Raspberry Pi type B + DVB-T Dongle to feed FR24

[Mutt08]

http://www.ebay.co.uk/itm/161012784882

Then you need to send about 12v-15v up the wire but not into the dongle power inserter http://www.ebay.co.uk/itm/121321050146 - the antenna is open circuit, so we don't need to worry about blocking the power going on to it.

Then just needs a 12-15v power supply - do measure the voltage it gives before connecting it, if it's over 16v you might damage the amplifier (if it's too low you might not get the results you hope for)

I think we do really well on the dongle - it's not uncommon to see 60-80 aircraft being processed at once.

That looks interesting, have got a spare psu that I just need to check the output on.

Regarding the power inserter shown in your link, am I correct in thinking that the psu will be fed to the 'IN' socket & then the 'HF+DC' feeds to the amp & antenna with the 'HF' then going back to the dongle?

[peterhr]

That looks interesting, have got a spare psu that I just need to check the output on.

Regarding the power inserter shown in your link, am I correct in thinking that the psu will be fed to the 'IN' socket & then the 'HF+DC' feeds to the amp & antenna with the 'HF' then going back to the dongle?

Yep, but do look at other similar items - you might find something with a more standard connector for the power supply to connect to.

[sgtbash]

I don't have a USB hub either, but you might want to get a short USB extender if to provide a mechanical soft or flexible link between the dongle and the Pi if you have a stuff antenna wire connected directly to the dongle.

You don't need to expand the image - I'm running happily on 1Gb

do go to http://forum.flightradar24.com/threa...-Flightradar24 and download the latest RPi feeder software

(on the pi do
wget https://dl.dropboxusercontent.com/u/...rm-rpi_242.tgz

tar -xzf fr24feed_arm-rpi_242.tgz

the update fr24.sh replaceing the feeder with fr24feed_arm-rpi_242

Placing the antenna outside above the roof line will make a considerable difference to the range

I got a mast http://www.ebay.co.uk/itm/321146611580 (but from a different supplier at lower price) on some wall fixed T&K brackets to mount the antenna (I didn't fancy climbing up the roof) - presently I'm using 3 mast sections to get above roof height.

(I can't view your attached image, sorry)

Hi Peter, Thanks for your reply - I've updated my FR24 feed files successfully and now running on those.

I've got to the bottom of my random bombing out also - Heat...

I've removed the cover of my USB dongle and its been running for 13:47 now without a hitch... about 13 hours and 45 minutes longer than usual...
I used a IR thermometer to get a rough temp from the chips on the little board and they are all running at about 50'... a lot more than I'd like really - I reakon they've (or one) possible has a 60' cutoff, one of them is a voltage regulator so maybe... Anyway - taking the cover off seems to have solved it for now.

Heres a snapshot from a good patch! =)


Thanks again!

[digiblur]

Try without the hub at all... I have no problem running my RTL dongle directly off of the Pi USB port.

As to antennna... the 1/2 wave dipole that you made from coax is a pretty darn good one...
I have been trying to build various co-liniears... and they never seem to do as good as the dipole.
I have the same problem with my house and nighbors houses blocking my line of site both east and west.. and some day when I finally settle on an antenna I will get it above the roof lines. Still.. tracking planes 40 to 50 miles away with a minimal antenna basically on the ground is fun.

Picked up a Pi to play around with myself. Toying around with the various builds of antennas can be time consuming but very rewarding. I even had very good results with a 50-1000mhz cable TV 15db amplifier. I had to back up the dongle gain to about 42 but I did have better results with it than without the amp.

I have some other build ideas I'm going to try this evening with some solid copper wire and scrap tubing.

[peterhr]

[snip...]I've got to the bottom of my random bombing out also - Heat...

I've removed the cover of my USB dongle and its been running for 13:47 now without a hitch... about 13 hours and 45 minutes longer than usual...
I used a IR thermometer to get a rough temp from the chips on the little board and they are all running at about 50'... a lot more than I'd like really - I reakon they've (or one) possible has a 60' cutoff, one of them is a voltage regulator so maybe... Anyway - taking the cover off seems to have solved it for now.[...snip]

Actually, the heat problem is not uncommon for people who do install in attics and does need thinking about if putting in a box on the antenna (solar as well as self heating).

[peterhr]

I've added an LNB (950mhz-2050Mhz) amplifier to the base of the Bulgarian collinear, it appears to have increased range (seeing more planes over central Ireland, English channel, north seat and into Scotland) and number aircraft being picked up from about 140 to 160)

[Mutt08]

I've added an LNB (950mhz-2050Mhz) amplifier to the base of the Bulgarian collinear, it appears to have increased range (seeing more planes over central Ireland, English channel, north seat and into Scotland) and number aircraft being picked up from about 140 to 160)

I think I will definitely give the amplifier route a try, would like to see if it improves my reception of ground traffic at Manchester.

[peterhr]

Effect of adding an LNB Preamp http://www.ebay.co.uk/itm/350657754545
to the base of my Bulgarian collinear (before the feed cable)

The pre-amp is wrapped in self amalgamating tape - hopefully water wont get in - feeding 15v up the wire.

Snap2.gifCapture.PNG

You'll notice the polar plot is rounder (corresponding to about 240nm SW - N), the contacts at 200nm is higher We reached 1417 (prev 1214) aircraft and a maximum range of 247nm (457km)

Dump1090 now seems to struggle sometimes with the map display updates - try the link "T-EGBB19" below

[Mutt08]

Those results certainly seem to prove that the pre-amps doing the job, I've ordered an amp & power inserter so hopefully in the next week or so I will be able to post some positive results also. Did you have to make any changes to the gain settings in dump1090?

Just looking at your Dump1090 output & its updating fine at the moment, your getting really good coverage well off the coast of South Devon which seems a total black spot for me beyond Bristol. Bit of a strange one really because thats my clearest line of sight from the antenna yet the direction thats partially blocked by high trees seems to give me the best distances, be interesting to see if this improves any.

[peterhr]

Those results certainly seem to prove that the pre-amps doing the job, I've ordered an amp & power inserter so hopefully in the next week or so I will be able to post some positive results also. Did you have to make any changes to the gain settings in dump1090?

Just looking at your Dump1090 output & its updating fine at the moment, your getting really good coverage well off the coast of South Devon which seems a total black spot for me beyond Bristol. Bit of a strange one really because thats my clearest line of sight from the antenna yet the direction thats partially blocked by high trees seems to give me the best distances, be interesting to see if this improves any.

No changes on the Pi apart from swapping a few plugs around to suit the connection requirements of the power inserter.

Go to http://www.heywhatsthat.com/ :
- put in to do a plot from your location (postcode works well in the UK - it's represents one small area), and a true guess at the height of your antenna.
- once the map appears click up in the air, then enter 40000 and 30000 as your heights

heywhatisthat will draw two rings which represent the distance where aircraft fall below the horizon from your antenna when flying at those heights, this gives a reasonable idea what the maximum achievable range for your rig should be if there are no trees or buildings blocking the signal. My rig is more or less plotting the outer line now, not much more I can do to increase it.

[updated FR24 stats image on prev post to reflect a full day]

[Filldo]

finally got this to work on rpi. i followed the directions from the earlier thread step by step. i hit a few snags along the way though.

first on the autoscript there is "--" that is loose after ./dump1090 that belongs to --enable-agc (i copied and pasted).

the feeder also doesn't seem to like being in the same script as the ./dump1090. i think there isn't enough time to let dump1090 start before the ./fr24feed stuff is executed in the same script, so my feed keeps showing as offline or intermittent. i ended up making two separate scripts one for ./dump1090 and one for ./fr24feed_arm-le****. i also added a few seconds to the crontab --- @reboot sleep 10;/root/dump1090.sh
@reboot sleep 45;/root/fr24feed.sh to give it time to fire up. and for the refire to start on the first minute instead of the zero minute (1 */4 * * * /root/fr24feed.sh).

it looks to be running harmoniously now. i'm not versed in linux but it was kind of fun to play with

[peterhr]

finally got this to work on rpi. i followed the directions from the earlier thread step by step. i hit a few snags along the way though.

first on the autoscript there is "--" that is loose after ./dump1090 that belongs to --enable-agc (i copied and pasted).

the feeder also doesn't seem to like being in the same script as the ./dump1090. i think there isn't enough time to let dump1090 start before the ./fr24feed stuff is executed in the same script, so my feed keeps showing as offline or intermittent. i ended up making two separate scripts one for ./dump1090 and one for ./fr24feed_arm-le****. i also added a few seconds to the crontab --- @reboot sleep 10;/root/dump1090.sh
@reboot sleep 45;/root/fr24feed.sh to give it time to fire up. and for the refire to start on the first minute instead of the zero minute (1 */4 * * * /root/fr24feed.sh).

it looks to be running harmoniously now. i'm not versed in linux but it was kind of fun to play with

The document is done in MS word, it has a tendency to help fully change a double dash to a long emdash ... happens, sorry

Try a sleep command between dump1090 and the feeder. Command is
sleep xxx where xxx is the number of seconds

Don't forget to update the feeder to the latest rpi feeder, see post 8 in tne thread again.

I'm pleased you feel you've got into Linux from this project

[Filldo]

i tried sleep and got almost the same results. also used the 242 version from scratch.

in terms of linux....although it might be fun at times, it's not for those who seek instant gratification =p

[peterhr]

The cron job I use has two lines
@reboot sleep xx; /root/fr24.sh (xx = number of seconds to pause)
* 2 * * * /root/fr24.sh

The reason the reboot line has the sleep is to give the Pi networking time to get going. My pi is connected to the router using homeplug / powerline adaptors, so there is little pause - if I was using WiFi i would need a much longer pause

the second line is just to restart the processes at 2AM ... just in case there memory leaks in the app, etc.

One thing I did find (not tried with current versions of the software) was that if dump1090 stopped and restarted while the feeder was running, then the feeder would re-connect.

I am wondering if the problems you were having could be power related ... that the power supply is a bit marginal and the dongle starting up was taking just too much power for the network processes to start too.

If anyone is interested - the 242 raspberry Pi feeder works much better than the 242 generic Arm feeder in terms of not leaving zombie processes all over the place.

However if you have a solution that works, stick with it

[Filldo]

i bought a surge protector that comes with two usb ports, one i use for the pi, the other to power the laptop cooling fan i rigged/ziptied to the pi to cool it. i never figured out why, but the act of plugging in the dongle forced a reboot. i didn't think much of it because a reboot never hurt a machine that bad, but if it is already plugged in on power on there is no problem. i'm also using a powered usb cable of 30ft. i read other people had bad luck with it but i tested the usb extension on windows rtl1090 and now rpi and both work fine.

the pdf said to go with the arm-le version. if you say the rpi version is better then i will give it a try tomorrow and refire the script/cron into one piece and test it again. if it works out better i might just set it to reset once a day vs every 4 hours (was afraid the reason for the feed dying was because of a crash/memory leak somewhere).

i'm on the fence about getting an amp for the antenna (from a cost and effort factor, since antenna is on roof at work). KSJC is sandwiched between two mountains to the east and the west and i figured if i can't see past the mountain the antenna won't see much further past it with an amp. anyone have experience with an amp seeing considerably far over mountains? my antenna is about 40 ft above ground and the mountain to the east is roughly 2500 feet (767m).