I received my Odroid N2 today, and have managed to get piaware 3.63/Dump1090-fa running fine with a RTL-SDR v3, but am not having any success with getting an Airspy running.
OS is ubuntu 18.04.2-4.9 running in 64bit - tried at first to just get the Airspy working on its own, but no joy at all.
The Airspy host software seems to be working fine as the Airspy is recognised, and I managed to run the test command “airspy_rx -r /dev/null -a 20000000 -t 3” as recommended by @prog which returned a rate of 20+ MSPS and cpu usage is very low on the RTL, so would be good to see what could be achieved with the airspy
I was hoping they would have released an armbian version for the N1 by now. Maybe in the near future. I am not sure how close Armbian is to Ubuntu. I may have time to give it a try this weekend. I should have time by the end of the month.
You need an airspy_adsb binary compiled for arm64 (64 bit kernel)
The only arm binary offered on the airspy website right now is for armhf (32 bit kernel).
So either you somehow manage to get a 32 bit kernel running or you’ll have to wait until an arm64 version is published.
(The hardware can run 32 bit as well just like with x86_64 hardware still being able to run 32 bit. But using a 64 bit kernel you can’t run 32 bit binaries without multiarch support or w/e, i’m not exactly sure.)
And no obviously my install script doesn’t work as it can’t get the correct binary.
Also the adsb-receiver script won’t help in getting the airspy_adsb program to run.
(It may help you to compile dump1090-fa)
So i guess i would advise you to contact airspy and ask for an airspy_adsb binary for arm64.
root@odroid:~# systemctl status airspy_adsb
● airspy_adsb.service - Airspy ADS-B receiver
Loaded: loaded (/etc/systemd/system/airspy_adsb.service; enabled; vendor preset: enabled)
Active: active (running) since Thu 2019-04-11 13:54:21 UTC; 3min 25s ago
Docs: HOWTO: Airspy mini and Airspy R2: Piaware / dump1090-fa configuration - #2 by wiedehopf
Main PID: 2071 (airspy_adsb)
CGroup: /system.slice/airspy_adsb.service
└─2071 /usr/local/bin/airspy_adsb -v -f 1 -b -l 29999:beast -l 47806:asavr -c localhost:30104:beast -g 19 -m 20
Apr 11 13:54:21 odroid systemd[1]: Started Airspy ADS-B receiver.
Apr 11 13:54:21 odroid airspy_adsb[2071]: Listening for beast clients on port 29999
Apr 11 13:54:21 odroid airspy_adsb[2071]: Listening for asavr clients on port 47806
Apr 11 13:54:21 odroid airspy_adsb[2071]: Acquired Airspy device with serial XXXXXXXX
Apr 11 13:54:21 odroid airspy_adsb[2071]: Decoding started at 20 MSPS
Apr 11 13:54:33 odroid airspy_adsb[2071]: Push client connected to localhost:30104 (beast)
root@odroid:~#
Sure enough, MyADS-B page claims it is connected and feeding, including MLAT.
In the next few days (barring business travel) I will start with a fresh sd-card from scratch. That can see if ln -s arm-linux-gnueabihf/ld-2.23.so ld-linux.so.3 is necessary. it is included today.
I agree… my guess is that you and I think the same about it. Let’s see what the minimum is that we can get away with to get a clean install. You are a rockstar for figuring out how to get it run. Thank you
I gave up because of the apt update error you had. Didn’t realize that wasn’t a problem.
(Only the hardkernel specific stuff like the kernel isn’t available in armhf (32bit), but the rest of the Ubuntu sources including the libraries are available as armhf)
@wiedehopf
Just to let you know, I have done another clean build, using the above you suggested and all working hunky dory - looks like that ’ ln -s arm-linux-gnueabihf/ld-2.23.so ld-linux.so.3’ wasn’t needed after all
Once again, VMT for all your hard work in getting this going.
Now, to compare the xu4 against the N2 with same gain settings to see which one wins!!
I don’t check for lost samples, but both units manage over 20 MSPS, so I don’t think that will be a problem.
I have been comparing daily position/plane numbers between both units (xu4’s), and interesting to see they are quite close with about +4% difference in AC reported between gains of 21 to 16, and positions reported difference under 1%.
gain 21 gives highest AC count and 16 gives most messages with me.
I will setup the logging sometime - using your manual setup which works very well
Aren’t you very close to SFO? I would expect a gain of 16 or 17 to give better results with an LNA and good antenna above the roof.
I am close to SFO, indeed. Before I left for the office this morning I set the gain to 18. Let’s see how it will run. I may have to play around with the gain a bit to find the optimal one.
There is an Armbian image now available for the N2 - ver.5.86.
I have just loaded the ‘stretch’ version from their download page https://www.armbian.com/odroid-n2/ and whilst it is not officially supported, some clever people have been working on various images for a while.
It works very well indeed - similar cpu% being used compared to the ubuntu image from odroid, but htop etc work on the Armbian image, where I had problems with ubuntu.
Well worth a try - have piaware/dump-fa 3.71 working here fine.