ADS-B DIY Antenna

Status
Not open for further replies.
The crashes of dump1090-mutability
22 times in last 3 weeks! :mad:

I had to manually reboot RPi, as restarting dump1090-mut was not enough. Collectd & pfclient often wont reconnect to dump1090-mut even after dump1090's restart.

Now I have done this to automatically reboot RPi:
Code:
sudo crontab -e
#above command opened crontab file for editing
#added following below last line of file
0 4,10,16,22  *  *  *  sudo reboot
#my time offset from UTC is -4 hrs,
# 4,10,16,22 hrs utc is 0,6,12,18 hrs local time
#saved (ctrl+o) and exit (ctrl+x)
#message displayed:
crontab:  installing new crontab

View attachment 2174
Still happening? Sorry to hear it. That's a tough nut to crack.
 
After 2½ day's trial run, the 6-legged F-connector spider is getting roughly 20% more positions than 4-legged SO239 spider. This may also be due to day to day fluctuations of traffic. I cant be very sure due to frequent crashes of dump1090-mut .....Grrrrrr.
Well, I'm glad to hear the antenna is working well for you. So have you tried the one with the amp yet? I've been lazy, I still haven't made one.
 
Well, I'm glad to hear the antenna is working well for you. So have you tried the one with the amp yet? I've been lazy, I still haven't made one.
The coax between antenna & receiver is so short, I am afraid amplifier may overload front end of RTL tuner chip. Due to crashes of Dump1090-mut, I have inserted a diplexer (3 dB attenuation) and lowered the gain from max to 40, still the crashes are there. Just a brain wave! The amplifier has also a range 950 to 2050 Mhz, that means it will not amplify lower frequencies. I will put amplified Spider in system and see if dump stops crashing or not.
 
The crashes of dump1090-mutability
22 times in last 3 weeks! :mad:

I had to manually reboot RPi, as restarting dump1090-mut was not enough. Collectd & pfclient often wont reconnect to dump1090-mut even after dump1090's restart.

Now I have done this to automatically reboot RPi:
Code:
sudo crontab -e
#above command opened crontab file for editing
#added following below last line of file
0 4,10,16,22  *  *  *  sudo reboot
#my time offset from UTC is -4 hrs,
# 4,10,16,22 hrs utc is 0,6,12,18 hrs local time
#saved (ctrl+o) and exit (ctrl+x)
#message displayed:
crontab:  installing new crontab

View attachment 2174

I'm running dump1090-mut on my rpi2 for two days.. no crashes :)
 
Hi Mukundan

Thanks for starting this thread. This is the first time you have posted in this thread after starting it!

After you started this thread, I was the first one to post there. For quite some time I was the only one posting in this thread. Occasionally some one will make few posts and fade out. Later trigger (Dave) started to post regularly. In Nov 2014, about 1-1/2 years after start, this thread had only 18 pages. Then in Nov 2014 giacomo1989, and jepolch joined this thread and started posting regularly.

Later in begening of 2015 others such as Sjacket99, garylovesbeer, xforce30164, radiostationx, bramj, TomW, and many others (forgive me if I forgot to mention their names) joined this thread and started posting, and as a result in a small span of time, this thread grew exponentially to 176 pages as it is today!

I am pleased to see your first & only post after starting the thread.

ab cd

Many thanks ab cd !! Yes. I started the thread and vanished only because, PF sent me their receiver setup to feed from my location. :)
Even though I didn't post anything , I used to read each and every reply in this thread
 
I've been running it for months on a RP1 2 and RPi B+ and no crashes. Go figure. :confused:
I have two independent systems with two antennas, two coax & two RPis.
The RPi in system 1 keeps crashing, in system 2 keeps running.
I swapped the two RPis, still system 1 keeps crashing, system 2 keeps running.
I swapped the two power supplies, still the system 1 keeps crashing, system 2 keeps crashing.
I swapped the two DVB-T dongles, still system 1 keeps crashing, system 2 keeps running.
I purchased new microSD cards and made fresh install of entire software, still system 1 keeps crashing, system 2 keeps running.
I inserted a diplexer in system 1, still the system 1 keeps crashing, system 2 keeps running.
I connected power adapter of system 1 to wall socket of another room using a 110V AC power extension chord, still system 1 keeps crashing, system 2 keeps running.
 
Last edited:
Still happening? Sorry to hear it. That's a tough nut to crack.
Yest, it has proved to be a hard nut to crack. I am unable to find the cause of this problem.

Now I have sent an email for help to Prof. Albus Dumbledore of Hogwarts, with cc to Ron Weasley, Hermione Granger, and Harry Potter :D
 
Last edited:
Yest, it has proved to be a hard nut to crack. I am unable to find the cause of this problem.

Now I have sent an email for help to Prof. Albus Dumbledore of Hogwarts, with cc to Ron Weasley, Hermione Granger, and Harry Potter :D
They'll tell you to use your magic wand. ;)
 
Yest, it has proved to be a hard nut to crack. I am unable to find the cause of this problem.

Now I have sent an email for help to Prof. Albus Dumbledore of Hogwarts, with cc to Ron Weasley, Hermione Granger, and Harry Potter :D

Hi ab cd,
Have you tried changing the net-ro size to something different than the default 500 as part of your diagnostics ?
Also could it be the RTL is going into power save mode and causing mutability to seem to crash when really its the dongle dying a death not the decoding software. There are fixes for that issue.
 
TRIAL RUN RESULTS:

dump1090-rpi-acs-30d-R.PNG


dump1090-rpi-range-30d-R.PNG


dump1090-rpi-cpu-30d-R.PNG
 
Hi ab cd ,
I have a cron job
* * * * * /ping c2 192.168.1.254 > /dev/nul

the ip 192. is my router, this sends a ping every minute, every hour , every day

i also have this


0 * * * * sudo /sbin/reboot
a reboot on the hour, every hour
The pi 2 boots into the decoder in around 20-25 secs.
The logic being that if I did suffer a crash,the worst that could happen would be the loss of 59 mins worth of data and the next timed reboot would cure the issue.


I run the wifi check script too in cron every 2 minutes but I dont know if you are linked to your router by wifi or eth.
The script for wifi is here, i put it in /usr/local/bin

https://raw.githubusercontent.com/dweeber/WiFi_Check/master/WiFi_Check

In this script the router address must be specified/altered to suit your network.
Running it in cron every 2 mins is overkill but system reliability/wifi network link reliability is the aim so every 2 mins it is ...

Finally,I turned off the power saving of my realtek wifi dongle by making a modification file with nano in /etc/modprobe.d
These combined made my pi super stable and reliable but I am not using Mutab due to the problems I had with it, maybe just me/being unlucky but I eventually settled for Malcom Robb version as it runs with no issues at all. I favored decoding data and getting aircraft messages "thro the box" rather than have greater control with bells and whistles (graphs) which was nice but having to put up with constant crashes/ssh freezes or other issues.

It would be interesting if you tried D1090 malcom robb (vanilla/basic decoder) on a new card/installation and compare. Leave it running and see if that does the business.
No graphs, No mods, no messing.. but it decodes messages and it works.
I find it very reliable indeed.
Sadly, You dont have much to lose by trying .
 
Last edited:
@radiostationx:
Thanks for the detailed explanation of your setup, which provides a very stable data feed.

What puzzles me is that if I swap Pis between two systems, crashing one stops crashing, and non-crashing one starts crashing. I have tried swapping DVB-T dongle, and DC adapter also, but result is same: system1 crashes, system2 does not. So it is neither RPi, software, dvb-t dongle or power supply. It is something specific to system 1 (excluding above noted swapped items) which is causing crashes.

I will do a fresh install of all software on RPi B+ and install dump1090 vanilla (malcom robson), and give it a try.

Actually I need the graphs with system 1 (crashing one) as I use it for trial run of antennas. This system has no amplifier and short (12ft/4m) coax, thus tests the raw antenna without amplifier modifying its real output.

I use system 2 (stable one) with long coax (50ft+15ft/15m+5m) and amplifier near antenna, to feed data to various sites.
 
@radiostationx:
Thanks for the detailed explanation of your setup, which provides a very stable data feed.

What puzzles me is that if I swap Pis between two systems, crashing one stops crashing, and non-crashing one starts crashing. I have tried swapping DVB-T dongle, and DC adapter also, but result is same: system1 crashes, system2 does not. So it is neither RPi, software, dvb-t dongle or power supply. It is something specific to system 1 (excluding above noted swapped items) which is causing crashes.

I will do a fresh install of all software on RPi B+ and install dump1090 vanilla (malcom robson), and give it a try.

Actually I need the graphs with system 1 (crashing one) as I use it for trial run of antennas. This system has no amplifier and short (12ft/4m) coax, thus tests the raw antenna without amplifier modifying its real output.

I use system 2 (stable one) with long coax (50ft+15ft/15m+5m) and amplifier near antenna, to feed data to various sites.

Hi ab cd,
You can still use "system 1" as an antenna testbed if you install pf client on top of your d1090 malcom robb.
Then simply use :30053 stats as a live message rate meter for antenna tests.
The live map comes in handy too on :30053.

and your place in the planefinder sharer list for "system 1"

https://planefinder.net/sharing/

"daily breakdown for <date>" tab


which will show you a polar plot which you can move the mouse pointer around for precise range.


The good thing with this is , it will keep plots per day in the historical data in your sharer list name.
There is not much that cannot be done with d1090mr teamed up with some feeder software.

Rolling minute view is nice in collectd but :30053/ stats gives the same and is live.

All this is not quite the same as the slick collectd graphs by xforce and cauis, I concede that, but its functional all the same for testing purposes.
In short, if the MR decoder works with no issues, its a no brainer,keep it simple.


One thing which confuses me, I am yet to figure out..
Planefinder feed (sharer list) says I get circa 1100-1300 aircraft per day, FA on the other hand says 2400 aircraft have been seen by the same station on the same day.

Shout to fellow pf/fa feeders , are your figures similar ?
 
Last edited:
Hi ab cd,
You can still use "system 1" as an antenna testbed if you install pf client on top of your d1090 malcom robb.
Then simply use :30053 stats as a live message rate meter for antenna tests.
The live map comes in handy too on :30053.

and your place in the planefinder sharer list for "system 1"

https://planefinder.net/sharing/

"daily breakdown for <date>" tab


which will show you a polar plot which you can move the mouse pointer around for precise range.


The good thing with this is , it will keep plots per day in the historical data in your sharer list name.
There is not much that cannot be done with d1090mr teamed up with some feeder software.

Rolling minute view is nice in collectd but :30053/ stats gives the same and is live.

All this is not quite the same as the slick collectd graphs by xforce and cauis, I concede that, but its functional all the same for testing purposes.
In short, if the MR decoder works with no issues, its a no brainer,keep it simple.


One thing which confuses me, I am yet to figure out..
Planefinder feed (sharer list) says I get circa 1100-1300 aircraft per day, FA on the other hand says 2400 aircraft have been seen by the same station on the same day.

Shout to fellow pf/fa feeders , are your figures similar ?
Thanks for advise about using pf stats & polar graph. Yes, I used it before installing collectd rrd tools. I will format my microSD card and do a fresh install using dump1090 vanila (mr).

But before that, I have done one more attempt to find the cause. Since the swap of RPi, DVB-T dongle, DC adapter, all have failed, I have now replced the remaining bits of system1, i.e I removed smx to F pigtail, the rg6 coax from pigtail to antenna, the diplexer, and the antenna. I then connected stock antenna directly to dvb-t dongle, cut antenna whip to 69mm, and placed the stock antenna over a metallic plate near window. I also changed gain in dump1090-mut from 40 to max.

I will run this for few days and see what happens now.
 
Last edited:
Couple of days ago, I added following line in crontab
0 0,6,12,18 * * * sudo reboot
To my surprise, RPi refused to reboot at specified times. I googled & checked a million web pages, but found nothing wrong in what I have done, till I gave up. Then finally I found it! Instead of typing numerals 0 0 at start of line, I have typed alphabet o o :(. What a silly mistake!

I could not find the problem because I did it using SSH from my phone, whose tiny screen could not clearly show difference between o and 0.

I even tried to debug using SSH from my phone, but could not find the problem. Only when I SSH from desktop, I could immediately notice I have typed alphabet o, not numeral 0 :mad:
 
Last edited:
Hi ab cd,
You can still use "system 1" as an antenna testbed if you install pf client on top of your d1090 malcom robb.
Then simply use :30053 stats as a live message rate meter for antenna tests.
The live map comes in handy too on :30053.

and your place in the planefinder sharer list for "system 1"

https://planefinder.net/sharing/

"daily breakdown for <date>" tab


which will show you a polar plot which you can move the mouse pointer around for precise range.


The good thing with this is , it will keep plots per day in the historical data in your sharer list name.
There is not much that cannot be done with d1090mr teamed up with some feeder software.

Rolling minute view is nice in collectd but :30053/ stats gives the same and is live.

All this is not quite the same as the slick collectd graphs by xforce and cauis, I concede that, but its functional all the same for testing purposes.
In short, if the MR decoder works with no issues, its a no brainer,keep it simple.


One thing which confuses me, I am yet to figure out..
Planefinder feed (sharer list) says I get circa 1100-1300 aircraft per day, FA on the other hand says 2400 aircraft have been seen by the same station on the same day.

Shout to fellow pf/fa feeders , are your figures similar ?
My numbers are very different between PF and FA. I see a lot more on PF than FA. There's a thread on FA where people were comparing their stats. I can't find it right now. The FA staff explained how they count differently on each site. If I locate the thread I'll post a link here.

Edit: I found the thread. It was a comparison between FR24 and FA, but I mentioned my PF stats in it. http://discussions.flightaware.com/ads-b-flight-tracking-f21/fr24-and-flightaware-stat-comparison-t24640.html

Yet another edit: Ah, my memory is faulty. It's just FR24 vs. FA.
 
Last edited:
Couple of days ago, I added following line in crontab
0 0,6,12,18 * * * sudo reboot
To my surprise, RPi refused to reboot at specified times. I googled & checked a million web pages, but found nothing wrong in what I have done, till I gave up. Then finally I found it! Instead of typing numerals 0 0 at start of line, I have typed alphabet o o :(. What a silly mistake!

I could not find the problem because I did it using SSH from my phone, whose tiny screen could not clearly show difference between o and 0.

I even tried to debug using SSH from my phone, but could not find the problem. Only when I SSH from desktop, I could immediately notice I have typed alphabet o, not numeral 0 :mad:
Sometimes it's necessary to give the full path, like this 'sudo /sbin/reboot'. Might help.
 
I recall that sudo/sbin/reboot is prefered over sudo reboot as the pi will prepare itself first and execute a clean reboot so that is why I used it in my crontab every hour, on the hour,every day, every week.
The pi2 unit running d1090 + 4 feeders completes the reboot in 20-25 secs typically.
I know it probably has log entries building for this timed task in var/log and i have not bothered to log rotate but i have a 32gb class10 sd and to date "df" tells me i have used 21% of the space available s i have some time to go before capacity is a concern.
 
Status
Not open for further replies.
Back
Top