Long term testing?
I tried it for several weeks. This setting gave me the best balance between range and number of reported messages per second.
There is a thread named “Thoughts on optimizing gain”. Tons of comments to read on a long winter evening. Check it out 
Could you please help me understand how a Message that was received, regardless of the signal strength, can be interpreted as “noise”.
Is it a message or is it “noise”
Thanks,
S
Already have, it’s what leads me to now wonder where the numbers come from since we all have different results and I’m not seeing the consistency that would give me statistical confidence in the numbers -3dBFS and 5%. It’s that that I’m keen to see if it exists anywhere.
In that thread I posted that I had 31%. I’ve since adjusted the setup and dropped the gain more and now have around 15%. I can’t say that I see any statistical significance in the changes. That’s good in one way – I’d prefer to use the lowest gain needed to extract the most value from the site, so if it’s not got worse then I’ll stick with 15% for now. But it’s all very ‘finger in the air’.
Anyway I don’t want to turn this into a spin-off of that thread, I was just curious to see any kind of mathematical source for these two figures, perhaps if it applies more generally to RF digital comms.
That’s not my interpretation but the result of several threads here.
The fact is if this number went higher, the number of usable positions/second goes down(at least in my environment)
I’ve since adjusted the setup and dropped the gain more and now have around 15%. I can’t say that I see any statistical significance in the changes.
For me i can say that the changes are clearly noticeable. Every setup is indivual, there is no “golden rule” on it. Try it and use the best for yourself.
For me they appear statistically insignificant hence my desire for objective info to support it. As I say I’m certainly not disputing what people see, but similarly if it’s about what people see then is this long-term testing and data by Mode S users which has been properly analysed and that has resulted in these numbers?
I can see that there would be best practices and ideal targets representing optimum installations. Of course every site is individual but that’s exactly where my query comes from – we can’t simultaneously have “You should target for a value of around 5% of all messages [being > -3dBFS]” and “There is no golden rule”.
Is there any objective mathematical info which supports these numbers, even just limited to the context of Mode S, or is it all 100% received wisdom and ‘what most people reckon from having a play around’?
of course you can. Set the 5% as target and check if it fits. I’m not saying that you have to stick to the 5%. That’s misinterpretation of my comment. “You should” is not “You must”
You should ask your second question in the “Thoughts” Thread. I am not the one brought this up, i am only repeating because for me it’s working
Note that the OP of this thread was asking about increasing Range with no mention of usable positions/second.
I too am interested in maximizing number of planes and maximum distance. Messages per second doesn’t matter to me.
I started off with two identical receiving systems. One I kept as a reference and the other is used for the experiments to understand the effects on aircraft count and range of any changes to antenna, feeder, amplifier, filter, receiver and gain and other settings. I can see actual changes in real time.
At my optimum range configuration at this location, “Messages > -3dBFS” is between about 35% if it is a nice day and all the trainers are out doing circuits nearby down to about 10% when it is bad weather or the bush fires are really bad with visibility at about 1km and not much more than the scheduled flights are about.
Every user is different and every site and location is different and quoting something as if it was a rule doesn’t progress the discussion.
YMMV
S.
It’s a rough rule of thumb, I never intended that value to be anything more than a “hey maybe you’re overloading things” indication.
-3dBFS is something of a magic constant because it means 50% of max power. Mode S messages effectively have a 50% duty cycle, so in a hypothetical perfect noise-free case, the strongest signal you could receive without overloading the ADC would have 50% of max power. After you add in some real-world noise to the nominally “off” half-bit periods, it’ll be somewhat above 50%. So a message with a signal strength >-3dBFS is probably saturating the ADC.
The demodulator is fairly tolerant of this, but it does eventually reach a point where very strong messages are undecodable.
Here’s a couple of graphs from some testing I did recently that show the effect. (Note that transmit levels are on an arbitrary scale, they don’t correspond to any particular absolute power level)
Things to note:
Since you can’t measure by how much a signal is overloading the receiver (and you can’t measure messages that you fail to decode), assuming that there’s some sort of distribution of signal levels and measuring how many fall into the “strong enough to be in the danger area, but not so strong that they couldn’t be received” can be used as a sort of proxy for measuring how many messages you’re losing to overload.
The actual shape of the distribution of signal levels, and your tolerance for losing strong/nearby signals vs. losing distant signals, is going to vary from site to site.
I’m likely out of my depth here…
Could there be a case for a dual receiver design (effectively two SDRs running off a single antenna) with each one’s gain set to cover strong and weak signals respectively. Would probably need USB3 to cover the extra bandwidth requirements of dual SDRs if the device were in a single dongle form.
Of course there would then be the issue of getting Dump1090 working with them.
You’re better off just using a SDR with better dynamic range, I suspect.
Very true! I imagine its a balancing act between making cost effective devices to encourage more receiving locations and that of satisfying the enthusiast who wants to get the best possible out of their location.
You can just use two receivers.
It requires quite a bit configuration to make it work and you need to feed as two different stations if you want FA MLAT.
The secondary receiver can have a very simple antenna as the signal strength isn’t that important.
Or you just get an Airspy Mini. But i’m well aware using two receiver is cheaper, especially if you already have an existing setup.
I was not replying to the opening question of this thread.
As i said in one of my replies, it depends on the environment. The 5% is a starting point for further analysis. If 35% is working for you, hey that’s fine.
I had 2 independent receivers running on single RPi model 2 B for almost an year. Since last 2 months these 2 stations are Orange Pi PC.
Below is the link to automated installation script which performs all installation & configuration automatically. The only manual part is to
Just did, good info thanks!!! I’ll watch for a while and see how it performs, I’ll also look at maybe adjusting the gain to possibly to improve.
Nice one thanks, that underlying analysis is what I was looking for. You can see where I dropped my gain at the start of the month.
If it’s too high the signal levels are very tight, and if it’s too low the upper level drops down. I’d instinctively want to find a gain which lets me see signals down around -25 dBFS but any lower and the upper level starts to compromise. Or to put it another way, I want to expand that graph to maximise its width to cover around -25 to -3 ish, either side of which the trade-offs start to steepen.
This is analagous to the squelch on a radio which needs to be responsive enough to the lower signal bound but no more than is needed.
Based on that I have a little more room to explore on mine, it looks like I can drop the gain a touch without compromising the upper level. I’ll also try it on max gain with no preamp since I remember what that looked like; it will be interesting to see if that naturally aligns with the -25 to -3 ish range.
Just pursuing the practicality (not the logic) of your suggestion -
Not only do two antennas allow the receivers to not operate at the extremes of their gain range, but their radiation patters are much more suited to each purpose.
@geckoVN I was thinking more along the lines of splitting the signal after a built in LNA makes up for the loss. Maybe not optimal - I suppose it depends on how much noise the LNA would introduce. An analogy would be a TV distribution amplifier restoring the signal level to each receiver. I’m no RF engineer so my assumptions may be invalid 
I agree, it would be better to have two aerials one of which would have less gain and run less risk of overloading.
That’s a great idea and kudos for scripting the install too! The setup is just as @wiedehopf mentioned too and the flexibility to pick and choose aerial types to suit the location’s needs gives some flexibility.
Before I realised geography and not the aerial siting on the property was my biggest issue, I was wondering if a dual receiver was possible and could stick an aerial to a north facing window. Having moved the setup entirely to the north window and received nothing at all, I didn’t pursue it any further.
Thanks for sharing the script!
