So I have an airspy mini, and a FA prostick. I have them running in tandem on a splitter.
The net of this is the airspy picks up more messages per aircraft, but overall less aircraft.
I have tried getting “support” from the airspy folks, but all I get is blame that its on my end, even blaming using php for the portal.
So In the end I feel its a huge waste of money to buy an Airspy for ADSB. I will post up some stats after the 2 systems run in parallel for a longer period of time.
here is the entire conversation with Prog the person that writes the ADSB code for Airspy. Just in case anyone else wants to see the interaction I had
[16:35:48] <|daver|> hey prog you around?
[16:35:49] Yum (~H2O-ScR@dynamic.rabatp1-117-59-137-41.wanamaroc.com](mailto:~H2O-ScR@dynamic.rabatp1-117-59-137-41.wanamaroc.com)) joined the channel
[16:39:10] yes I am
[16:39:10] nah he’s a square
[16:39:22] I’m round today
[16:39:33] triangular tomorrow.
[16:39:38] <|daver|> haha
[16:39:46] <|daver|> see if you can get to this webpage
[16:40:10] <|daver|> 162.236.27.45:9000/dump1090.php
[16:40:20] <|daver|> 162.236.27.45:9001/dump1090.php
[16:40:28] .php ?
[16:40:49] <|daver|> yup
[16:41:01] splitter?
[16:41:03] <|daver|> yup
[16:41:06] good
[16:41:06] <|daver|> first one is airspy
[16:41:10] <|daver|> second is FA prostick
[16:41:31] <|daver|> aircraft on airspy is more because its picking up ground , wait until those get filtered
[16:41:37] still loading
[16:41:56] <|daver|> ok
[16:41:58] picking ground is bad?
[16:42:05] <|daver|> it gets filtered, its not consistent
[16:42:11] <|daver|> I think dump1090 does that
[16:42:20] <|daver|> they just disappeared
[16:43:30] what gets filtered?
[16:43:31] <|daver|> should be a good comparison now
[16:43:35] <|daver|> the ground aircraft
[16:43:39] <|daver|> they are gone now
[16:44:19] why are you filtering them?
[16:44:30] <|daver|> im not
[16:44:38] <|daver|> they just will show up then disappear
[16:44:53] check your cables
[16:44:57] <|daver|> my guess is dump1090 eliminates it
[16:45:07] the number of aircraft changes heavily
[16:45:27] <|daver|> yup
[16:46:40] <|daver|> I am using some male to male adapters
[16:46:43] doesn’t look right
[16:46:51] <|daver|> which doesnt look right?
[16:47:00] jumps between 160 to 240
[16:47:08] <|daver|> hmm I dont see that
[16:47:22] <|daver|> when it goes to 240 thats when there is ground
[16:47:46] check your cables
[16:47:54] <|daver|> right now Airspy is 177 aircraft
[16:48:01] <|daver|> FA is 186
[16:48:04] <|daver|> thats what I see currently
[16:48:15] why do I see 240 ?
[16:48:37] fix your setup
[16:48:41] <|daver|> its probably lag dude
[16:48:45] <|daver|> let me check something
[16:48:58] why are you running php in the rpi’s ?
[16:49:03] skhreze_ (~debian@ip-5-172-247-198.free.aero2.net.pl](mailto:~debian@ip-5-172-247-198.free.aero2.net.pl)) left IRC (Ping timeout: 240 seconds)
[16:49:07] <|daver|> because thats the software dump1090 uses
[16:49:13] <|daver|> yours uses it too
[16:49:21] dump1090 doesn’t require php
[16:49:24] <|daver|> its part of the port softwarte
[16:49:26] <|daver|> its wraps it
[16:49:34] <|daver|> click on performance graphs
[16:49:34] use the official thing
[16:49:38] <|daver|> and you can see more stuff
[16:49:46] <|daver|> this is offical, it just has a portal around it
[16:50:15] interesting
[16:50:17] <|daver|> both setups are using same software
[16:50:22] the max range is now better on airspy
[16:50:38] right.
[16:50:48] fix your stuff for good statistics.
[16:50:57] it behaves erratically
[16:50:59] <|daver|> I dont follow on what you are suggesting to fix
[16:51:03] <|daver|> its the connection most likely
[16:51:06] <|daver|> my bandwidth is limited
[16:51:11] remove that php thing
[16:51:21] <|daver|> I cant just “remove” it
[16:52:09] do you realize you are running php cgi on a poor RPi ?
[16:52:22] Looks like 198/179 vs 203/184.
[16:52:40] <|daver|> yes and its only using 25% cpu total
[16:52:50] <|daver|> so I dont think the php wrapper is affecting me
[16:53:06] it’s not letting airspy_adsb to breath
[16:54:10] <|daver|> airspy is on an odroid-c2
[16:54:20] <|daver|> cpu is around 30% on all cores
[16:55:51] Kurlon (~Kurlon@cpe-104-228-216-170.maine.res.rr.com](mailto:~Kurlon@cpe-104-228-216-170.maine.res.rr.com)) joined the channel
[16:56:26] <|daver|> its interesting to look at the map and compares which planes 1 setup sees vs the other
[16:57:19] good
[16:57:20] have fun
[16:57:48] <|daver|> so no idea why the airspy sees less planes using the same antenna?
[16:58:28] fix your stuff for good statistics.
[16:58:44] <|daver|> dude, thats vague, my stuff isnt “broken”
[16:58:57] <|daver|> just because its using a php wrapped doesn’t mean its broke
[16:59:18] you will know it’s fixed when you no longer see spikes in the number of planes
[16:59:24] <|daver|> I dont see spikes
[16:59:32] <|daver|> I think thats because its going over the net
[16:59:37] <|daver|> connected locally its fine
[16:59:47] 162.236.27.45:9000/graphs/dump10 … 608730.269
[16:59:56] that’s a spike
[17:00:08] <|daver|> correct, thats when it picks up aircraft on the ground
[17:00:18] <|daver|> it happens all the time on every setup I have ever run
[17:00:31] <|daver|> I am 13 miles from DFW airport, it will pickup like 50+ aircraft
[17:01:45] make it see them permanently. planes on the ground do not just disappear.
[17:01:59] <|daver|> I am not aware of a way to do that
[17:02:13] check your cables
[17:02:25] and good luck!
[17:05:36] <|daver|> pretty sure the aircraft show up on the ground because I am on the fringe on what it can see that close
[17:05:42] <|daver|> dont think it has anything to do with setup
[17:09:24] <|daver|> look at the aircraft on the TX/LA state line
[17:09:30] <|daver|> thats what I was describing before
[17:10:49] you’re the expert
[17:11:23] <|daver|> thanks for being so condescending
[17:12:02] <|daver|> I do everything you suggest and you blame me, thats cool man. Or the fact a $20 dongle beats your $100 dongle
[17:12:09] <|daver|> if you dont wanna make a better product than fine
[17:12:27] <|daver|> ill make sure to write an honest review for anyone else that wants to avoid using your overpriced stuff
[17:12:49] man I’m trying to help
[17:13:06] <|daver|> by blaming stuff that has no effect?
[17:13:15] <|daver|> or that I cant help ground aircraft show up
[17:13:26] <|daver|> pick 1 plane and compare the signal
[17:13:39] my only problem, I admit, is that I never insist when someone doesn’t understand on the first hit
[17:13:41] <|daver|> i cant make this shit up did
[17:13:58] so, good luck! 
[17:14:22] <|daver|> yeah thanks, so sincere
[17:14:49] with some more experience you will be able to detect these problems and fix them
[17:15:13] meanwhile, you will be on a roll of false assumptions
[17:15:44] <|daver|> so again blame the end user, great way to support a product
[17:15:52] <|daver|> hey its not my fault you are the idiot
[17:16:03] <|daver|> go get a splitter, go do this, oh yeah, still your failt
[17:16:18] no pb.
[17:17:58] <|daver|> your bedside manner lacks quite a bit
[17:20:13] I’m just too tired to explain from A to Z
Looking forward to seeing the stats.
A maybe-dumb suggestion: have you tried swapping the two outputs of your splitter, just incase the splitter isn’t splitting evenly?
I just installed the splitter today,but will be doing that in the future just in case. I got the splitter to make sure my eyeballing of the results was backed up by facts and not just an assumption.
Yeh, he can be abrasive. I had to work out how to get it to work correctly myself. He was half helpful.
There is a dump1090 for the Airspy but I couldn’t get it to compile.
Does your splitter isolate the feeds well? If not, then the noise they generate can interfere with each other (the dongle).
The Airspy really needs more CPU and USB bandwidth than an RPI3. I have seen great reviews using an Odroid XU4(I have one on order). Here is a recent review groups.yahoo.com/neo/groups/Mod … opics/2603
It works OK on an RPI3 provided you keep the RPI3 cool and don’t access dump1090. It overloads the shared USB bus (shared with Ethernet, SD card and all 2/4 USB slots). MLAT drops out when it detects the CPU is too hot or too heavily used.
I found that I needed a HAB/Nevis amp/filter to get it to work better than an FA Pro stick.
I just added a fan to keep the RPI3 cool and that seems to help.
The Airspy is a great multi use receiver. It has a really wide range, wide continuous receive bandwidth and better sensitivity than the RTL style dongles. I think that if you used it in a more powerful machine, it would perform much better than the FA Pro stick.
Splitter is a mini circuits. I am running the airspy on an ODROID-c2 so CPU usage hovers around 30%
as for Prog, he just banned me from their IRC channel, he has pretty thin skin if you try to suggest his product might actually have an issue.
Dave,
Could you provide the airspy_adsb command you are using?
Jon
Same one you are -p -b -w 4 -g 21
looks good.
I was after the port setup. Wanted to make sure you were getting MLAT too.
(Sounds like you know your stuff well enough to have this correct too. Maybe you have it working better than I do)
I set it up just like you did where airspy_adsb sets up a server on port 30005 and feeds the piaware MLAT port.
I tell Piaware to connect to 30005 for MLAT.
airspy connects to 30104 (dump1090-mut) for the non-MLAT traffic.
my airspy_adsb run command
sudo /home/pi/airspy_adsb -c localhost:30104:beast -l 30005:beast -g 21 -p -w 4 &
(bias-t is removed. The amp is next to the RPI and I am powering it directly via USB. Trying to eliminate one item at a time to fault find)
I think I am duplicating some traffic but haven’t worked out a better way to do it.
Agreed. Probably has great utility in other applications, but for ADS-B the R802T2 and FA Prostick are good enough at <20% the cost.
The Prostick has an inbuilt amplifier and the Airspy doesn’t… So if the Airspy is picking up less aircraft and more messages is it not the case that the Prostick is picking up planes further away (Due to the amplification) but the Airspy isn’t seeing these?
During the development of the airspy mini they will have tested under more consistent conditions that we will at home so he will know exactly how the airspy mini will perform for ADSB when set up correctly. The SDRSharp twitter account said that ADSB was the use case they optimised the most. So if he’s confident in the software and the hardware then in his mind there must be setup issues. Also, i think that him replying in a second language makes his response look worse that it actually is
Ste
(Forgot to add, i have an Airspy R2 attached to a raspberry pi 3 using a HABamp, with a seperate pi running dump1090-mutability to view the output from it via a push feed from the pi3. The range and message count are consistently better than the Flightaware stick when used on the same antenna. I’m not knocking the flightaware stick though, its excellent!
I am using a HAB supplies LNA and SAW filter in front of both dongles in my testing.
also to make sure I wasnt exceeding the limit of the Pi, I bought an ODROID-c2 to run it on.
I just swapped my Airspy2 with an Airspy mini(arrived yesterday).
I did a quick reboot and the Overall CPU Utilisation stats have dropped significantly.
Down from 25-30%(for the last few weeks) to 18-20%.
It is a quiet time of the day, only 80/60 aircraft seen by my Basement Antenna.
Edit: NEVER MIND. For some reason, after a reboot, the CPU is back up to Airspy2 levels.
A fan, however, does really help with temps. Consider connecting it to the 3.3V source if it is too loud.
In my basement, the CPU temp went from a peak of 140-145F during the day, a peak of 105F.
I am curious to see what your stats are. R2 compared to the mini
Aircraft wise, the stats look the same.
CPU temp is also down a little, however, this is to be expect with the lower CPU usage. I recently added a fan as the CPU usage was causing MLAT to drop out.
I will try to connect it to my attic antenna for a test, in the next week. The FA Dongle, using the attic antenna, gets twice the A/C as the airspy at the moment. Attic is 20ft higher and has a 9db antenna (basement has $5 cheapo antenna, 3ft agl next to the house), so it is hard to compare apples to apples. When they were side by side, I think the airspy got 10-20% more A/C.
so if you look at my stats flightaware.com/adsb/stats/user … tats-13373
site 13373 is a FA prostick behind a HAB supplies LNA/SAW filter running gain of 30
site 23634 is Airspy mini behind same HAB supplies amp and both are connected to a FA antenna mounted on the roof.
The net of this is the airspy picks up more messages per aircraft, but overall less aircraft.
More messages from an AirSpy is what I would expect not necessarily more aircraft.
well after looking at my stats today, it got less messages and less aircraft.
