FlightAware Discussions

V5.0.5-airspy dump1090-fa with native AirSpy support now available

Oops :slight_smile:
But only 3 messages will still result in distorted average if i’m not mistaken.
I suppose aircraft with less than 8 messages i could just exclude in graphs1090.
But just emitting -49.5 until a solid average is established might be the better solution and done in dump1090-fa.
It’s pretty minor anyhow.

These pics were both live but  very different  in stats 
i will have to read and read  to get a better profile  

I must have the filter before the LNA because of these. :slight_smile:
image
With filter first:
image
With LNA first, it’s just trash.

@g7ruh 's numbers, even with the airspy mini just don’t “look” right. :slight_smile:

Yeah, that’s why I’d prefer to see the results from the dump1090 stats output. I know how those are calculated :slight_smile:

Strange
i just added caius conf above
only to notice i don’t have the file --wisdom /etc/dump1090-fa/wisdom-airspy.arm64"

i also dont have my “sample option conf” has no --wisdom /etc/dump1090-fa/wisdom.local
so # out that line and gosh the whole page exploded with planes and messages
but coming down to earth i spose i do need these proper files to work
so looks like i got a bit missing

Ok so far i have  cp  all 3 files 
none have been altered so no permissions  etc 
which i had problems last time 
i have tried  chmoding them but with no success 

pi@raspberrypi:/etc/dump1090-fa $ ls
wisdom.aarch64  wisdom-airspy.local  wisdom.local
pi@raspberrypi:/etc/dump1090-fa $ sudo chmod a+r /etc/dump1090-fa/wisdom.local
pi@raspberrypi:/etc/dump1090-fa $ ls
wisdom.aarch64  wisdom-airspy.local  wisdom.local
pi@raspberrypi:/etc/dump1090-fa $ sudo chmod a+r /wisdom.local
chmod: cannot access '/wisdom.local': No such file or directory
pi@raspberrypi:/etc/dump1090-fa $ sudo chmod a+r /etc/dump1090-fa/wisdom.aarch64
pi@raspberrypi:/etc/dump1090-fa $ ls
wisdom.aarch64  wisdom-airspy.local  wisdom.local

I see why you need a “big” filter first, all that RF will desense any receiver!!!

My location is very, quiet RF-wise at and around 1090Mhz.

The only thing different from a comparision point of view is that the airspy is not as good at the more distant messages, but is a little better at the close in ones.

The airspy seems to have a better dynamic range than the ‘standard’ dongles.

in good conditions or bad (had both while testing) the airspy is consistently 15-20% down on number of aircraft and these are the distant ones (from about 160 nm).

Sadly for this discussion, it is a single package here

OK, how about the airspy using airspy_adsb?

Gotcha.

Happy to run a test for Airspy Mini using airspy_adsb.
What config should I use for dump1090-fa?
Just want to make sure you get the data you need so we can compare apples and apples :slight_smile:

It doesn’t matter much. None of the gain, sample rate, sample format or demod settings will have any effect.

Thanks, that is what I assumed, but better to check the actual requirement!

Here is the airspy_adsb signal graph for 2 hours

2hr-airspy_adsb-signal-level

and the reference system for the same period

2hr-reference-orange fa dongle_adsb-signal-level

Let me know if you would like a longer timeframe or if this is enough?

Observations: airspy_adsb is much closer to the reference system comparing the message rate on skyaware. The number of aircraft is closer to reference system.This is compared with the dump1090-fa-airspy and reference system as reported in a previous post.

That is what I also found when comparing airspy_adsb with reference system for about a month. The Aircraft Reported graph for was about the same for both in differering weather conditions, so much so that the test system line was hidden by the reference system line.

airspy_adsb config file settings

GAIN= 21
OPTIONS= -v -f 1 -e 10 -w 3 -t 300
SAMPLE_RATE= 12
#network settings
NET= -l 47787:beast -c localhost:30004:beast
#Don't change:
G=-g
M=-m
1 Like

Futher to the above post,

here is a comparison of airspy mini sample rate 12 and 20 mhz. Less aircraft than previous post as it is approaching evening in UK.

signal-level-airspyadsb-12mhz-20mhz-12-mhzedit

and CPU use for 20mhz, 20mhz with packing and 12 mhz

12mhz-20mhz-CPU-use-edit

hope that may help interpret what I am seeing here.

Observations: I think that the number of aircraft on test system (airspy_adsb) gets close to the reference system. The message rate as reported by skyaware is closer to reference system compared with the dump1090-fa-airspy setup: perhaps 3/4 to 7/8 compared with the dump1090-fa-airspy which is about 2/3 or a biot less than the reference system.

The test system (airspy_adsb) seems to decode more messages than the reference system too.

Let me know if there is anything else I can provide using airspy_adsb before I revert to testing dunp1090-fa-airspy

This is the reference system LHS and test system (airspy_adsb) skyaware page sorted by RSSI, just as an example of data at the weaker end of signals.

Any chance  we can have a ''peek'' at your 5.0.5 dump1090-fa conf
not really much to compare  as yet  
many thanks 
john 

What is the ‘-t’ option for? I don’t see it in the examples in /etc/default/airspy_adsb

It is part of the whitelist (-w) option

the items in the whitelist will not be deleted until as in -t . I have a lot of trees around and setting a higher means that an aircraft is accepted as valid for longer as it drops in and out of the leaves and branches, it does not have to be reaquired and be subject to validation and entry in the whitelist.

Hope that makes sense.

1 Like

see post 312 for the config for both reference and dump1090-fa-airspy

OPTIONS= -v -f 1 -p  -x -e 10.5 -w 3 -t 300

Aha, thanks! I will have to add that to my config.

This is my config with ‘-t 300’ added.

/usr/local/bin/airspy_adsb -v -f 1 -w 4 -e 18 -x -t 300 -l 47787:beast -c localhost:30004:beast -g 20 -m 12
e 18 wow  i'd have lost that bet .... but see you are on samples 12 
````think im doing this < > thing wrong
`