SatSig topic:

Receiver shows higher signal at edges of alignment

(Read 14691 times)

Jun 12^th, 2013 at 12:05pm

I recently aligned my girlfriend's satellite dish and was surprised to discover that the receiver reports a peak signal not at the center of alignment, but at the edges. That is to say, when adjusting either the azimuth or elevation I expected the reported signal to follow a bell curve, where the signal shows a maximum at the center and a minimum at the edges before dropping off to 0. However, what happens as you slowly sweep either horizontally or vertically is that the signal reported by the receiver looks like a "U", where at the first edge it quickly jumps from 0 to a maximum of around 75%, then it gradually goes down to a minimum of around 62% when at the center, and increases back to around 75% on the other side before dropping quickly off to 0 again.

After some hours of watching, the picture is most stable when I have it aligned at the center, even though the signal and quality from the receiver show lower numbers, so I leave it there. It look me an hour of adjusting and testing to get it this way (without any professional equipment, just iteratively tweaking, testing and gradually tightening). I'm very pleased with our picture's stability now (whereas it would freeze in the past), but a 62% signal seems a little lower than I'd expect, and I'm left scratching my head as to why I'm seeing this "U" shaped signal during alignment.

Does anyone know why?

Setup

- GoSAT GS7055HDi receiver
- 10 meters coax from receiver to motor
- Optibox DM3800 motor
- 80-100 cm dish (not there now to measure it exactly, but I could if it matters)
- IDLP-40SL+ universal LNB
- Approximate Location: 51N, 15E (Czech Republic)
- Alignment Target: Astra 3B (23.5E), Transponder: 11797 MHz, DVB-S2 8PSK, 3/4, 27500 (Skylink)
- Direct line of sight (no obstacles)

Background

This is my first dish alignment, and I have no professional equipment to use, only a joy for understanding things. This project started because my girlfriend was seeing frequent picture freezes on some channels, and when the older SG2100 motor that she had finally failed, I decided to dive in. I replaced it with the DM3800, and at the same time we upgraded from an older KAON KSC-570 CX receiver to the GoSAT GS7055HDi, for HD support. After installing the new motor, I went to realign the dish and discovered that the offset tilt previously set on the dish by a professional appeared to be 4.5deg (on the previous SG2100, that was 40 minus the 35.5deg set at the dish). That didn't match the calculated dish offset tilt of 6.7deg from the online satsig calculator, but it did match what I'd expect if the installer tried to peak out the signal at the edge, which might explain our previous picture instability. In my alignment, I found that the signal center occurs almost exactly at 6.7deg (for the DM3800, it's 35 minus 6.7 = 28.3deg set at the dish), so that makes me assume that I've got it right now. But still, why this "U" shaped signal during alignment?

Jun 12^th, 2013 at 5:02pm

It is possible to measure an antenna pattern and see two equal peaks, but I've never seen it on an antenna as small as 80 or 100cm at Ku band.
The explanation, which is probably not relevent in this case, is as follows:
There is a single big main beam shaped like a bell or parabola, but surrounding this is the first sidelobe ring, typically 10 to 15 dB lower. Many lower level sidelobes rings exist further out.
If you make an antenna pattern cut across the main beam you see a picture like that on the left, below:

The above (left) is the normal situation and it is quite easy to work out where to move the antenna to peak up.
There is a special case however to be aware of, particularly with large antennas: You see two similar height 'main' beams, and this pattern is similar in both azimuth and elevation. See right hand picture. Why ?
The reason is that the nearest you are getting to the main beam is when you are in the central low trough (first null) between the main beam and the first sidelobe ring. When you 'peak up' you are in fact 45 deg away from the main beam and sitting on the top of the first sidelobe ring, in one of four possible 45 deg quadrant positions.
As you move, you simply cut through the first sidelobe ring in two places and miss the main beam altogether. The same applies when you do the azimuth or elevation cut. Think of potentially moving in a full square around the main beam, missing it all the way!
The cure is to put the antenna into the bottom of the null between the two approximately equal first sidelobe peaks and then sweeping the antenna at right angles. You will then get dramatic improvement as you find the main beam itself.

The above phenomena is very visible on large teleport dishes where the receive system is likely to work suprisingly well, even if the dish gain is 15 dB low, when peaked on the first sidelobe.

In the case of a small antenna having two such obvious peaks is a mystery to me.

I would try:

a. Check out the above (improbable) theory !

b. Check the front rim of the dish is flat using the crossed threads test. Pull the two threads tight and check they just touch in the middle.

c. Optimise the polarisation. Turn the feed one way and the other, until severely degraded and measured signal quality levels are noted. Record the angle. Repeat going the other way, and then half the angle between the two equally degraded points. The ASTRA satellite polarisation may differ from calculation (e.g. 7 deg error?).

Note:

Severe distortion of the dish shape could result in a double hump main beam.
Severe distortion of the dish shape or out of focus feed distance or feed position could result in high sidelobes and adjacent satellite interference, where the first sidelobe hits an adjacent satellite.
Poor polarisation adjustment could result in x-pol interference being severe on axis. Signal power level wou

Jun 14^th, 2013 at 9:22am

Ok, that helps a lot, thanks for the info. To be more specific, the dish diameter is 87 cm. In regards to the things to try and notes you sent:

a. I'm not sure how to move the antenna to experiment with it. It's fixed on its aluminum feed arm and the dish hangs out over a balcony 7 stories up. I could swing it around with the motor and play with it, but am not sure how to move it around in a flexible way that allows me to place it where I want it then go back to the receiver and take measurements.

b. Interesting, the crossed threads test shows that the horizontal thread is about 1.2 cm in front of the vertical thread. I don't know if that's by design, or the dish is warped. I attached a side profile where you might see some curvature.

Do you think this amount of curvature of the front rim could be enough to cause what I'm seeing?

c. So far, I haven't seen where I can adjust the polarization on the LNB as I don't see where I can rotate it. I also read that Astra 3B is 0.7deg out of vertical (or is it a whole 7deg?). Regarding the signal quality, it is always roughly proportional to the signal and is pretty good, so at a signal of 62% (in the center) the quality is 74%. At the edge where I can get the signal up to 74%, the quality goes up to around 86%. But any slight movement in that peaked area and it drops to 0, so I can't leave it there.

According to your notes, it doesn't make me think that this is a polarization problem (because of the reasonably high quality), but more likely to be a dish shape or feed distance/position problem. I attached a couple of pictures of the dish, one of the whole thing and one of an interesting "feature" on the feed arm that was probably the result of a repair some years back (not me). It's a cracked mount on the feed arm that was repaired with wire, but it appears to be a non-load-bearing thing that holds the LNB's coax able. Despite this, the feed arm appears to be straight left to right, now whether it's bent higher or lower I don't know because I don't know where it should be.

Maybe a feed arm bent higher or lower than it should be could cause what I'm seeing?

The dish is about 15 years old, so this story might end in a new dish and LNB, but I wanted to give a try at figuring out what's causing it first, because it's interesting. Thanks again for the guidance...

Jun 14^th, 2013 at 9:42am

Ignore the crossed strings test. Your dish is obviously designed such that the front rim is NOT flat.

The LNB arm may well have sagged down due to part broken bracket just below the dish. If this can be tightened up, and the LNB raised slightly, note that the beam will go down and the elevation will need adjusting.

Polarisation adjustment is the rotation of the LNB in its clamp. With a motor driven polar mount the LNB is normally set, and fixed, exactly upright when the motor is centralised (aimed at a due south satellite). Astra may be off by 7 deg so if you spend most time watching Astra it will help to optimise the LNB polarisation for Astra.

A polar mount requires considerable skill to set up and may take several days work to re-align, once it gets misaligned. There are three adjustments: two elevation adjustments, behind the dish itself and the motor axis angle. The clamp to the pole needs to be adjusted exactly north-south.

Take great care working on the antenna in view of risk of falling from height and risk of dropping tools on people below. If in any doubt get a local satellite TV supplier/installer to help.

Best regards, Eric.

Jun 14^th, 2013 at 1:27pm

A new 'start all over again' thought has occurred to me.

It may be that the two peaks you see are not real. If you are driving the antenna using the motor and using the set top box readout as an indication of the antenna position then you may be misled. According to the direction in which the antenna moves towards a stop you can get two different position readouts for the same physical stop position.

Have you tried peaking up using the motor and then very gently lifting or lowering the LNB arm, or gently pushing or pulling on the lower dish edge to alter the elevation slightly.

Best regards, Eric.

Jun 14^th, 2013 at 1:27pm

Ok, the second graph you posted ("Starting just below main beam") is almost exactly what I'm seeing. I'll try to get the feed arm up a little higher and see what happens. I won't be on site there for a couple of days, but will post later on how it goes.

On Polar Mount Setup

I see how challenging polar mounts can be, even for an engineer (albeit in a different field). The whole thing took a weekend, for my first time. For the benefit of others, or if you want to point out if I did something wrong, here's what I did (at the risk of repeating information from elsewhere):

1. Read about polar mounts here: https://www.satsig.net/polmount.htm. Read about geostationary orbits and communications satellites on Wikipedia (https://en.wikipedia.org/wiki/Geostationary_orbit), for background. Found my longitude and latitude here: https://www.satsig.net/maps/lat-long-finder.htm.

2. Made sure the mounting pole is as vertical as possible using a level on multiple locations around it. It was.

3. Adjusted the elevation on the motor mount based on my latitude (mine had latitude markers on one side).

4. Adjusted the offset angle behind the dish. My original SG2100 motor had a 40deg crank, whereas DM3800 I now have has a 35deg crank, so the setting on the dish for me is 35deg - "polar mount dish offset tilt", where the offset tilt is available here: https://www.satsig.net/ssazelm.htm.

5. Found true south using a mobile phone compass (the iPhone one allows switching between magnetic and true south), making sure to calibrate it first and being careful about interference. Verified true south by using Google maps and a distant southern landmark. I used https://www.satsig.net/maps/lat-long-finder.htm. Rotated the entire motor/dish assembly towards true south as best as possible.

6. Set the receiver to USALS mode and input my longitude and latitude, so it can calculate where to point to other satellites for me. Told the receiver to go to Astra 3B (it already knew it's at 23.5degE).

7. Made small physical adjustments to the motor mount's azimuth and dish's declination angle until I found Astra 3B. This I knew by doing a channel scan. This took a while!

8. Tried to maximize the signal reported on my receiver by making iterative, very tiny adjustments to the azimuth and declination angles. For me, the center of the signal was surprisingly a minimum, which is why I posted in the first place.

9. Verified the motor could track to other satellites, like Astra 1L at 19.2degE. It could.

On Polarization

I can rotate the dish around with the motor and safely reach the LNB, so I'll try checking the polarization. I think I see what you mean, that the circular part of the LNB itself looks like it can rotate. and in fact, I think polarization should start properly set before adjusting the elevation and azimuth, right? Then it might need to be optimized later to maximize signal quality.

Thanks again..

Jun 14^th, 2013 at 1:44pm

Quote:

It may be that the two peaks you see are not real. If you are driving the antenna using the motor and using the set top box readout as an indication of the antenna position then you may be misled. According to the direction in which the antenna moves towards a stop you can get two different position readouts for the same physical stop position.

Have you tried peaking up using the motor and then very gently lifting or lowering the LNB arm, or gently pushing or pulling on the dish edge to alter the elevation slightly.

Thanks for the idea. I'm not using the receiver as an indication of position, only for signal and quality levels. I've verified that these two peaks happen with both physical adjustments and with adjustments using the motor's east/west buttons. (The TV is fortunately close to the dish so I can easily make adjustments and check the screen).

In fact, to verify more accurately that I'm "peaked up" (which is my case is "troughed down") at the signal center, I was using a poor man's technique. After I tell the motor to go to Astra 3B (23.5degE), I check the signal, which is 62%. I then click the East button on the motor and count the number of clicks until it rises to 63%. I tell the motor to return to Astra 3B (23.5degE) then do the same thing towards the west. It's around 12-13 clicks in each direction that the signal goes up to 63%, so that's how I know I'm centered. The motor makes finer adjustments more accurately than I can by hand.

I'll try exactly what you suggest with the LNB arm and elevation, and see if I can get it sorted out...

Jun 14^th, 2013 at 2:14pm

In a polar mount the LNB polarisation is physically set to zero relative to the LNB clamp ring mark and a line up/down the middle of the dish.
When the motor is centralised and you are looking at the due south satellite then the polarisation angle will actually be zero.
When the motor turns either way the polarisation tilts automatically to the side, as the whole dish and LNB arm is tilted sideways as a whole.
The above works perfectly assuming all satellites have their vertical polarisations aligned north-south parallel with the earth axis or at 90 deg to the equator.
A few satellites do not follow this normality. Eutelsat and Astra may be off by 3 or 7 deg. Not sure.

Best regards, Eric

Jun 14^th, 2013 at 2:47pm

There are nearby satellites with carriers on the same frequency:
19.2 H 11798
23.5 H 11797
28.2 H 11798
I wonder how many 'clicks' it takes to get from one to the next ?
Adjacent satellite interference could be a factor, if the antenna focus is poor and the beamwidth larger than it should be or if the sidelobes are unusually high. Quite how this can cause the two peaks symptom is not clear.

Best regards, Eric.

Jun 14^th, 2013 at 3:30pm

Before any arm bending of the LNB sort, I'll make a table of clicks vs signal and quality to see if it says anything.

I was actually initially having trouble picking up 11797. I thought I was pointed properly because I was receiving ČT2 (which according to https://www.skylink.cz/jak-na-to-24.html#Jak+naladit+programy is supposed to be on the same satellite at 12525) but channels on the 11797 transponder, like the Golf Channel, weren't working. I still had to make a small elevation adjustment, then optimize the azimuth adjustment for 11797 to work, which it now does. Unfortunately I can't be more scientific than that because I was doing it by hand and not documenting my steps at the time.

It doesn't make sense to me that orientation adjustments would have to be made for different transponders on the same satellite, I thought. I then thought of Astra 3A, which is now at 23.7deg and in an inclined orbit, according to https://en.wikipedia.org/wiki/Astra_3A. Was Astra 3A still transmitting and confusing me somehow? I reckoned that no, they can't have both transmitting on the same frequencies at the same time. They probably wouldn't even split services between the two. And in fact, the line on Wikipedia that says "As of October 2012, Astra 3A is in an inclined orbit at 23.7degE with all services carried by the adjacent Astra 3B craft." suggests to me that Astra 3A is just a floater at this point.

So the mystery of how I could find an orientation to receive 12525 but not 11797 is not yet solved. But I know now to keep an eye out for adjacent satellite signals in case they also confused me, and think this will get sorted out...

Jun 18^th, 2013 at 1:56pm

I attached a table of signal and quality vs clicks. Some features I noted:

- The increase in signal is a little faster when moving West for some reason.

- The signal and quality at both ends tops out at around the same place- signal=71, quality=84. That suggests to me that I'm not seeing interference from adjacent satellites but just the physical characteristics of the dish.

Also, I had a helper watch the signal and quality while I tried to bend the feed arm up. At rest, it showed signal=62, and as I bent it up I was able to get it to signal=65, so that suggests that the feed arm is too low and missing the main beam. But I was only able to bend it a little bit by hand, and to go further would probably have to do it with a torch or something else. Even then, I don't know how far to bend it to maximize the signal, and there's the risk of breaking it.

So there it sits until I think of some way to solve it, but for now, at least we've got a reasonably stable picture. I may yet play with the polarization to see what impact that has...

Mar 5^th, 2017 at 8:52pm

Is it possible you are using the same boom and LNB for your new larger dish as you were for the old small dish? If so, would this be your problem? The "reflective" pattern from the larger dish could be different than the pattern from the small dish. Ed.