Some parts of the day I see that but then the delta drops to 0%. Probably it depends of how many planes are close by and how many are far away.
Maybe I need more gain, because even at 12MHz I am at gain 20-21 and effectively I can’t go higher. I already have a LNA with 27dB, followed by a very long RG6 cable (45m). Maybe I can move the antenna closer somewhere, but that will lose some height.
If i remember correctly, you always choose far higher gain that i would anyhow 
Also the different sample rates are only gonna perform differently if you have either lots of overlapping messages or lots of messages that are barely above the noise.
If you have neither, the sample rates are probably gonna be almost the same.
There are not so many signals right now, maybe 1/4 of the heydays. I am hitting max 300 msg/s now and I was at about 1,100-1,200 in December last year.
I have no means to accurately measure the level with Airspy. They claim a maximum RF input of +10 dBm, so I would think there is plenty of headroom. This is what your graphs are reporting.
That’s indeed not that much gain 
I just remembered when you had all the signals between 0 dBFS and -10 dBFS!
Anyhow not everyone sees benefits with 24 MSPS.
I thought of trimming gain a bit when lowering the sample rate as I’m effectively tuned to the higher rate by default as that’s where I normally run, but truthfully, I’ve never noticed the Airspy to be very reliant on gain at all - at least not like the rtl radios. I’ll need to muck around a bit more to find the most comparable setting for each sample rate like I did with -e and do another round of tests I guess. I suppose there may actually be a linear(?) scale between sample rate and gain, but I wouldn’t know where to begin to calculate that.
When I tested yesterday, there were between 100-120 aircraft in my range at any given time. It would be interesting to replicate those tests during peak, medium and slow hours to get a better picture of the effects. As with testing anything in this game, the dynamic nature of data makes it very difficult to maintain a valid control to compare against - even when testing side by side, but it’s the best we have.
I’m definitely open for ideas on how to improve the results/validity to these tests.
Could you post graphs of the signal for the test?
Also i’m curious why you’re not getting matching figures compared to the reference receiver?
As in i’d expect one of the settings to be 100% ?
Sure - but to enable me to do so, I’ll need to test for longer strings of time the next go-around.
As stated above, the setups as they sit are not close to identical. My current “test” setup averages 93.8% (@24MSPS) of my main setup so far as positions are concerned (aircraft counts about the same difference, but I count positions due to rule of large numbers). This is mainly because I currently have a splitter upstream of the radios spouting into the Airspy R2 for 1090Mhz and the other side going into an RTL radio for 978Mhz - so the ~3db hit is apparent and the main difference between the two besides antenna.
For my tests above, when I dropped to 20MSPS, that 93.8% baseline dropped to 85% and at 12MSPS, 85.2%.
When I equalize my setups (again) for the next antenna shootout I have planned for when it cools down, I’ll take the time to re-run some of these tests again on otherwise dead equal setups for some better data.
For now, signal graph on main:
vs. test:
EDIT/ADD: Test rig after knocking gain down a couple notches where it should be. Amazing how the dynamic range shows itself:
Can see where I’ seeing more dynamic range on my main setup without the splitter in the way (and better antenna). I should actually drop the gain a bit more on the test rig now that I look at them side by side… Can see where I hit the peak level 0.0 yesterday on the test rig during my tests.
We’ve veered way off course on this thread…Apologies
Hey… new here. I also just setup this same antenna today, also to replace a dumb little spider antenna I put together a couple weeks ago. My dongle is from Amazon:
“RTL-SDR Blog V3 R820T2 RTL2832U 1PPM TCXO HF Bias Tee SMA Software Defined Radio with Dipole Antenna Kit”
https://www.amazon.com/gp/product/B011HVUEME/ref=ppx_yo_dt_b_asin_title_o04_s02?ie=UTF8&psc=1
I first started feeding only to FlilghtRader24, but when I got the antenna, I realized hey there’s another provider, so let’s feed to them too (so here I am). After setting up piAware and stuff, and then modifying the FR24 config so it would run too, I started watching my stats to look for an improvement.
I haven’t been up long enough on FA so I’ll be quoting FR24 statistics. On the Spider anetnna, over the period of about ten days, I got a best of 136nm. This was with the antenna at the top of a 16’ tall wooden pole, using two 25’ lengths of RG6 TV cable (not new either) running to an adapter to an N-cable (the one that came with the dongle), to the dongle.
When I installed the new antenna this morning, I put it on the same pole, in place of the spider antenna, using the same cable. I figured let’s leave everything else alone so I’m really testing JUST the antenna.
That being said, over the last couple of hours I’m already seeing a max range of 153nm and I expect it will go up more over the next 24 hours.
I’ll be out of town for the weekend, so next week I’m going to buy some fresh RG cable and see about getting the antenna up higher.
You Sir, have been bitten by the bug!
Make sure you get cable with reasonable specs for microwave (GHz) frequency.
“RG6” covers a lot of ground, from very poor to very good. 50’ of satellite grade RG6 wouldn’t be the worst you could use.
The RTL-SDR blog V3 is an excellent general purpose receiver, but can be significantly improved for ADS-B with a filter+amp.
(my best performing site runs this combination)
And there is more chance of finding a vaccine for Covid-19 than the bug!
Now you just need a filtered LNA
RTL-SDR Blog ADS-B Triple Filtered LNA (Bias Tee Powered)
And some graphs of course: GitHub - wiedehopf/graphs1090: Graphs for readsb / dump1090-fa / dump1090 (based on dump1090-tools by mutability)
Thanks; I just installed the graphs. For the filter, do I just add it inline?
Yes, but the device linked needs power injected via bias-t as it’s not only a filter.
In your case it’s sufficient to activate the rtl-sdr v3 builtin biast (once the LNA is connected in line):
https://www.rtl-sdr.com/getting-the-v3-bias-tee-to-activate-on-piaware-ads-b-images/
It happens to us all eventually 
Ok, it’s been 2 weeks since I changed the gain and installed the graphs.
Errors are 4.2%
Suggestions???
I have no idea if these mean anything??
Maybe not call it errors when it nowhere says those are errors.
All looks reasonable, settings are fine.
If this is at default gain, your antenna / coax setup is probably a bit weak / coax has quite a bit of loss.
These are not “errors”, but messages > 3dBFS.
Can be real messages, but also some “noise” which cannot be processed by your device. That’s why you shoult target to below 5% which you did already.
In a very frequent are like mine, smaller changes in gain does not end up in differencies.
Currently i am running gain 42.1, but going up to 49.6 is showing only a small change.
Only a change to -10 is heavily overloading my setup and the ratio is up to 25%
And with everything below 39.6 the range is slightly decreasing.
What i am trying to say: It’s a try and error to find the proper gain settings.
I set the gain at 42.1 as advise over stock settings
I have a Proxicast 10 ft Low-Loss Coax Extension Cable (50 Ohm) with the FA antenna.
Is there a better cable that I should purchase?
What is the give away that he cable is not good?
Thanks
With that gain setting the results are what i would expect and look good.
Note that i said “If this is at default gain”
