Networked dongle using Linux

[IanH]

Just to show there's more than one way to skin a squirrel cat.

Dongle attached to Raspberry Pi pumping data over a network to a Linux PC running modes_rx and displaying output on PlanePlotter. Limited contacts from downstairs study window.





More details:

Having established that my USB/network hub setup doesn't seem to work, I went back to an earlier idea - using a Raspberry Pi. This £30 computer doesn't have the processing power to run an ADSB decoding programme but people have been using it to stream packets from the dongle to another system, eg to FR24.

The RTL-SDR suite provides "RTL_TCP" to send data from a dongle across the network. A recent addition to "modes_rx" (the decoding programme) allows an IP address to be included from where it gets its data.

After some trial and error, and after solving some strange error codes, I've got it working!

Details to follow in case anyone else wants to try.

And Raspberry Pi and dongle off to the loft soon.

Ian

[IanH]

Dongle in loft using supplied aerial. Networked to my study via Raspberry Pi. Sitting on same metal base plate as an AirNav RadarBox 1090 aerial.

Virtual radar output from dongle at:
http://virtualradar.servehttp.com/VirtualRadar/GoogleMap.htm

[Edit: this link may be down at times - the temporary combination of SSH connection to the the Pi and the use of a Ubuntu VM means that they will not be available overnight until I put them all on one box.]

SBSPlotter run for about 2 hours this afternoon. Comparison below simply because aerials in similar positions, NOT for any ulterior motive:

E4000 Dongle


AirNav Radarbox


Radarbox has greater range but shapes of plots are similar. Message rate was similar for the two receivers.

To give an idea of the area covered, this is captured from the dongle output on PP.



Overall I am quite impressed. The dongle easily covers the visual range for spotting. There are some spurious hex codes in the output as can be seen occasionally on the Virtual radar output from the dongle.

I know Mike was going to try some comparative tests using the same antenna. Those will be interesting.

[Anmer]

Thanks Ian, well done.

I will be in touch this week for some tips.

[IanH]

Replaced the E4000 dongle with the R820T version that arrived today. Definitely more sensitive. Can't quite match the Radarbox for range and area covered but then it it is using the generic aerial rather than a 1090-specific version.

I'll post the SBSPlotter output tomorrow.

As before the Raspberry Pi is pumping out RTL_TCP data but now to a virtual Linux box. The virtual Ubuntu is running SSH to control the Pi, running modes_rx to decode the RTL_TCP stream from the dongle and also running Virtual Radar Server.

Cost: Raspberry Pi~£30, Dongle~£12, Pi power supply~£7 and powered USB hub (since Pi PSU is not enough to power dongle)~£7, ethernet cable~?depends

Anything that you already have reduces the cost. Possibly an old laptop could replace the Pi - need to check whether a Windows version of RTL_TCP would work.

Ian

[junglejet]

On windows you could use RTL1090 via TCP connection, couldn't you?

[IanH]

Andy

Agreed - I realised that after I wrote about RTL_TCP on Windows.

But I don't have any low power PCs that can run Windows for a permanent solution so I am stuck with Linux.

I'm going to try an old laptop in the loft  running WinXP with a dongle to test RTL1090. But the 6Gb hard disk is noisy and I wouldn't trust it to keep going for much longer. And the noise will keep me awake! But it will be ok for testing during the day.

What I want is a low power PC to send data from the dongle to a Windows decoder programme - RTL1090 looks very promising but I don't have a reliable low power data source.

My SBS-1er runs over ethernet, my Radarbox runs using a USB/ethernet hub.

BUT the dongle does not work using the USB/ethernet hub (and from a recent comment on the SDRSharp website it will not run via a "USB extender cable").

So those of us who don't want to run coax need a method to get data from the dongle to a Windows PC.

Ian

[dongle]

Very interesting Ian.

Given that you are not using a resonant antenna, you would not really expect the sensitivity of the system to match another which was using an antenna cut to operate at the frequency of interest. The difference in the antennas would deliver a very different outcome. On the face of it, the fact that with the newer dongle you are getting only a little more sensitivity from the Air nav box with its better antenna, the two systems may not be any different. I don't suppose you can swap the antennas around. i expect they will have incompatible terminations.

If you could cobble together some adaptors with short cable and plugs / sockets, you could swap the antenna's over. It would produce interesting data for the comparison.

Thanks for sharing your findings. Very interesting.


EDIT:

By the way, I have been running my E4000 dongle for about six weeks on a 10 meter USB cable. I think it is an active one. It works fine on the eeepc in my radio cupboard with the dongle up in the loft near the pre-amp and antenna.


E4000 dongle output below:

[IanH]

Thanks for the info re the active USB cable. That gives me the inspiration to try one since needing a Windows PC in the loft is a problem. But a Windows front end is going to be a lot easier for many people.

My old Pentium 3 laptop doesn't want to run RTL1090. It sees the dongle, starts and then locks up. Suspect inadequate CPU power. Committing a newer Core2 laptop is not going to happen when the Raspberry Pi is doing a good job.

Edit: Finally got some plots for the R820 vs Radarbox.

ANRB plots are similar on each occasion but I think the improvement for the R820 over the E4000 is evident.

R820 dongle


Radarbox


Ian