The line up here is a home made quarter wave ground plane antenna on the roof with a small home brew pre-amp with about 12 - 16db gain inline about four feet donw from the antenna. This connects directly to my RTL2832 / E4000 dongle.
I live about 5 to 6 miles from Newcastle Airport and have pretty much line of site to the runway. Not quite, there is a very small rise between the two locations, but only about six feet. I can follow planes almost down to the tarmac as this screen grab shows.
(https://lh5.googleusercontent.com/-8xm9OcTqJTo/UIqy0KcO7II/AAAAAAAAESY/Uy7xnI4S2bg/s800/adsb%2520landing%2520at%2520NGNT.JPG)
Now - here are the problems:
Firstly, when planes land at NGNT, they come down and then, while on the ground, something weird goes on for about five minutes, probably while they are taxiing and switching off the adsb system. What happens is that the plane disappears off my map for a moment, and then re-apprears about twenty or more miles away, stationary on the map, and then it flips about thirty miles south, and then again, it flips back north. This goes on like I said for a few minutes and then stops. I know it is the plane that landed, because if I had it selected, it stays highlighted.
Secondly, since I fitted the pre-amp, the system may be overloading on local aircraft. I now rarely see planes that have taken off five miles away, even if they fly really near, and more weird than that, even when they are far off, but well within range, they don't re-appear on my system - even over a hundred miles off. meanwhile, planes picked up far off, are on screen right overhead and right down to the ground if they are landing at the local airport.
I can see that the dongle with its low dynamic range might easily be overwhelmed with its pre-amp by string local signals, but why when the plane is far off does it not appear?
I am certain that neither of these weird behaviours happened before the pre-amp went on.
These things are no problem to me, I just want to understand what is happening - the whole project is just to see if I could make a rock bottom cheap system and make it work.
(https://lh3.googleusercontent.com/-72s1znC4038/UIj9tkiYNqI/AAAAAAAAESA/5HGuT2YsLSo/s800/adsb%2520map%2520planes%2520too%2520big2.JPG)
Probably poor message decoding.
Kinetic used to have an issue with on-ground aircraft being displaced but it was resolved some years back. Even the most experienced still have decoding issues:
http://radarspotting.com/forum/index.php/topic,1699.msg7573.html#msg7573
http://radarspotting.com/forum/index.php/topic,1707.msg7631.html#msg7631
http://radarspotting.com/forum/index.php/topic,1708.msg7628.html#msg7628
Yes I supposed it was likely to be a software issue brought on by overloading the dongle. If the analogue/digital converter in the dongle is overwhelmed and it likely is, by virtue of local traffic and a pre-amp, the ability of the eight bit system to manage these signals will be compromised. I can see that. Since the flipping positions happens in both Virtual Radar and Plane Plotter, I conclude that it is gr-modes that is making erroneous decodes with the bad data.
However, what I can not understand, is why I can receive a landing aircraft perfectly until it is on the ground five miles away, but can't see a taking off one anymore, even when it flew over my house at 5000 feet and then travelled a hundred miles away. For some reason, if picked up as a massive signal and not decoded, it won't be decoded. It is almost as if it was blacklisted after the initial swamping signal.
Last week before the pre-amp was put in line, neither problem happened, I just couldn't see planes as reliably at long range and probably had about a fifty miles less 'seeing power' in some directions. The last couple of days I have even been spotting high flying aircraft around Chelmsford which must be at least 275 miles from here and overland with some high ground in the path around the North York Moors. Before the pre-amp, I was lucky to get as far south as Nottingham in a similar direction.
I might remove the pre-amp for a bit and check again without it. It's a pain getting ladders out and going up there though...... so, maybe it can wait.
Could be a doughnut affect from your home brew antenna?
Are the "missing" aircraft actually transmitting location data? Bear in mind that a fair amount of traffic transmits altitude but not position. Today on my dongle I had 1 transmitting position but 3 not. The aircraft symbol in Planeplotter (or "a" from the keyboard brings up a separate window) shows what is being received which may not match what is being plotted.
Ian
Edit: if it is showing on FR24, it should be showing on the dongle!
Hi Ian,
Yes they were on FlightRadar24 but not on the dongle. This only affects aircraft taking off five miles down the road. I see landing planes fine. Quite a lot of traffic is not of the Type 17 message type which contains locator information. So they don't get plotted on the map, obviously. You can see them streaming through the terminal window at speed if you pause the display. Maybe half the messages on the screen don't translate into a plot on the map because they aren't the right type of message. These missing take off planes are transmitting type 17 messages which don't get decoded, but I'm getting plenty of pretty far away ones. Overloading from the pre-amp I think....
Thanks for the tip about 'a' in Plane Plotter. neat trick to conveniently see what exactly the messages contain. At this moment, 11 out of 24 planes in my view are not transmitting Type 17 messages and are not being plotted.
I wrote an email to Nick Forster who produced the gr_air_modes software that is doing all the work of decoding ADS-B from dongles and other such hardware. He thanked me for the big report and says he thinks he knows exactly what is causing it and will sort it for future releases.
I like the way these Linux people provide all this free software for other people and are so responsive to the people who use it. Fantastic altruism.
I got a new version of gr-air-modes that solved this problem and another one that I hadn't mentioned here. The new version of the software can be obtained by following these instructions from Nick Foster, the guy who wrote it.
"If you installed from the git instructions, go into the gr-air-modes source directory and type:
$ git remote update
$ git pull origin master
$ cd build
$ make
$ sudo make install
$ sudo ldconfig
...and that should do it."
A great job, and great communication from the guy that dongle operators owe for the opportunity to set up a £25 ADS-B receiver.
Very impressive to get that sort of response. I have had similar interaction with a guy who was writing an online photo database.
I think Nick Forster and the guy who is working on SDRSharp are pretty popular right now having figured out how to provide the community with software to make use of these dongles.
But once the enthusiasm fades or real life gets in the way, the project can fold. Hopefully not yet!
Ian
Yes - that's true. Fortunately for now, he is beavering away at making it better.
I'm waiting for the R820T dongle to arrive. It will be a while yet, as it is coming from Hon Kong, but when that one comes, I will reinstate my 2 meter hb9cv and see what I can do in the way of VHF listening for starters using SDR Sharp. It works on my netbook, but for some reason, not on my old Dell Inspiron. I wasted half a day trying to get that working on the Dell. I'll try one more version of Net Framework on that and then if it won't work, I'll just use the netbook for it.
There is contradictory info around about the capabilities of that tuner chip. Some say it stops at about 1ghz, but there is other stuff that has it going well up above that. It will be interesting to play around with it anyway.
I wonder if it will need different drivers to make it work.
Dongle
Looked on Google about drivers and this list
http://lists.osmocom.org/pipermail/osmocom-sdr/2012-September/000253.html
indicated those for the R820T are being included in the main library. If someone takes that library and uses it in one of the DLLs for SDRSharp, that would be it.
Indeed change #964 for SDRSharp says "Added R820T tuner type"
The link also mentions the frequency ranges achieved.
Ian
Based on the information here:
http://www.hamradioscience.com/10-ads-b-receiver-rtl2832u-r820t/
I've ordered one of these from eBay, so when it arrives I'll have a dabble. I have no idea about Linux so if it requires that - it'll be a steep learning curve. Maybe before then Mike will have written a guide to help us out ;D
Tim
I've written a guide to run Ubuntu in Virtualbox under Windows and then install the ADS-B receiving software. Lots of pictures!
But it is fairly straightforward thanks to the efforts of the guys who have written the software.
Currently with Mike to test whether it makes sense with his new dongle.
Ian
Ian,
Great news, as soon as it's available I'll study it, I doubt my dongle will arrive before the end of Nov - it's coming from HK, so no rush :)
Tim
My DVB-T dongle has arrived today but I've been out to lunch and three bottles of Merlot later I think it can wait until tomorrow!
Hi All K.Anderson reckons in his book sbs1 and basestation .that you can build a Bias tee & uhf pre amp for around £20 will build one on breadboard an see if you can
Hi again just found this one its the same as K,A. http://www.hamtv.com/pdffiles/biasT.pdf
Also this one But a bit complicated. Dongle pre-amp http://sivantoledotech.wordpress.com/2012/08/03/a-high-end-uhf-preamp-for-the-vhfuhf-dongle/
Shaky
What language are you speaking? I had a quick look at those articles and to those who don't have a clue (such as me) this is all very strangle and unknown!
Ian
Hi Ian .The pre-amps and tees cost about £129 if they can be made for £20 enough said
You could make a fortune selling to those who don't have the skills. Just need to convince us we need them ;)
Hi All . Not so the bias tee yes it works, Pre-Amp no wrong info by K.A Also he printed a dia/of a J pole its wrong. Back to the workbench
Billko
Unlike Lionel K Anderson to make a mistake. ;)