Optimize Gain - which would you chose

FlightAware ADS-B Flight Tracking
1

So trying to optimize with my new set up. Which would you chose. Data seems to be all over the place? Im thinking 37.2 . but not sure.

pi@piaware:~$ sudo ./optimize-gain.py
test 1 of 5
gain= 49.6 messages= 8142 positions= 958 planes= 40
gain= 48.0 messages= 7694 positions= 1006 planes= 39
gain= 44.5 messages= 7717 positions= 1094 planes= 35
gain= 43.9 messages= 8051 positions= 1140 planes= 38
gain= 43.4 messages= 8226 positions= 1261 planes= 42
gain= 42.1 messages= 7808 positions= 1268 planes= 41
gain= 40.2 messages= 8344 positions= 1275 planes= 37
gain= 38.6 messages= 8625 positions= 1426 planes= 41
gain= 37.2 messages= 9462 positions= 1502 planes= 42
gain= 36.4 messages= 9559 positions= 1597 planes= 38
gain= 33.8 messages= 9148 positions= 1630 planes= 42
gain= 32.8 messages= 9694 positions= 1635 planes= 40
gain= 29.7 messages= 10399 positions= 1845 planes= 42
gain= 28.0 messages= 9807 positions= 1639 planes= 38
gain= 25.4 messages= 9589 positions= 1572 planes= 38
gain= 22.9 messages= 8963 positions= 1451 planes= 33
gain= 20.7 messages= 7941 positions= 1273 planes= 30
test 2 of 5
gain= 49.6 messages= 11834 positions= 1904 planes= 45
gain= 48.0 messages= 11689 positions= 1894 planes= 45
gain= 44.5 messages= 11440 positions= 1832 planes= 46
gain= 43.9 messages= 11361 positions= 1757 planes= 45
gain= 43.4 messages= 10267 positions= 1623 planes= 45
gain= 42.1 messages= 10138 positions= 1514 planes= 47
gain= 40.2 messages= 8744 positions= 1355 planes= 42
gain= 38.6 messages= 8357 positions= 1300 planes= 41
gain= 37.2 messages= 7992 positions= 1189 planes= 41
gain= 36.4 messages= 6607 positions= 1026 planes= 38
gain= 33.8 messages= 6364 positions= 1039 planes= 37
gain= 32.8 messages= 5769 positions= 936 planes= 39
gain= 29.7 messages= 4606 positions= 773 planes= 37
gain= 28.0 messages= 4047 positions= 600 planes= 36
gain= 25.4 messages= 3402 positions= 502 planes= 25
gain= 22.9 messages= 3537 positions= 571 planes= 25
gain= 20.7 messages= 3063 positions= 572 planes= 19
test 3 of 5
gain= 49.6 messages= 6762 positions= 1169 planes= 43
gain= 48.0 messages= 7108 positions= 1182 planes= 44
gain= 44.5 messages= 7235 positions= 1249 planes= 44
gain= 43.9 messages= 7768 positions= 1421 planes= 42
gain= 43.4 messages= 7738 positions= 1410 planes= 43
gain= 42.1 messages= 8433 positions= 1415 planes= 47
gain= 40.2 messages= 8364 positions= 1422 planes= 48
gain= 38.6 messages= 8235 positions= 1442 planes= 45
gain= 37.2 messages= 8387 positions= 1450 planes= 41
gain= 36.4 messages= 8155 positions= 1326 planes= 42
gain= 33.8 messages= 7722 positions= 1192 planes= 38
gain= 32.8 messages= 6952 positions= 1095 planes= 37
gain= 29.7 messages= 6674 positions= 1076 planes= 35
gain= 28.0 messages= 6271 positions= 1072 planes= 30
gain= 25.4 messages= 5033 positions= 829 planes= 26
gain= 22.9 messages= 4042 positions= 618 planes= 27
gain= 20.7 messages= 3536 positions= 532 planes= 22
test 4 of 5
gain= 49.6 messages= 6677 positions= 1112 planes= 47
gain= 48.0 messages= 7798 positions= 1285 planes= 46
gain= 44.5 messages= 7177 positions= 1233 planes= 46
gain= 43.9 messages= 7453 positions= 1192 planes= 45
gain= 43.4 messages= 7583 positions= 1094 planes= 48
gain= 42.1 messages= 8387 positions= 1201 planes= 46
gain= 40.2 messages= 8610 positions= 1177 planes= 47
gain= 38.6 messages= 8425 positions= 1153 planes= 48
gain= 37.2 messages= 7161 positions= 913 planes= 48
gain= 36.4 messages= 6887 positions= 1041 planes= 45
gain= 33.8 messages= 7748 positions= 1070 planes= 47
gain= 32.8 messages= 6847 positions= 887 planes= 46
gain= 29.7 messages= 8241 positions= 1019 planes= 42
gain= 28.0 messages= 7453 positions= 876 planes= 37
gain= 25.4 messages= 7072 positions= 858 planes= 32
gain= 22.9 messages= 6775 positions= 701 planes= 31
gain= 20.7 messages= 5966 positions= 540 planes= 28
test 5 of 5
gain= 49.6 messages= 12806 positions= 1252 planes= 59
gain= 48.0 messages= 13600 positions= 1482 planes= 56
gain= 44.5 messages= 13759 positions= 1667 planes= 58
gain= 43.9 messages= 15372 positions= 1742 planes= 55
gain= 43.4 messages= 15441 positions= 1766 planes= 57
gain= 42.1 messages= 15318 positions= 1758 planes= 61
gain= 40.2 messages= 15076 positions= 1848 planes= 61
gain= 38.6 messages= 15397 positions= 1942 planes= 56
gain= 37.2 messages= 15002 positions= 2046 planes= 63
gain= 36.4 messages= 16245 positions= 2171 planes= 55
gain= 33.8 messages= 16532 positions= 2122 planes= 59
gain= 32.8 messages= 15877 positions= 1992 planes= 59
gain= 29.7 messages= 15565 positions= 1975 planes= 60
gain= 28.0 messages= 14770 positions= 1965 planes= 58
gain= 25.4 messages= 13540 positions= 1699 planes= 46
gain= 22.9 messages= 12254 positions= 1624 planes= 45
gain= 20.7 messages= 9289 positions= 1307 planes= 46

===Totals===
Gain, Messages, Positions, Aircraft
49.6 46221 6395 189
48.0 47889 6849 184
44.5 47328 7075 186
43.9 50005 7252 186
43.4 49255 7154 189
42.1 50084 7156 197
40.2 49138 7077 193
38.6 49039 7263 185
37.2 48004 7100 188
36.4 47453 7161 171
33.8 47514 7053 176
32.8 45139 6545 169
29.7 45485 6688 168
28.0 42348 6152 163
25.4 38636 5460 140
22.9 35571 4965 138
20.7 29795 4224 123
pi@piaware:~$

2

Yes, that’s a good starting place, but in addition to that, I would look at the noise floor changes for the gain settings you select.

3

I’m unknowledged in that.

4

Looking at the total values, 42.1 has the highest number of planes. You could One or two values close to this and run the script again, to validate the result.

5

Quick related question: please could someone direct me to previous postings for guidance on optimizing gain settings using a FA Prostick Plus and RPi3 i.e. how to identify the optimum gain setting and how to make the change to the setting

Many thanks

7
8

I did the Optimize Gain Biekerc. Thats how I got the graphs above. But its a disappointment that Im receiving about the same planes as before. Different gain settings than my old set up.

9

That was a reply to NigelRichardson, who wanted to find out how to do this.

Generally, the number of planes depends on the flight patterns traffic One message from a plane is enought to count as one plane. Even if you increase maximum distance or message rate, in the end you might still see the same number of unique planes.

10

Sorry :slight_smile: Biekerc, Thats the thing about this new board you cant really tell whom is replying to whom unless they quote it, which I still have trouble doing.

I upgraded because I used to be top in the Puget sound. A gent about a mile from me is now just slaughtering me in planes and position counts. planes by a couple hundred at times and 90,000 positions. I thought I would go to the flight aware antenna instead of my home made COCO as well as the Prostick + instead of the Prostick. Both receivers have the Flight Aware filter too. Since it was a Birthday gift I ordered the PI3 just to upgrade of of my old B+. My Counts are pretty much the same, just a different gain setting. Kinda Disappointed actually.

I dont care IM in 2nd or 3rd, but I thought I was missing data to feed to flightaware and wanted to optimize the site to improve data. I wrote the gent to see what he is running to find out what I could do better :). Im wondering if I should try it without the filter to see if it improves.

12

fwiw, just noticed, lowering gain too much lowers the number of MLAT receivers synchronized.
i don’t know how/if that impacts MLAT positions/aircraft received as yet…

Evangel

13

Quoting is easy, just highlight a text passage, click on the quote-symbol that appears, and then a message windows opens with the quote in it.

Just looked at your stats, compared to most of the sites around you your results pretty high. The gent I think you are refering to, has two sites, a new one close to you and one further away.

Right now, the site nearer to you has just around 30 planes more, which is not a lot, but already 100.000 message more.

Maybe another examples of better reception, more messages, but not necessary many more planes.
Looking at the coverage graph, it has signals further away and fewer gaps in one direction.

I would guess it is located somewhere higher up and therefore unobstructed “view”. Otherwise I do not think you are so much apart, be careful not compare the values of both sites combined with yours.

I do not think changing the RPi makes a difference at all, I think David Baker wrote somewhere recently it is mostly amplifier and filter, then antenna, that make a difference.

14

Makes sense. Less gain, less data received, less data available and/or valid for MLAT calculations, less common planes seen by nearby receivers, therefore no sync at all, or sync with fewer receivers.

15

Hi all
Been using optimize-gain.py for few weeks now on n off
but after getting correct figure to use and set it couple days later figures have all changed again as if it is constantly changing the gain setting
is this correct ?

john

16

You should not use that gain script anymore after finding the optimal values.
Even if that “optimal” might look like it changes, based on your local and far-away traffic.

17

cool thanks SoNic67

john

18

I lived in a location where i can barely get aircraft over 120nm and my antennas where 10 metres high since changing location my antennas are actually nearly 9 metres high and the minimum i get is around 240nm and furthest i got so far has been 515nm even when i use to go adsb mobile some locations could get over 100nm but there was one location that i use to get over 300nm with a 5db antenna…

Regards Lino…

19

@linofalzon

Range mainly depends upon terrain arround antenna location which in turn depends on geographical location. Incresing antenna height also increses range, if antenna is surrounded by tall buildings/trees/hills. Once antenna is above these obstructions and can “see” horizon in all or most directions, any further increase in height brings very little improvement.

You can find your maximum reach at each of different locations (which you have mentioned in your post above) by method described in the FIRST POST of this thread:

What is the Maximum Range I can Get?

.

20

ok, i need some opinions. i have just installed a minicircuits splitter and a second RPI3B and am running a second site in parallel, the antenna is common as are all cables. my hope was to optimize the gain by using a single target and looking at two data sets. (i have a flightaware antenna and a rtlsdr 1090 lna at the antenna (fixed gain of 27dB). this is mounted on the roof. i then have 105 feet of LMR400 (running thru a lightening arrestor (outside) and a minicircuits bias-t (inside the house). the output of the bias-t is feeding the splitter. there are two RTL-SDR v3’s. i estimate about 9dB of loss from cable,lightening arrestor, bias t, splitter etc. the spec for the spliter is -3.6dB, insertion loss and split loss per side)
the “reference” side gain is set to -10. the “experiment” side gain is currently set to 42.1 AND has a 3dB pad between the splitter and dongle. (so, the experiment side has a gain of 13.9dB less than the reference side)

when i get reference side RSSI values of -3dB or less then i see approximately 13.9dB less in the experimental RSSI values than the ref side…so far so good. now (finally) here is my issue and where i need opinions. when the reference side RSSI is, for example -1.9dB i see experimental side RSSI values of -2.3dB. earlier today i had a plane come by at about 2000 feet and only 1/4 mile away (the engines were easily heard). i saw a reference side RSSI of -1.5dB and an experimental side RSSI of -1.6dB
is my antenna over-driving the LNA?
opinions or alternate explanations are very welcomed

21

I’m running a homemade spider antenna with the rdl-sdr LNA and the rtl v3 dongle.

Gain set at 20.7 so i don’t get overloaded by airplanes flying close by.
Not much cable though. On the quiet end I get messages with RSSI around -34 dBFS.
When i increase the gain to 22.9 the faintest signals i get are around -32 dBFS.
So i figured i’m still receiving mostly everything far out but very rarely get overloaded when small planes come buzzing by close. Back to your setup though.

Setting at -10 means you don’t know what your current gain is as the receiver can switch it around as it chooses. So hard to do comparisons.

Just do the experiments with a gain of 28 on one station and 38 on the other station so you don’t get that close to saturation. Maybe even lower.
If the signal peaks are overloading the dongle the RSSI reading is gonna be off.
I would only compare readings below -3dBFS. Otherwise the receiver with the higher gain may already be losing the stronger messages so the reading will be same while the signal is actually stronger. (You are getting quite a few messages from each plane, some messages will be stronger than others)

You should also really compare the number of messages received from the same target on the different receivers. It will show you much more than the RSSI number, that’s a rough reading anyway and can fluctuate quite a bit.

22

very true. can you post some comparison charts ? :grinning: