I assumed that anyone with an Amateur Extra licence could search the forums on their own.
Well at least the 3rd link is one you don’t instantly find searching for “gain”.
And in my opinion it is a useful thread, so why wouldn’t i link it?
Anyway i’ll stick to receiving radio waves as i have no radio license
(My internet connection over one twisted pair might be radiating quite a bit of radio waves but it seems you don’t need a license for that)
I do have the external filter.
The ‘strong signals’ stat says 44% which from what I read sounds too high. The gain is set for -10 (AGC enabled) now. Maybe I need to insert a physical pad inline with the filter.
"samples_processed" : 709230592, "samples_dropped" : 10878976,
That’s about 1.5% dropped samples, I wonder if a beaglebone black can’t keep up.
AGC does not work correctly for ADS-B signals.
Just use a gain of 42 and try with that.
Interestingly MLAT still seems to working for you.
But optimally there should be no dropped samples.
Still those dropped samples wouldn’t be the reason for only close in planes not being shown.
That is absolutely right. Lower the gain in steps. 38-42 dB should be about right.
What do the rssi figures look like for the close aircraft?
You want them to peak around -3db.
Also, which filter do you have?
i am about 1 nm away from my airport. i also have flights that “grey out” because of overload on the dongle ( i am using an airspy mini…thanks to wiedehopf for the help). i agree that you will have to adjust the gain (this can take a while to experiment…use rssi as your guide, not the “gain scripts”). we are both very close to local airports so must accept that some overload will occur if we wish to receive distant aircraft as well. this is a “game” of balancing close traffic with maximizing distant targets.
The airspy overloads my RPI 3B+. It is there temporarily
I have to run it on an odroid XU4. The flash card died and I haven’t finished rebuilding it.
There is also a bigger Odroid N1 due out this week.
You may need to drop the gain another setting of two.
Also, native UAT978 support is due within a few weeks. You may want a bigger box if you want to run both(edit: sorry this only applies if you use an airspy).
Yep 40dB seems to be the magic number. Interesting. I’ll have to watch it tomorrow (when there is more traffic) but with a fedex Cessna 206 that just landed, I was able to see it all the way down to touchdown then the full length of the runway.
The only weird thing is that piaware-config wrote the new setting to /etc/piaware.conf and it did not take. I’m testing it by running dump1090-fa manually. I have to figure out why that is - the docs suggest maybe an override file in /boot (but there is not one). So something strange there…
You need to edit /etc/default/dump1090-fa
The piaware setting only works for the piaware sd-card image.
I think the BBB is a little noisy for ADS-B. I haven’t looked at it on a spectrum analyzer, but there is apparently a strong emission on 1 GHz (don’t know how wide). I switched to a RPi-1B.
MobaXterm will tell you the load on the BBB after you SSH in.
The radarcape has a BBB. Maybe they modify it to clean it up.
Wz2b has about 7 times the traffic of gliderp. He may need more than an RPI1B.
I didn’t intend to recommend a 1B. With my traffic, the 1B runs into the 60%-70% load regime, or less. I have a spare 3B, and 3B+, should 978 MHz ever come online and I want to run both frequencies concurrently.
I intended to say that the BBB, and mine is an A6, is noisy. It certainly is in UHF, and I’ve read it is at 1 GHz. JFFT.
That’s why Radarcape and FlightFeeder Blue come in metallic enclosures/boxes. Shielding.
Except the receiver is inside the shielding.
Antenna is what matters, in my experience.
The receiver’s 50 ohm input impedance is low enough that will squelch any EM captured at the input.
If clipping is an issue, it would seem expedient to keep RFI to a minimum to help the dynamic range to be as useful as possible.
Perhaps the shielding does mitigate the RFI, but I don’t have a Radarscape, nor access to a GHz class spectrum analyzer, to investigate. The Reddit poster (above) recommends shielded Ethernet cables, if you’re using that medium. I tried RotM Cat5, and the BBB seemed to radiate UHF RFI from the cable, and that’s after completely disabling HDMI - another violent source of VHF/UHF RFI on my unit.
Again, I don’t have a GHz class analyzer, but I’m not partial to the discounting of evidence, even when it is out of band.
My BBB doesn’t have a metal case. An RFI engineer working for AAI once told me, eons ago, that the only way to really control noise is to enclose electronics in a box with no openings. Perhaps, but I’m not an EE, and I got the gist of a Faraday cage from University physics.
My BBB is an early model, and may be worse than the newest redesign (Rev C). However, the Reddit poster provides evidence that correlates with my experience.
I am by no means an expert in credentials, experience, or education; but I am curious within the bounds of my means.
I have no interest in buying a $100k network analyzer to fully investigate, and assist in potentially mitigating, the condition I hypothesized here. JFFT