ADS-B DIY Antenna

Status
Not open for further replies.

jepolch

Active Member
I spent about an hour and a half working on the roof today. I replaced the coax to the second antenna position and added decoupling coils to each coax. I wish I had made more turns, but too late now. Maybe another day. I also replaced the cantenna with the duplicate spider. Everything looks perfect - BUT - the original spider still outperforms his newer sibling with the newer coax!!

Finally, it occurred to me that maybe the difference is between the two Pi's. After all was said and done, I swapped the antenna feeds between the Pi's and now the new spider outperforms the original one. Both Pi's are identical, running the same version of dump1090-muta and Piaware. The main difference is that the underperforming Pi is running FR24 (node). It runs at a higher load level than the "better" Pi. In a while I'll shut down FR24 on one Pi and PlaneFinder on the other Pi and see how they do when they're both just running Piaware. What a mystery. Note the "before" and "after" in the dump screen shots.
 

Attachments

  • spiders 3-18-15.jpg
    spiders 3-18-15.jpg
    111.4 KB · Views: 78
  • decoupling coils.jpg
    decoupling coils.jpg
    567.1 KB · Views: 80
  • pi2 3-18-15 before.jpg
    pi2 3-18-15 before.jpg
    33.5 KB · Views: 70
  • pi3 3-18-15 before.jpg
    pi3 3-18-15 before.jpg
    32.2 KB · Views: 67
  • pi2 3-18-15 after.jpg
    pi2 3-18-15 after.jpg
    31.7 KB · Views: 64
  • pi3 3-18-15 after.jpg
    pi3 3-18-15 after.jpg
    32.3 KB · Views: 63
Last edited:

jepolch

Active Member
The plot thickens. I moved the antennas back to their original locations. Original spider to Pi2 and new spider to Pi3, but I swapped the dongles between the Pi's. Now Pi2, with the original spider but with Pi3's dongle, is doing slightly worse than Pi3. The difference is in the dongles. They both have the R820T2, but the apparently better performing receiver is the NooElec brand. The underperforming receiver is the RTL-SDR.COM brand. Now I need to try to figure out what they are doing differently and if I can equalize them. At least I won't have to do that on the roof! :D
 

ab cd

Senior Member
@jepolch:
Did you try your new RPi Model 2 ?
I put mine in operation couple of days ago, using the microSD card from RPi B+. The dump1090-mutability suddenly died after running for couple of hours. I rebooted the RPi Model 2, and it started working ok. After couple of hours, dump1090-mutability died again, I rebooted RPi, jt was ok, but same scenario repeated again.

I formatted microSD card, did a fresh install of everything, but above scenario persisted.

Next, I used a brand-new microSD card, installed wheezy & all, and above scenario persisted.

I dont know weather the new RPi model 2 is defective, or dump1090-mutability is incompatable with RPi Model 2 :mad:

I am now going to do a format of microSD card & fresh install of everything, but will install dump1090 Malcom Robson instead of mutability, and see the outcome.
 

ab cd

Senior Member
The plot thickens. I moved the antennas back to their original locations. Original spider to Pi2 and new spider to Pi3, but I swapped the dongles between the Pi's. Now Pi2, with the original spider but with Pi3's dongle, is doing slightly worse than Pi3. The difference is in the dongles. They both have the R820T2, but the apparently better performing receiver is the NooElec brand. The underperforming receiver is the RTL-SDR.COM brand. Now I need to try to figure out what they are doing differently and if I can equalize them. At least I won't have to do that on the roof! :D
The best way to compare different antennas is to just swap antennas, using everything else same.

When I test/compare antennas, I only change antennas above my launch pad. Everything below launch pad is static (i.e. same location, same coax, same dongle, same RPi with same microSD Card, same usb cable to router, same desktop. If I use an amplifier, then same amplifier, same power inserter, same DC supply, & same DC blocker).
 
Last edited:

ab cd

Senior Member
i build a weird antenna, but it performs quite nice for what it is. around 134msg/sec and tracking 17 planes right now, had higher peaks. It basically consists of two dipoles at opposite ends of a beam and one of them is turned 90 degrees clockwise compared to the other one and is a slightly different type of dipole.
View attachment 1365

image of antenna is here
It is unconventional antenna, but great to try. Often unconventional things prove to be a break-through (Earth is not flat, it is round → Columbus Sailing West to reach East→ Discovery of Americas). I like your ingenuity.

I suggest that in place of your 2nd antenna (the center-fed U), try a center fed Z.

The center-fed Z + first normal dipole makes three dipoles, 1st along X axis, 2nd aliong Y axis & 3rd along Z axis. Covers all polarizations :D
 
Last edited:

jepolch

Active Member
@jepolch:
Did you try your new RPi Model 2 ?
I put mine in operation couple of days ago, using the microSD card from RPi B+. The dump1090-mutability suddenly died after running for couple of hours. I rebooted the RPi Model 2, and it started working ok. After couple of hours, dump1090-mutability died again, I rebooted RPi, jt was ok, but same scenario repeated again.

I formatted microSD card, did a fresh install of everything, but above scenario persisted.

Next, I used a brand-new microSD card, installed wheezy & all, and above scenario persisted.

I dont know weather the new RPi model 2 is defective, or dump1090-mutability is incompatable with RPi Model 2 :mad:

I am now going to do a format of microSD card & fresh install of everything, but will install dump1090 Malcom Robson instead of mutability, and see the outcome.
Wow that is weird. But I wouldn't be totally surprised that the RPi2 might crash if you used the SD card from your B+. The two have different hardware, so surely when you installed Raspbian on your B+ it would have been configured for the existing hardware. Put that card into another model and it might have problems. I'd compare that to taking a hard drive with Windows installed out of a computer with an Intel processor and putting it into one with an AMD processor. Turn it on and it might try to boot, but things aren't going to work. The two Pi's are more similar than AMD vs. Intel, but the principle is similar. I would think a fair test would be a fresh install in the RPi2.

OK, the second scenario is different. You set up the RPi2 from scratch and things still aren't working. That puzzles me. I haven't even powered up my RPi2 yet. I'll try to do that tomorrow and let you know what happens. I'll install Wheezy and dump1090-muta like you did and see what happens.

When you said the RPi2 died after a couple of hours, what exactly did you see? You couldn't SSH to it? Couldn't ping it? Were any lights blinking, or on solid? Did you plug the HDMI into a TV to see if you got a command prompt?
 

jepolch

Active Member
The best way to compare different antennas is to just swap antennas, using everything else same.

When I test/compare antennas, I only change antennas above my launch pad. Everything below launch pad is static (i.e. same location, same coax, same dongle, same RPi with same microSD Card, same usb cable to router, same desktop. If I use an amplifier, then same amplifier, same power inserter, same DC supply, & same DC blocker).
You're right. I thought I had swapped the antennas before, but maybe I hadn't. As it turns out it's still more complex than I mentioned earlier. Now that Pi2 is back with original (better) spider, but has the dongle from Pi3, it's (Pi2) not performing as well as it did with good spider and the NooElec dongle. It's like swapping the dongles more or less evened out the performance between the two Pi's and while Pi2 is doing a little worse, Pi3 is doing a little better. Tomorrow I'll put them back in their original configurations and see how they perform. I really don't know what to think! o_O
 

ab cd

Senior Member
Wow that is weird. But I wouldn't be totally surprised that the RPi2 might crash if you used the SD card from your B+. The two have different hardware, so surely when you installed Raspbian on your B+ it would have been configured for the existing hardware. Put that card into another model and it might have problems. I'd compare that to taking a hard drive with Windows installed out of a computer with an Intel processor and putting it into one with an AMD processor. Turn it on and it might try to boot, but things aren't going to work. The two Pi's are more similar than AMD vs. Intel, but the principle is similar. I would think a fair test would be a fresh install in the RPi2.
My reaction was same as yours, so I formatted the microSD Card, and after copying Wheezy image, inserted it into Pi2, and all further installation & configurations were done with microSD card inside Pi2. Still the problem persisted.

OK, the second scenario is different. You set up the RPi2 from scratch and things still aren't working. That puzzles me.
That puzzled me too. I thought may be my microSD card has worn out due to repeated formattings & reinstalls. I took a brand-new microSD card and did a fresh install using desktop then Pi2, but to my despair, result was same :mad:

When you said the RPi2 died after a couple of hours, what exactly did you see? You couldn't SSH to it? Couldn't ping it? Were any lights blinking, or on solid? Did you plug the HDMI into a TV to see if you got a command prompt?
Everything was normal, but when I ssh from desktop & give command "top", I noted cpu load is very low. I checked carefully, found dump1090-mut is not running. I checked my feeder data on sites of flightaware & FF24, found these are showing me live, but data collectd since last two hours is zero!

I give ssh command "sudo reboot" from my desktop, run top command, found dump1090-mut & all are running ok.

After 4 hrs, I ssh from my phone, ran top command, found everything is running ok, but dump1090-mut is missing from tbe list. Again checked Flightaware & FR24 sites for my statistics, found shown alive, but number of planes for last 3 hours is zero!

" sudo reboot" or "sudo shutdown -h -P now"+unplug/replug dc adapter brings dump1090-mut back to life, but only to die again after an hour or two! :mad:
 
Last edited:

giacomo1989

Member
My reaction was same as yours, so I formatted the microSD Card, and after copying Wheezy image, inserted it into Pi2, and all further installation & configurations were done with microSD card inside Pi2. Still the problem persisted.


That puzzled me too. I thought may be my microSD card has worn out due to repeated formattings & reinstalls. I took a brand-new microSD card and did a fresh install using desktop then Pi2, but to my despair, result was same :mad:


Everything was normal, but when I ssh from desktop & give command "top", I noted cpu load is very low. I checked carefully, found dump1090-mut is not running. I checked my feeder data on sites of flightaware & FF24, found these are showing me live, but data collectd since last two hours is zero!

I give ssh command "sudo reboot" from my desktop, run top command, found dump1090-mut & all are running ok.

After 4 hrs, I ssh from my phone, ran top command, found everything is running ok, but dump1090-mut is missing from tbe list. Again checked Flightaware & FR24 sites for my statistics, found shown alive, but number of planes for last 3 hours is zero!

" sudo reboot" or "sudo shutdown -h -P now"+unplug/replug dc adapter brings dump1090-mut back to life, but only to die again after an hour or two! :mad:

Are you using a system image built for rpi2? It is different from the b+
 

xforce30164

Active Member
@jepolch:
Did you try your new RPi Model 2 ?
I put mine in operation couple of days ago, using the microSD card from RPi B+. The dump1090-mutability suddenly died after running for couple of hours. I rebooted the RPi Model 2, and it started working ok. After couple of hours, dump1090-mutability died again, I rebooted RPi, jt was ok, but same scenario repeated again.

I formatted microSD card, did a fresh install of everything, but above scenario persisted.

Next, I used a brand-new microSD card, installed wheezy & all, and above scenario persisted.

I dont know weather the new RPi model 2 is defective, or dump1090-mutability is incompatable with RPi Model 2 :mad:

I am now going to do a format of microSD card & fresh install of everything, but will install dump1090 Malcom Robson instead of mutability, and see the outcome.

Hmm, Just to give my info, my RaspyTwo receiver at my parents house is a raspberry pi 2 and is running just fine, I did indeed put a clean install of rasbian on it. Good luck fixing your RPi 2! Also check the dump-mutability log maybe it gives some insight into what is happening.
 

xforce30164

Active Member
It is unconventional antenna, but great to try. Often unconventional things prove to be a break-through (Earth is not flat, it is round → Columbus Sailing West to reach East→ Discovery of Americas). I like your ingenuity.

I suggest that in place of your 2nd antenna (the center-fed U), try a center fed Z.

The center-fed Z + first normal dipole makes three dipoles, 1st along X axis, 2nd aliong Y axis & 3rd along Z axis. Covers all polarizations :D

I did some quick google searching for the center-fed Z, but haven't been able to find any good information about the Z-type in particular. Do you maybe have a link for me or any other "keywords" I could use in my search?

Also I'm thinking of maybe making an antenna that slowly bends over larger area, someone like a "center-fed C" antenna

Currently working on dynamic page/script generation through php:

http://dump1090.ayadex.com/page.php?g=dump1090&c=y&r=1

(r=1 or r=2 to switch between receivers). In the end I plan to eliminate most javascript files and have them be generated dynamically based on which receiver you want. this should give me slightly more modular and flexible system.

Edit2: I currently got a basic table working that is generated by php based on the combination of both receivers, if it has data for both it will display distance, messages, age and rssi values separately. (I do need to implement several calculations for values)

aircraft combined table php
 
Last edited:

jepolch

Active Member
My reaction was same as yours, so I formatted the microSD Card, and after copying Wheezy image, inserted it into Pi2, and all further installation & configurations were done with microSD card inside Pi2. Still the problem persisted.


That puzzled me too. I thought may be my microSD card has worn out due to repeated formattings & reinstalls. I took a brand-new microSD card and did a fresh install using desktop then Pi2, but to my despair, result was same :mad:


Everything was normal, but when I ssh from desktop & give command "top", I noted cpu load is very low. I checked carefully, found dump1090-mut is not running. I checked my feeder data on sites of flightaware & FF24, found these are showing me live, but data collectd since last two hours is zero!

I give ssh command "sudo reboot" from my desktop, run top command, found dump1090-mut & all are running ok.

After 4 hrs, I ssh from my phone, ran top command, found everything is running ok, but dump1090-mut is missing from tbe list. Again checked Flightaware & FR24 sites for my statistics, found shown alive, but number of planes for last 3 hours is zero!

" sudo reboot" or "sudo shutdown -h -P now"+unplug/replug dc adapter brings dump1090-mut back to life, but only to die again after an hour or two! :mad:
Just by coincidence someone over at FlightAware is reporting the same situation you are having with dump1090-muta, although I think they are using Pi (not a 2). http://discussions.flightaware.com/ads-b-flight-tracking-f21/crashing-for-some-reason-t20422.html Check out that thread and see if obj replies with ideas.
 

jepolch

Active Member
@jepolch have you find to have more aircraft after you make the common mode currents blocker loop?
So far, the numbers are down today. It could be the weather. It's overcast and foggy and my reception is usually lower on foggy days. The best reception is on clear, sunny days. So, I can't really tell anything yet regarding the coils. As for my receivers, and the issues I was having yesterday, I have my Pi's back to their original configurations and things are about equal between them now, but like I said, the numbers are low today. :(
 

giacomo1989

Member
So far, the numbers are down today. It could be the weather. It's overcast and foggy and my reception is usually lower on foggy days. The best reception is on clear, sunny days. So, I can't really tell anything yet regarding the coils. As for my receivers, and the issues I was having yesterday, I have my Pi's back to their original configurations and things are about equal between them now, but like I said, the numbers are low today. :(
Extensive test needed. Thanks for the moment.
PS. I'm also waiting for abcd bazooka dipole lol
 

jepolch

Active Member
@ab cd: I have my Pi 2 set up now. It's running dump1090-muta and lighttpd. No feeder package yet. I'm just watching it for now while I get the OS set up to my liking. So far, so good. I'll let you know if it runs into any problems.
 
Status
Not open for further replies.
Top