lining up microwave antenna's
 

Maybe somebody can shed some light on the following matter:

Recently I have installed an 18 GHz Microwave link, consisting of NEC-
Pasolink gear and Andrews Microwave Dishes.
Because it was a first for me, I had thought over the process of lining
up the dishes. I had heard of and read about the side-lobe vs the main-
lobe.
So I figured out the my coworker on the other end of the link should move
his dish and I would tell him the reading on my voltmeter. The higher the
better and at the peak reading he would fix the dish. Then I would do the
same on my end. And then his end a second time and then my end a second
time
We would repeat this procedure for the vertical line-up as well.
I had calculated an expected attenuation which could be converted in a
voltage reading.

Assume the reading should say 3,5 V. Well I never got anything better
than 2.8 / 2.9 volt.

COnsequently I asked NEC for advice and they said that 'you should line
up you own end using the voltmeter, and not the opposite end'. With this
advice we repeated everything and reached the expected reading of 3.5
volt. The 2.8 volt was a sidelobe of the antenna-beam

What I don't understand is that when I rotate my dish (either
horizontally or vertically) I can go from sidelobe to mainlobe to
sidelobe. I don't move the dish further left or right, I only rotate it
around a vertical or horizontal axis. When I would move the dish further
left or right or up or down, than I can visualize going from lobe to
lobe. Not by rotating the dish
That is why I used the opposite end to move the beam.

Anybody can explain where I go wrong

Regards, Ad

On Fri, 11 Jul 2003 21:35:16 +0200, wrote:

Anybody can explain where I go wrong

Regards, Ad

Hi Ad,

Looks like cross-polarization issue. As you rotate, they come into
conflict (the polarizations) and then re-emerge from conflict.
Actually, it sounds fairly bullet proof with what you got (plenty of
signal, the system probably doesn't need more than a volt, as you
would measure it).

73's
Richard Clark, KB7QHC

On Fri, 11 Jul 2003 21:35:16 +0200, wrote:

Maybe somebody can shed some light on the following matter:

Recently I have installed an 18 GHz Microwave link, consisting of NEC-
Pasolink gear and Andrews Microwave Dishes.
Because it was a first for me, I had thought over the process of lining
up the dishes. I had heard of and read about the side-lobe vs the main-
lobe.
So I figured out the my coworker on the other end of the link should move
his dish and I would tell him the reading on my voltmeter. The higher the
better and at the peak reading he would fix the dish. Then I would do the
same on my end. And then his end a second time and then my end a second
time
We would repeat this procedure for the vertical line-up as well.
I had calculated an expected attenuation which could be converted in a
voltage reading.

Assume the reading should say 3,5 V. Well I never got anything better
than 2.8 / 2.9 volt.

COnsequently I asked NEC for advice and they said that 'you should line
up you own end using the voltmeter, and not the opposite end'. With this
advice we repeated everything and reached the expected reading of 3.5
volt. The 2.8 volt was a sidelobe of the antenna-beam

What I don't understand is that when I rotate my dish (either
horizontally or vertically) I can go from sidelobe to mainlobe to
sidelobe. I don't move the dish further left or right, I only rotate it
around a vertical or horizontal axis. When I would move the dish further
left or right or up or down, than I can visualize going from lobe to
lobe. Not by rotating the dish
That is why I used the opposite end to move the beam.

Anybody can explain where I go wrong

Regards, Ad

When aligning the antennas you must swing each all the way left and
all the way right to where you are well past the beam width and any
lobes. Of course you note the signal strength as you do this. That is
the only way to be sure that you are not on a side lobe. Just moving
it a little each side of a peak or even moving it until you loose the
signal is not sufficient. You could hit a null between the main and a
side lobe and think that you have gone far enough but you still could
peak on a side lobe.
By swinging way past the lobes, each way, and noting the signal
strength as you go will assure that you find the main lobe.
Do the same in the vertical plane also.

73
Gary K4FMX

Gary, K4FMK wrote:
"Do the same in the vertical plane also."

Never had a minor lobe identification problem.

Never improved over the bubble-level set of vertical elevation angle on
long paths either, but I always tried.

I had always calculated my path gains and losses, and my best received
carrier power was very nearly always within a db of my calculations. If
not, I had a problem. Fortunately, that was very rare.

Best regards, Richard Harrison, KB5WZI

Floyd Davidson wrote:
"Did you actually do very many?"

I`ve done single-hops, several-hop systems, and transcontinental
systems. I`ve done them on-shore, off-shore, and in a multi-hop loop
system on-shore and off-shore. I`ve done several systems with paths
sandwiched between tall buildings. I`ve done 960 radio, 2-GHz, and 6-GHz
systems. I`ve done space-diversity systems, hot-standby, and unprotected
systems. I`ve done solid-state systems, vacuum-tube systems, etc., etc.

I have no reason to say anything which is untrue. Floyd knows of an
anomalous hop in the desert. The path suffers reflections, else it would
not have great variation of signal with height. I know of many anomalous
systems, but I never built one. All of mine worked as designed.

Best regards, Richard Harrison, KB5WZI

Floyd Davidson wrote:
"---I`ll bet you just exagerated a little, that`s all."

Too many hours of daylight on Floyd are taking their toll.

Everything you work with is known. precisely, including path attenuation
under normal propagation conditions. Normally, you don`t have a path
grazing at a highly reflective point. Your path survey discloses path
detractions and you adjust for the possibility of distructive
interference. You may opt for a high / low antenna placement for the
path ends, diversity, more clearance, shorter paths, and brute-force
fade margins. The high / low option lets you move the reflection point
and the reflection. Long microwave systems must have huge fade margins
anyway due to noise buildup from individual path contributions. A
receiver not too much below the overload signal point is a very quiet
receiver and contributes almost no noise to a system.

When the path design is right, the as-built numbers are almost exactly
as calculated, whether you believe it or not.

Best regards, Richard Harrison, KB5WZI

Floyd Davidson wrote:
"Richard you can cut the bull**** out."

Floyd claims to have measured microwave paths for 40 years. I`ve been
doing it since 1960, so that`s about as long.

I`ve made repeated measurements over a number of years on the same
repeaters. During normal propagation, which is by far most of the time,
path loss like other system losses is very constant.

Of course there are periods of anomalous propagation. It depends on
location, season, and time of day. It`s worse when the atmosphere is
stagnant.

I`m sure that marginal paths with insufficient clearance and other
problems may have propagation which comes and goes. I`ve seen some, but
I haven`t built any like that.

Best regards, Richard Harrison, KB5WZI

Floyd Davidson wrote:
"Suggesting your designs have never collided with such an anomaly just
means you didn`t do many or you`re failing to recall them."

It means care was taken in path planning. Microwave propagation is very
predictable. The only variable is the atmosphere and long statistical
records are available which allow a design for a percent reliability,
99.99 out to the limit of your budget.

Of course, on any particular path, there will be fades on some foggy
morning, but that doesn`t guarantee failure if you`ve planned for it.
The phone company uses alternate routing. I`ve designed similar systems.

What I said was, the as-built hops performed as designed during normal
propagation. That`s nearly all of the time. The received carrier power
measures remarkably close to the predicted value nearly every time and
if it doesn`t when the the alignment is complete, in my experience, it
has been always due to a defect in the system, not in its design..

I have never had to go back and "beef up" a hop to obtain reliability,
and I`ve put some hops in tough territory. The arithmetic is simple and
so are the criteria.

Best regards, Richard Harrison, KB5WZI

Floyd Davidson wrote:
"This thing (TV microwave relay station) was located 50 miles from town
at 8500 foot up on a mountain top. They had a 50 foot tower (at the TV
studios I suppose)----they turned it on and it worked great."

So, if it wasn`t broke, why did they fix it?

Floyd also wrote:
"But somebody had the smart idea to see what happens if they slide the
dish down the tower to see if the signal would improve. It did!."

I wasn`t there, so I can only speculate, but I might have not been
surprised by those results.

On a 50-mile path with plenty of mid-path clearance, propagation is
similar to communications with a satellite. One difference at the
Arizona latitude is the vertical angle the dish path makes with the
Earth. The low angle the dish on the terrestrial path makes with the
Earth, makes it vulnerable to reflections from the Earth. The higher the
dish is placed, the more vulnerable it becomes. That`s a reason to go
high / low on a reflective path, and not high/high. (Low/low won`t make
the trip on most long paths due to Earth curvature).

Best regards, Richard Harrison, KB5WZI

Floyd Davidson wrote:
"But, it doesn`t emulate "a satellite" shot in the slightest."

A satellite is a microwave repeater in the sky which is usually not
susceptible to earth reflections.

Terrestrial microwave paths are often designed to minimize earth
reflection.

Reason is, on a long path, the lengths of the direct ray and the Earth
reflected ray are almost the same. Reflection reverses the phase of the
reflected ray so that the reflected ray detracts from the direct ray
when they combine at the receiver. They arrive out-of-phase, at least at
some antenna altitudes.

The phase of the signal undergoes 180-degrees of phase change as we
examine the signal either forward or backwards from the plane of the
signal. That is to say, the phase of the signal is a function of
distance along the path. When a second signal source is created by Earth
reflection on the path, the combination of the direct and reflected rays
produces strata of stronger and weaker signal intensities as altitude
changes. No reflected signal, no significant variation of signal with
altitude, given sufficient path clearance. This is a worthwhile design
goal.

No matter which end of the TV relay path was varied to produce a
significant signal strength change, the attenuation at the higher
elevation was probably caused by an Earth reflection. Another reflecting
surface near the microwave path could have been the reflector. I`ve seen
that in the city canyons I`ve had to squeeze signals through. But on
most paths, the Earth is the most likely reflector.

Best regards, Richard Harrison, KB5WZI

Floyd Davidson wrote:
"---the look angle for a geosynchronous satellite here is about 12
degrees if the satellite is directly south."

Floyd is in Alaska. Floyd`s example of anomalous propagation from
reflection to a terrestial transmission was in Arizona. On the equator,
the "look angle" may be nearly straight up.

The reason a satellite dish is less susceptible to earth reflections is
that the satellite dish is not aimed to pick the reflections up. The
satellite dish is aimed at the sky.

A terrestrial microwave dish aimed directly at the satellite dish is
likely not transmitting an interfering frequency, but if so, it is
unlikely to be aligned well enough or above the horizon of the satellite
dish.

The approximate specifications of a 6-foot dish for beamwidth and gain
versus frequency a

1.3 GHz 9 deg. 25 dbi

2.3 GHz 5 deg. 30 dbi

3.5 GHz 4 deg. 33 dbi

6 GHz 2 deg. 36 dbi

10 GHz 1.5 deg 43 dbi

25 GHz 0.5 deg 50 dbi

Sources of the above are the "RSGB VHF-UHF Manual" and the ARRL Antenna
book (they agree).

In moderate latitudes, the satellite earth station antenna is really
looking up. It is quite likely terrestrial signals are not within range
of its bandwidth, beamwidth, or distance.

Best regards, Richard Harrison, KB5WZI

Floyd Davidson wrote:
"And darned if a few years later a build up of rhime ice on the 43 feet
of tower above the dish didn`t fall down and bend that dish into a pile
of rubble too,"

By Floyd`s logic the lesson must be that a dish must always top the
tower, out of harms way.

The antenna does not have to receive zero interference, though that
would be nice. We are always susceptible to some interference from
somewhere at some time. Multiphop systems often must reuse just a few
frequency pairs over and over. It`s all the regulators will allow.
Situations arise when anomalous propagation provides strong signals at
extraordinary distances.

Planning includes avoiding azimuths which would repeat to present
interference at a great distance along with a repetition of a frequency
which might interfere. A solution to interference is coordination.
Another is often high performance dishes which do a better job of
rejection. I`ve used the shielded variety from Antennas For
Communications (AFC) with good success.

Ice may be falling, but the sky isn`t in the case of microwave
interference.

Best regards, Richard Harrison, KB5WZI

Floyd Davidson wrote:
"The only sky falling is your claim that you calculate path losses to
within 1 dB for every single microwave shot."

I wrote many boring postings ago:
"---my best received carrier power was very nearly always within a db of
my calculations."

That was very nearly, not every single microwave shot. That was during
normal propagation conditions, not during fades. It was and is a true
statement. Such results can be achieved by anyone who plans the system
correctly using the required accurate information.

Best regards, Richard Harrison, KB5WZI