterfall display moving from right to left, which, depending on my screen layout, takes around 10 seconds. File names are date (d-m-y) followed by approx. time (hhmm). Nothing's automatic here!

Meanwhile, life's been taken over by Christmas preparations!

I find this one interesting because the frequency starts high but drops below the frequency of the static ionised column signal.

In the next two grabs the signal trace shows the frequency to be above and below the main signal, though it is not clear from the audio trace.

This one also shows two hits.

Ian

[SteveNickolls]

Hi Ian, welcome to the fascinating world of meteors. I'm pleased you picked up detections during the Geminids shower recently while having your aerial in the loft, it will be interesting to see how your day to day counts go. When I did a check on the duration of trails I was getting back in October over 91% of all detections were from trails lasting less than a second so it is difficult for me to see much frequency changes in such a short time however many of the longer trails do show frequency changes over time. My understanding is that where the frequency is rising that corresponds to a meteor headed some direction towards the receiver. The first screenshot below shows a trail with a slight decreasing frequency and the second an increasing frequency.

As for the GRAVES signal changing segments every 3.2 seconds I haven't noticed anything on the longer trails. Part of the difficulty/fun with the longer trails is deciphering them and I have started to classify trails into different categories, for example 'L' shapes,  'cross' shapes, 'broader' trails and trails that fade and reappear. I hope that by building up a reference someday I can work out the possible causes for them.

I'm using the FunCUBE dongle pro+ on a home built aerial of the S@N type and use Spectrum Labs software which takes automatic screenshots (and audio files if you want them). You can tailor the 'conditional actions' script in the software to, for example only record meteor trails longer than a specific period which saves on disk space and vary the signal to noise ratio employed to determine what is/is not a meteor.

Best of luck with your system Ian.

Cheers,

Steve

[The Admiral]

Thanks for your response Steve. At present I'm nowhere near as methodical as yourself; in fact, I'm still at the dead chuffed stage finding that the thing works! I think the next step will be to use Spectrum Lab to avoid sitting there and pressing "Screen Print" whenever anything interesting crops up, or using many GB of disk space!

I'm getting a little confused though. What I understood to be the case (which may be wrong of course),  was that a meteor strikes the atmosphere at speed, is decelerated until obliterated, creating a ball of ionised air around it so doing, and leaving behind it a trail of ionised air that moves (relatively slowly) at the whim of the atmosphere. I tend to think of it a bit like an aircraft leaving behind a vapour trail. The ionisation in the air slowly recombines to form neutral molecules. What is detected are the radio waves scattered from the air whilst it is still ionised.

So, in the following trace (which I've rotated to match yours), I'd assume that A-B is the deceleration phase, from which one can estimate the initial velocity, and B-C is the reflection from the essentially static ion column, before re-combination. The signal A-B is therefore the equivalent of a radio reflection off of the moving aircarft, and B-C is a relection from the vapour trail, so to speak. The frequency change from point A to point B is the doppler shift due to the meteor's motion, and whether it goes from high to low, or vice versa, depends on whether the meteor is tracking away from or towards the detector. My comment was that I don't think I've seen any, within the relatively small sample I have, where A to B increases in frequency (that is, the meteor tracking towards us), they've all been going away. I was wondering if this impression was meaningful, and indeed if there is any explanation for it, like the earth's motion through the swarm. The examples you gave seem to refer to the frequency shift for the B-C phase, but maybe I'm just interpretting your spectrum incorrectly.

Ian

[SteveNickolls]

Hi Ian,

I'll work through my screenshots and try and post some up for you to see. I'm sure there are a variety to look through.

Cheers,

Steve

[The Admiral]

Thanks Steve. Actually I think I got it a little wrong, what I've seen are meteors coming towards me, and nothing going away. That's the opposite of what I said above. That is, if the frequency starts high and falls, then the meteor should be coming towards the detector.

Cheers

Ian

[SteveNickolls]

Here are two screenshots showing what I think you were describing Ian.

Cheers,

Steve

[The Admiral]

Ah yes, I see that you have an example of both. Does your categorisation system allow you to estimate the proportions of each type?

I will have to install Spectrum Lab at some point. Is the installation straightforward (I have a Win7 machine)? Having looked at other posts on this site I'm guessing I'll need a virtual cable to connect the audio output of SDR Console into SL?

Ian

[Gasman]

Hi Ian

If you just run the funcube controller and set it (for Graves) to 143048kHz and then run Spectrum lab you don`t need the virtual cable or SDR-console

Steve

[The Admiral]

Hi Steve

I don't have the Funcube plus controller, nothing was ever shipped with the device. Are you referring to the frequency control programme which is downloadable from the Funcube site? I used SDR-Console as it supports the Funcube without any additional file installations, and was seemless in installation (in comparison to SDR#). Presumably the control programme will still allow the use of SDR-Console (which I find good for other radio applications).

Ian

[Gasman]

Thats the one Ian just done a screen grab running the controller and Spec lab -

cheers

Steve

[The Admiral]

Thanks Steve, I'll give it a try, though Xmas preps are holding things up at the mo

[SteveNickolls]

Ah yes, I see that you have an example of both. Does your categorisation system allow you to estimate the proportions of each type?

I will have to install Spectrum Lab at some point. Is the installation straightforward (I have a Win7 machine)? Having looked at other posts on this site I'm guessing I'll need a virtual cable to connect the audio output of SDR Console into SL?

Ian

Hi Ian,

I do keep screenshots of all the larger trails (about 3 months of data now) and I could work out the proportions of those showing an increasing, decreasing or no change in frequency-hey I got a project to do over Christmas! Will post the results. I have SL running on a Win 7 64 bit machine fine. SL installed ok but I found that I had to 'scan for hardware changes' before it was happy with the FunCUBE dongle.

Best regards,

SteveN

[The Admiral]

Thats the one Ian just done a screen grab running the controller and Spec lab -

fcdcontroller.png

cheers

Steve

Steve, I've now jumped in and installed Spectrum Lab (and at this stage it seems quite impenetrable ), but I have at least got it seeing the Funcube. I've lost the sound output from my speakers though. I realise that the signal paths in a PC can be quite tortuous and may need a bit of untangling, but are you still able to monitor the audio output?

SteveN, what about with your set up?

Ian

[The Admiral]

Ah, I've found a britastro site that has a pdf by a Paul Hyde (aboutspectrumlab v1) which has got me to the point of at least hearing as well as seeing. Must say though, not really sure what I'm doing! Picking up a few nice echos, but I think there's quite a learning curve with setting up this bit of software.