Play time has arrived

@tomvdhorst

Flightaware counts the aircraft itself only once per 24 hrs

Interesting - so it’s physical aircraft and not “flight numbers”. I had been presuming (not sure why) that when “United 123” landed and later the plane took off as “United 456” heading somewhere else, that would be “two counts”, but if it’s the same aircraft I suppose it makes sense to count as only one item.

That actually “helps” me to some extent, as due to geography I can see pretty much every departure from KSAN, but probably only about 1/2 of the arrivals. I was pondering how to try to pick up the rest of the arrivals, but if it tracks planes instead of flights, then I don’t need to - I see every plane eventually; they all have to take off sooner or later.

1 Like

It’s listed in the FAQ :slight_smile:

Why do the hourly aircraft counts on my ADS-B site’s statistics not sum to the daily aircraft total? SHARE | BACK TO TOP

An aircraft is counted only once per UTC day to calculate the daily aircraft totals. A given aircraft is counted only at most once per hour for the hourly totals.

1 Like

I have an original Chinese made Pi 1B - two USB ports and 256MB RAM

Boots from SD card but Raspian Buster OS is on an attached powered partitioned HD, there is no OS on the SD card.

It runs apache2 which serves my videos and music on my home network, videos viewed using KODI on devices attached to networked TVs. Friend in US able to view some home made video files via web server. Media kept on partition of attached HD.

Also runs a saned server so can scan from computers not attached.
SSH obviously
A desktop that does not boot by default (used rarely)
VNC server for when Desktop rarely runs - gave it a 512MB swap to make up for low onboard RAM.

Geoff

1 Like

@tomvdhorst - I got my ADSBExchange Blue stick, and replaced the RTL-SDR I was using. Immediately, the max distance I was seeing planes jumped from ~220nm to >300nm. I let it run for a day, and my Total Aircraft reported increased from ~3000 to ~3450 - all in all I cannot complain!

Since I have two Pi’s I think the next experiment will be to split the antenna signal and feed each one (different dongles) to compare “real time” the difference in detections.

At this point I haven’t done anything with trying to optimize the gain. I saw you had some good resources on that (Github), and I’ve bookmarked them to come back to. I’m hopeful that some tweaking there may improve detection rate even more - I need to get smarter on that.

I also need to figure some more objective way to measure performance of the dongles than “roughly how many planes I see before/after I switch”. I have also found your github on graph1090, which I’ll try to employ as well.
Are there other things to try? I’d like to try to create a “range plot” (I think I need VRS to do so). Lots more tinkering to come…

1 Like

It’s not my Github :slight_smile: that very valuable resource is made by @wiedehopf

The only tweaking I do is to check if I need to raise or lower the gain but most of my sites have the autogain option enabled so they correct themselves.
This is not always foolproof so a critical eye is advised over time.

I always use Grapsh1090 to compare the performance of my sites.
Like in this example I switched an RTL-SDR without the LNA towards a Flightaware Prostick Plus (with built in-LNA)

You can see the increase in range and the number of aircraft is also slightly up.
Installation is very easy for this piece of software, installation instructions are here :slight_smile:

Just copy and paste the commands in your Rapsberry Pi and the installation wil be done via a script.
after installation you can acces the screens via your browser ( read the last message when the installation is done).
Data will be flowing in after 10 minutes so be a little patient after installation.

As for range plotting, I don’t have anything for that installed, I just look daily at the individual stats provided by Flightradar24 :wink: they provide a range plot for the individual recievers.
If I need a combined overview I use Skyaware Anywhere, that combines the 10 sites all together in 1 picture.
I have a seperate Intel NUC computer running that has that webpage permanently open so all I have to do is to siwtch to that view to have a global overview of my recievers.

1 Like

Thanks will try that. I just found the FR24 stats (didn’t know there was such a page until I went poking for it), and I see that from RTL-SDR to ADSB-Exchange-Blue My max range went from 245nm to 333nm - so a pretty significant increase. Will start on the gain optimization today.

 

https://github.com/abcd567a/set-gain#readme

 

 

1 Like

I completed @abcd567 's install to include gain to be listed on my local skyaware page. I’m curious “what units is this reported in?” I’m showing “60” (presumably the default, since I haven’t started playing with it), but 60 “what’s”? I’d have thought it might be dB, but that seems like a pretty high value (60dB = 20 x 3dB doublings, so ~1,000,000x ??)

I was thinking of putting a coax splitter into the feed from my roof antenna, so I could do a side by side compare of my two sticks (on two RPi’s). Presumably that will hit my signal -3db (at least - assumes the TV coax splitter isn’t filtering at 1090MHz, which is well above the TV band, so it might be further degrading the signal). Can I simply up the gain by “3” (or something else) to try to account for this?

I’m also curious, looking at @abcd567 post here (31/51):

My 192.168.1.27/skyaware page now shows the current gain and allows for resetting, but I’m still looking at the map plot, not a graph. After following @wiedehopf’s Github to add Graph1090, I see plots similar to what @tomvdhorst is showing a couple posts up - not what abcd567 is showing in the link (which looks like a skyaware page but with a graph, not a map). Is there a skyaware page setting I’m overlooking?

Looking at @wiedehopf 's Github on automatically setting the gain
https://github.com/wiedehopf/adsb-scripts/wiki/Automatic-gain-optimization-for-readsb-and-dump1090-fa
The first line says:
only for rtl-sdr/DVB-T USB receiver, not Airspy or Beast receivers

I had an RTL-SDR in use, but replaced it with the ADSB_Exchange Blue dongle (which I presume is more or less equivalent to the FA Blue stick). What does that count as? I am not sure I know what a DVB-T receiver is. Googling I find another “dongle-y” thing at amazon, but it doesn’t look like the Pro Stick. I’ve seen Airspy as a brand, but not “Beast” (although I’ve seen references to that other places). Is there a glossary somewhere for the various terms encountered here?

And back to the other matter at hand, is my Blue stick “eligible” for this Auto gain program (perhaps it counts as one of the options)?

Thanks for any help!

Yes, 60 is 60dB.
In fact all values on the set-gain drop-down are in dB.

The dongle is actually designed as a TV receiver (DVB-T i.e. Digital Video Broadcast - Terrestrial). Its AGC is designed for steady and fairly stable TV signals. However the ADSB signals are very short pulses from a large number of aircraft, rapidly changing in amplitude. This makes dongle’s AGC to go crazy and it locks itself at around 58 to 60 dB. Setting dongle to 60 is same as setting it to AGC or -10

The actual setting of the dongle is shown in red above the drop-down & set-gain button as “Current Gain”.

image

 

1 Like

I wouldn’t use a TV splitter in order to devide the signal between the two recievers.
Looking here there’s enough choice to get a decent splitter based on SMA. Just mind to get the correct connectors.
https://www.amazon.com/sma-splitter/s?k=sma+splitter

As for the remark that is stated on the script from Wiedehopf, almost all dongles are based on the same chip and recognized in the software as a rtl-sdr reciever.
It uses the same name for a flightaware dongle, radarbox dongle and a rtl-sdr dongle.
Same goes for your ADSB dongle, it will be recognized as an rtl-sdr dongle and that will make it perfectly fine for the autogain options ( either the method of abcd567 or the automated gain adaptation from dump1090-fa itself).

Beast is a different protocol to report ADS-B signals and that is a different kind of reciever.
Examples are an Airsquitter or a Radarcape reciever

1 Like

The autogain doesn’t work for dump1090-fa since that changed the config logic.
Well it still works on piaware images if you run one of those.

The autogain logic is not smart, it targets the 5% as described elsewhere looking at the previous day and making a one step adjustment.
It’s described pretty clearly in the readme.
It’s mostly for people who don’t want to deal with it.

1 Like

Thanks to all here. My thought with using the splitter was to get a more accurate comparison between the RTL-SDRv3 and the ADSB stick. Somewhat anecdotally, I can see a “big difference” between them just by hopping back and forth, and it’s probably easier to just run each one for a whole day now that I have graphs1090 up and working.

Yes, you are right. Much easier and simpler to run each one for one day without splitter.

Following is just for browsing & information, and NOT a recommendation to use Wilkinson splitter, though these are available very cheap on Ebay.com and Aliexpress.com

https://forum.planefinder.net/threads/wilkinson-splitter-combiner.1377/

 

 

I have two Pi’s running now - one connected to my “good antenna”, and the other to the little stick that came with my ADSB dongle; the ADSB antenna (indoors) is seeing about 20 planes at a time while the good antenna (on the roof) is seeing 200. The data reported on my SkyAware page at FlightAware is kind of messed up; it won’t give me a “distance” column, since it thinks I have two sites (doesn’t realize they are colocated), and it can’t compute a distance.

I want to continue to monkey with the second site, so I want to be able to look at 192.168.1.26/skyaware and see what it’s detecting inside my LAN, I just don’t want FA to be paying attention to it. I do expect at some point I may actually use it - so I don’t want to kill/delete the second site; I just want to pause it being received/used by FlightAware. Is this possible - can I ask FA to “ignore” one of my sites, but not actually delete it (presuming I’ll turn off the “ignore” someday)? I’m guessing even if I stop the feed from the house to FA (not sure how to do that, but presume it’s possible), that FA will still think I have 2 sites feeding it, and will not give distance on SkyAware.

sudo systemctl disable --now piaware

No, you’re the best of us :rofl:

1 Like

Ah glad to know, thanks for the confirmation :rofl: :joy: :smiley:

1 Like

And not even going double dutch :rofl:

1 Like

You do realise you probably have more Raspberry Pi’s on your shelf than most of the suppliers have left in stock! :rofl:

Don’t bother trading the stock market or padding your investment portfolio with Bitcoin…just sell some Pi’s and cash in on the ridiculous price they are going for at the moment! :wink: :rofl:

-=Glyn=-

1 Like

@wiedehopf that seems to work to stop sending data to FA, while still generating it, so I can see the effects of adjusting things (like gain) by looking at 192.168.1.26/skyaware on my local network. However it looks like FA is still listening for that site; if I go to https://flightaware/skyaware - from outside my home network, FA won’t compute a “distance” for any aircraft, even though all the data is coming from only one site:

From poking around elsewhere in the forum, it seems the issue here is that “generically” someone with multiple sites could have data coming from multiple physically different locations and as such, computing a distance becomes less meaningful; “distance from where?”

I accept this is a bit of a small issue - if I’m at work and I want to take a peek at what my home setup is receiving, it’s not the end of the world if I can’t get a distance measure (I can’t get a graphs1090 output either). I was just looking to see if there were a way to effectively hide the site so FA stopped listening for it.
I guess I could actually delete it, create a new one, and then leave it unclaimed, but that seems like it’s impolite to the FA structure - adding extraneous sites that aren’t really independent/new.