Thursday, May 14, 2009

Improving the Signal to Noise Ratio (AKA Before & After)

At the suggestion of Jordan Arndt, VE6ZT I spent most of last evening removing my 222, 432, and 1.2ghz Yagis from my indoor, attic mounted "stack". The cluttered stack is pictured above. After completing that task, I relocated the 3 Element 6 Meter beam to the top of the mast which places it at about 22 feet above "true ground".

More importantly, it moved it a good 11 feet or so from the AC wiring and other cables that run along the attic floor.

The 2 Meter 7 Element Yagi remained it's it's location on the stack, about five feet above the wiring, and 16 feet above "true ground".

Prior to relocating the antennas, I did basic noise floor plots for both antennas using the WSJT software. Tuned to 50.125mhz and 144.200mhz, I adjusted the software for a 0 db noise level at an azimuth reading of 00. I then moved the antennas through each 10 degrees of azimuth and plotted the noise level for each one.

As you can see from the "before" numbers on both bands, noise floor levels would rise to as high as 11db on 6 Meters in some directions, and 7db on 2 Meters. Noise floor levels this high made it impossible to do any "weak signal" work in these directions. 6 Meter Meteor Scatter was very difficult, though e-skip work was certainly possible. Last season saw 149 grids worked on the Magic Band.

With the antennas removed and the 6 Meter beam relocated, I repeated the noise floor mapping without adjusting any settings. The readings were taken approximately 2 hours apart. In an indoor environment, the clock can be a major factor in regards to noise. As neighbors switch off appliances, TV's and other devices, the noise floor can change significantly in some directions.

But even allowing for that, these readings were taken at approximately 9:30PM local time, and as you can see from the plots, the results were impressive on both bands. Six showed the most improvement, as much as 20db better in some directions.

2 Meters did not benefit as much, but improvements of over 4db were noted there as well. For an installation like mine, that's a huge improvement!

I allowed WSJT to run overnight on 50.260 with the beam pointed out West. When I checked this morning just prior to 10am, I had captured three complete QSO's from Texas stations. That's a big improvement over any other unattended recording that I've tried in the past.

Bill Olson, K1DY also suggested adding additional ferritte cores to the feedlines of the 6 Meter antenna, which I did while it was on the ground. Jordan has suggested that using a balanced "T" match feed might help to reduce noise pickup as well. I'll likely try replacing the 6 Meter beam with an M2 CM3 when the budget will allow. For now, the additional ferritte chokes should keep feedline radiation and noise pickup to a minimum.

The plan now is to play with the antennas at length and see what the improvement translates to in terms of real world improvements. One thing I immediately noticed is that my lowest noise floor on 2 Meters is now at 40 degrees azimuth. It was almost due North, so most of my SSB and CW grids have been worked in that direction. The lower noise at 40 degrees opens up the Carolinas and there VHF population to me. Very excited about that.

The other antennas will be added to the rover package, and I'll pursue most of my other band work strictly during major contests. As Jordan pointed out, I wasn't gaining grids on those bands at home anyway, and having the antennas up there only served to degrade performance on the "work horse" bands of 6 & 2 Meters.

I can't thank Jordan & Bill enough for all the advice. Everyone on the VHF Reflectors has been helpful too.

Here are the "before and after" noise floor plots:

SIX METER NOISE PLOT FOR WSJT 50.125 (Before & After)

36 0db -13db
35 1db -13db
34 1db -13db
33 1db -14db
32 2db -12db
31 2.5db -9db
30 4db -8db
29 5db -8db
28 6db -4db
27 8db -2db
26 10db 0db
25 10db 1db
24 11db 2db
23 11db 4db
22 10db 4db
21 10db 4db
20 8db 4db
19 7db 3db
18 7db 2db
17 8db 0db
16 9db -2db
15 8db -4db
14 8db -4db
13 8db -4db
12 8db -5db
11 8db -7db
10 8db -7db
9 10db -8db
8 11db -7db
7 11db -7db
6 11db -8db
5 11db -8db
4 11db -9db
3 10d -10db
2 8db -11db
1 3db -12db
0 3db -12db

2 Meter Relative Noise Floor Readings.
Icom IC-746 Pro. Pre-Amp off. S Meter set to "5" to avoid AGC "OFF" AGC Setting. WSJT calibrated to read "O" db at heading 00. Freq=144.200mhz

"After" readings are after the removal of 222, 432, and 1.2ghz antennas on May 13, 2009. 6 Meter 3 element beam moved to top of stack, approx. 22 feet above true ground.


00 0db -2db
01 2db -2db
02 4db -3db
03 4db -6db
04 5db -6db
05 5db -1db
06 4db 0db
07 4db 0db
08 4db 0db
09 3db -2db
10 3db -3db
11 3.5db -4db
12 4db -4db
13 6db -4db
14 6db -1db
15 6db 0db
16 5db 0db
17 4db 0db
18 3db 0db
19 3db 0db
20 4db 0db
21 6db 2db
22 6db 3db
23 7db 3db
24 6db 3db
25 5db 3db
26 4db 2db
27 2db 1db
28 2db 1db
29 2db 1db
30 2db 0db
31 1db -1db
32 0db -2db
33 0db -2db
34 0db -2db
35 0db -2db
36 0db -2db

No comments: