Side by side antenna testing

[TomW]

I have completed the first portion of my first side-by-side antenna tests.
First, the gear: I purchased two complete sets of gear, in order to make two identical trains. These consist of RPi 2s, power inserters, diplexers, amplifiers etc (details in photos below). The software is identical (except for ip address and hostname). The RPis each reboot once a day at the same time via crontab. The antenna masts remain in my home office (approximately 20 feet above grade), but still inside the house. Their locations differ only in about 24 inches in longitude, and throughout the testing will be called "West" and "East", and in all screen captures, the systems will display as if looking North (West on Left, East on Right). The systems are completely interchangeable, allowing for example, swapping one amp for the other to see if differences are contributing to results. Each test will be two runs of at least 3 days, the second of which will be with the antennas switched between East and West.
My location is approximately 23 NM northeast of DFW and 18.4 NM north-northeast of DAL.
I set up a number of tests and let the systems run unmolested for 3 full calendar days. The test indicators are shown in the attached photo (from Flight Aware, Dump-Mut, and Plane Finder).

So Test #1, run 1: West = 8 leg spider, S0-239 whip, legs measuring 69mm from teflon in So-239. East = 8 leg spider with N whip, legs measuring 69mm below the copper disk.
And the winner is East
Antennas have been swapped and the second half of Test #1 will begin tomorrow.
Comments and suggestions will be appreciated
Tom

 

[trigger]

Nice work Tom. Looking forward to the next results.

 

[bramj]

Very nice setup Tom,

 

[ab cd]

Very scientific approach Tom!
With this type of comparison setup, one can be very confident of the results.
Well done, congratulations.

By the way, what is the outer diameter of copper disc? I just wanted to compre this copper disc with Cantenna's bottom (Pepsi Can's bottom has a dia of 66mm).

 

[TomW]

Very scientific approach Tom!
With this type of comparison setup, one can be very confident of the results.
Well done, congratulations.

By the way, what is the outer diameter of copper disc? I just wanted to compre this copper disc with Cantenna's bottom (Pepsi Can's bottom has a dia of 66mm).

The copper disks are 2.5 inches diameter. I just received another batch of 5, because I see more tests on the horizon...

 

[ab cd]

The copper disks are 2.5 inches diameter. I just received another batch of 5, because I see more tests on the horizon...

Thanks.

Your copper disc is almost same dia as a Cantenna made from a standard Pepsi/Coke/Beer can, whose bottom has a diameter of 2-5/8 inches (2.6 inches / 67 mm).

 

[ab cd]

I am just wondering what will be the effect of increasing/decreasing the disc dia, keeping the radials same length, i.e. 69mm from the point these are bent down.

 

[TomW]

I am just wondering what will be the effect of increasing/decreasing the disc dia, keeping the radials same length, i.e. 69mm from the point these are bent down.

I'll be happy to test that...some day.... I must keep to the schedule and plan......

 

[ab cd]

I'll be happy to test that...some day.... I must keep to the schedule and plan......

Thanks. I will wait for the turn of this test.

 

[TomW]

Test #1 Run #2 results are in. The N whip with 8 spider leg 69mm below the copper disk out performed the other in the second (West) position. It out performed in both positions. Next test will be Test #2 Run #1, everything same as Test #1 Run #2 except the underperforming antenna will receive a new N base whip.

 

[ab cd]

Test #1 Run #2 results are in. The N whip with 8 spider leg 69mm below the copper disk out performed the other in the second (West) position. It out performed in both positions. Next test will be Test #2 Run #1, everything same as Test #1 Run #2 except the underperforming antenna will receive a new N base whip.

Great! Thanks for sharing your findings.

I feel that changing SO239 to N should not make any difference. However the trial run results will tell the truth.

 

[TomW]

Great! Thanks for sharing your findings.

I feel that changing SO239 to N should not make any difference. However the trial run results will tell the truth.

I agree, however with different connectors, the question would constantly arise... so best nip it in the bud now. Also there were several differences: so239 to f converter plus f female to f female vs N male to f female; and the so239 has an 12 gauge whip whereas the N has a 14 gauge whip. Getting to all Ns will make it easier later..... Also, now I'll be able to test the two N whips against each other!

 

[ab cd]

I agree, however with different connectors, the question would constantly arise... so best nip it in the bud now. Also there were several differences: so239 to f converter plus f female to f female vs N male to f female; and the so239 has an 12 gauge whip whereas the N has a 14 gauge whip. Getting to all Ns will make it easier later..... Also, now I'll be able to test the two N whips against each other!

This is great! Eliminate differences of connectors, then what remains is pure comparison of antenna elements.

 

[TomW]

Test #2 Run #1 is complete. Amazingly, the new N whip improved the antenna with the shorter spider legs (69mm to the teflon). The results are quite close. Next move (Run #2) will be with antenna locations swapped.

 

[TomW]

Test #2 Run #2 is complete. Only change from Run #1 above is swapping the antennas east to west 24 inches. Again, this time the east antenna wins.... Test #2 Run #3 will be with the antennas in current position, with the RPi's and other components swapped to see if the current east system is better than west.

 

[trigger]

Great work Tom.

 

[ab cd]

Thanks Tom for sharing you findings. Very scientific approach to find effect of various factors.

 

[TomW]

Test #3 Run #2 is complete....the plot thickens. In this test, the Pi-receiver-diplexer-power injector-amp trains were swapped from their positions in Run #2. The west antenna (with the former train from the east) improved in aircraft seen and mode-s traffic, but performed worse in range but slightly better in coverage (note in the coverage map, the pattern coverage colors match the color in the results spreadsheet...Green=East).
Next test will be with the trains returned to original, and I'll start moving individual components (e.g. amp) between the two antennas.

 

[ab cd]

Test #3 Run #2 is complete....the plot thickens. In this test, the Pi-receiver-diplexer-power injector-amp trains were swapped from their positions in Run #2. The west antenna (with the former train from the east) improved in aircraft seen and mode-s traffic, but performed worse in range but slightly better in coverage (note in the coverage map, the pattern coverage colors match the color in the results spreadsheet...Green=East).
Next test will be with the trains returned to original, and I'll start moving individual components (e.g. amp) between the two antennas.

This time West won! So it is not only antenna, the train also counts. Thanks for sharing a very valuable information.

 

[TomW]

Amplifier Performance Matters - Test #3 Run #1 proves this. Comparing a previous test with this one, with the only difference being amps swapped one of the amps wins decisively each time. Next test will be swapping the power supplies with the current configuration.