Thanks. I guess I pay the price for dragging my feet, or maybe it was just one leg. Hope I was standing upright.
Thanks for the heads up on the sudo shutdown now. I did see that in Wiedehopf’s notes but there is nothing like a kind reminder. Wish I could slip a micro SD card behind my ear so I don’t have to buffer all the time.
Yup, I’d given that advice already a while back, but you are the master of your own destiny 
. It’s the first thing I install after the FA and FR24 software.
Then I go about on other ventures 
Got hold of the Vinnant COL1090/8-P today along with a FA Prostick Plus and Cable (25 ft) for 80 Euro’s. All brand new.
Bought it second hand in the area form someone that didn’t find time to set it up so decided to get rid of it.
So I’m gonna set it up on one of the existing 5.5 dB antennas to see how it compares to the COL1090/9-PSE 
1 Like
Jolly good show. Brand new on the cheap without shipping is pretty sweet.
2 Likes
Curious ehat the results will be. Thinking about the Vennant 9 dB as well. Just the length could be a „problem“
You have to be patient then 
first results will be checked by the end of the week. I want to compare it over the full working week and will let you know. So far it hasn’t been better then the 5.5 dB Vinnant COL1090/5-S over the weekend. That’s why I want to look a little bit longer 
Bring Patient…… well not my best forte.
Early comparison:
COL1090/8-P
COL1090/9-PSE
8P was installed on Saturday at 18:00 hrs
It looks like the 9-PSE is outperforming the 8-P.
But, 8P is located inside, 9PSE is located outside.
Distance between them is 1,5 meters.
9-PSE is on a fixed gain of 42.1, 8P is on a dynamic gain.
But the gain on the 8P is always 49.6 dB.
More conclusions later this week
Thanks for the details. Curious to see what else develops in your test bed there this week. I run the Max gain on my 8P which equates to the same 49.6 steady gain on the graphs 1090 tracking data. Wondering if that could be too much gain.
I’m going to run a test to determine the best gain setting. From The piaware perspective it seems to be the 49.6 value, but that is pretty much for all the Prostick Plus dongles. The test will show what is the best number for the number of positions and what value will be best for the number of aircraft. That’s something that I have to plan since it doesn’t make sense to do during low traffic hours.
I will look forward to seeing your test results. I seem to do well on positions most of the time but don’t seem to do as well with the aircraft numbers.
Well the results will be computed for my location so it might be a different value that might be working for you. I will share the test script that you can run in order to test for your location.I have to look it up, I don’t know where I’ve stored it.
I’ve been a bit quiet here while I’ve made further improvements. You’re right about the noise not being an issue. Nothing I did since has lowered it. In fact, my noise figure is a bit higher, but hasn’t had any negative impact. What’s changed since I started this post? My LNA popped, so I bought a higher quality one. Was using a cheap no-name brand one and replaced it with a Sawbird dual ADS-B and UAT LNA. That made a huge improvement in range, coverage and message rates. I then decided to go “pro” and bought a Radarbox antenna. Another big improvement. Since this topic started, I’ve gone from 280k messages to 420k per day, max range increased from 220nm to 278nm, average range from 190nm to 221 and coverage from 78% to 88%. All with an increase in noise going over the weak signals!
3 Likes
Step 1
First step is to install Python on your raspberry Pi
sudo apt install python
Step 2, create a blank file named optimize-gain.py
(2) Create a blank file:
sudo nano optimize-gain.py
Step 3
Copy and paste script 2 from this post into the blank file
Step 4
Make sure you copied all the lines
Save file Ctl+o and exit editor Ctl+x
Step 5 Make file executable.
sudo chmod +x optimize-gain.py
step 6 Run the script.
sudo python optimize-gain.py
Step 7
Go to kitchen and prepare a cup or two of coffee. You will need it during the long wait the script is running and outputs the final results. 
steps above are from @abcd567 , credit is towards him for this instruction.
This test wil run 5 times the same values, this will give you an insight in the total values for the different gain steps.
You will have one value with the highest postion count, maybe the same or another with the highest plane count.
It is then up to you to decide to go for the postions or the number of planes.
Set that gain number then as fixed in your setup.
Edit the dump1090-fa config file:
sudo nano /etc/default/dump1090-fa
Look for the following parameters
set ADAPTIVE_DYNAMIC_RANGE= from yes to no
set RECEIVER_GAIN=60 to the desired gain value ( use a point and not a comma when implementing decimal values)
Save the file
Restart dump1090-fa to have the setting take effect
sudo systemctl restart dump1090-fa
During the measurement your stats on graphs1090 will be messed up, this due to the fact that for every new gain setting the dump-1090-fa proces is restarted.
Note, this takes some time, average time on each pass is 15 minutes so it wil take you 1,5 hrs in total to complete the test. Be patient while the script goes through its moves.
In the end it wil display the weighted result of the 5 passes
So the results of my testing on the COL1090/8-P setup
| Gain | Messages total | Postions total | Aircraft total |
|---|---|---|---|
| 49,6 | 139069 | 9179 | 267 |
| 48 | 135014 | 8949 | 268 |
| 44,5 | 134996 | 8844 | 273 |
| 43,9 | 138880 | 9117 | 262 |
| 43,4 | 138886 | 9119 | 262 |
| 42,1 | 138336 | 9064 | 269 |
| 40,2 | 137209 | 9010 | 272 |
| 38,6 | 138115 | 9135 | 253 |
| 37,2 | 145152 | 9504 | 255 |
| 36,4 | 144016 | 9382 | 246 |
| 33,8 | 136907 | 9006 | 258 |
| 32,8 | 138954 | 9174 | 252 |
| 29,7 | 136063 | 9126 | 253 |
| 28 | 141160 | 9584 | 268 |
| 25,4 | 135749 | 9192 | 256 |
| 22,9 | 128711 | 8926 | 250 |
| 20,7 | 125657 | 8493 | 253 |
The highest postion count is on a gain setting of 28
The highest aircraft count is on a setting of 44,5
For me the aircraft count is the most important so the fixed gain will change from 49,6 to 44,5
So I give up some postions in order to get the most aircraft.
Run the test on your setup and determine which values are the best fit for your system.
Nice, different perspective here to setting desired gain. My chosen method is chasing 5% messages above -3ddB. That currently has me set to 10dB with my new setup. LNA has very high gain. How does your percentage of strong messages look like after this approach?
At 40.2 you only have one fewer aircraft than at 44.5. And really, the counts are close enough between 40.2 and 49.6 (range of only 10) to be almost meaningless. Also, there is only a ~10% difference in positions between 28.0 and 37.2
Given that, I would either run the script again, at other times of the day, and different days, or just go with 40.2 for a while and then increase by a step or two for a while to see what serves better overall. (i’d probably do both, honestly)
For me, I want my gain set at the lowest value that gets me good results. I don’t want to give up a bunch of count, but I don’t want my receiver deafened by a small plane or helicopter flying directly over my house, either.
Remember that the test runs for an hour & a half, so you are not tracking the same planes over the full duration. I am near PDX airport & I would bet that whatever setting was in use between 6:45 and 7:00 am on a weekday would show the best results, as flights arrived for the 8:00am bank of departures.
Just offering an alternate view of your stats.
1 Like
I’m having trouble interpreting the results/differences between tests especially Test1 and Test3. The numbers are significantly different, but the actual air traffic has not changed by much during the hour and a half tests.
Test 1 gain= 20.7 messages= 9805 positions= 1798 planes= 37
Test 2 gain= 20.7 messages= 8476 positions= 1529 planes= 34
Test 3 gain= 20.7 messages= 13667 positions= 2248 planes= 37
Test 4 gain= 20.7 messages= 6426 positions= 1144 planes= 29
Test5 gain= 20.7 messages= 7935 positions= 1504 planes= 29
The percentage has dropped from an average of 7.5 to 2.3 in the last 8 hrs.
But then again this was during the night so in the upcoming evening you will have a better view of the performance during the day.