This is an update about my 3.7m dish.
Sometime in late summer last year, we had a lightning storm which must have damaged the LNB as it just stopped working. I replaced it with a new Invacom C120, and although the system worked during the day, BBC and ITV would break up early evening. Then came some really heavy winds that rotated the dish on its mounting pole and bent one of the support bolts. So the winter was spent with no satellite reception, just poor quality streaming TV over the internet.
With spring weather over the last few weeks and the World Cup looming, I had another go at getting the pictures back. First off, I bought a ROVER S2 meter as even on my first attempt, my Satlook Digital NIT monitor was too noisy to get a reliable lock on any of the Astra 2D transponders.
The starting point was to install two anchors in the concrete pad behind the dish and use wire cables and turn-buckles to secure each side of the the rear mount framework. (Pic2). I then found that there was some interaction between the azimuth and elevation adjustments. Tightening the bolts to lock the dish in the best azimuth would move the elevation angle slightly. With hindsight, I should have remove my original mounting pole and replaced it with a much larger, correct diameter pole when I installed the 3.7m dish rather than just using longer securing bolts.
In the end the procedure I used was to slacked the bolts just so that I could rotate the dish – set best azimuth using the signal meter – tighten the pole bolts one at a time in fractions of a turn and then make a fine elevation adjustment.
Moving to the front of the dish, I machined out the centre of the cast aluminium support and then fixed it back to the support legs with much longer bolts and multiple nuts so that I could make fine adjustments of the position of the mount relative to to the axis (Pic3). I make a new mounting collar with single grub screw to lock the feedhorn into place. Getting the feed horn in the right place was difficult as there are four things to adjust, the x and y positions relative to the axis and the two tilt axes. So the feed horn could be off centre with both x and y errors, but tilted so that it still points at the centre of the dish.
My procedure was to first measure the distances from front surface of the mount to four equispaced points on the edge of the dish. I then used the screw adjustment on the base of the support legs to get the distances equal. Then I mounted a laser pointer into the feedhorn and adjusted the bolts at top of the mount to get the laser pointing at the centre of the dish base plate. I could find no way of getting the laser exactly concentric with the axis of the feed horn, so the best I could do was to ensure that the spot moved in a circle around the centre point as I rotated the feedhorn. I estimate that I got the laser axis within 2mm of the centre which about 0.1° error.
I printed a scale of |||||||| characters (Times 12pt Bold) spaced 0.85mm apart on a sticky label and stuck it around the rim of the LNB flange (Pic4). With a pointer glued to the support base, and given the flange diameter of 50mm, I could read the skew angle to an accuracy of about 2°
The following measurement were made between 2-3pm local time on 3/04/2010. Once I got a lock onto transponder 45 (10773H), I used MER(Modulation Error Ratio) to fine tune the system. See
hXXp://pacificsatellitenews.com.au/2009/06/mer-modulated-error-ratio-what-is-it/
Angle MER
21 11.3
22 11.4
23 11.6
24 11.8
25 12.2
26 12.2
27 12.1
28 11.7
29 11.2
The correct skew angle for my location is -9.3°. The scale above is somewhat arbitary as it was difficult to set a horizontal level. However, it does show a clear peak at mark 26 with an error band of about +- 2°
These results were taken with the front of the feedhorn set 1369mm from the base plate. This the focal length given by Prodeline (1371mm) minus 3mm to bring the point just inside the tube. I tried moving the focus point in and out by +/5mm but with no measurable change in the MER. I set the spacing of the annular ring at 4.3mm to match the F/D ratio of 0.37. Again moving the ring in and out by about +/- 2mm produced little measurable difference.
The final measured results for the transponders that I am interested in were
Frequency Station Noise Margin MER
tp41 10714H Channel 4 4.5dB 12.1dB
tp42 10729V Channel 4+1 3.8dB 11.9dB
tp44 10758V ITV London 3.4dB 11.0dB
tp45 10773H BBC1 , 2, 3,4 4.5dB 12.2dB
tp46 10788V BBC 1 Regions 3.4dB 11.0dB
tp49 10832H ITV Regions 4.2dB 11.8dB
tp50 10847V BBC HD 3.6dB 11.0dB
tp53 10891H ITV Regions 4.2dB 11.7dB
It is clear that the horizontal polarized stations are generally better than the vertical ones, but then I did set the skew using a horizontal station. I would rather have the best I can for these rather than try balance the performance over both polarizations. In any case the ROVER S2 meter gave a Quality = PASS on all transponders.
The outcome of all this careful work?. Last night at 10pm, I was able to see all the channels, including BBC HD with no breakup. All's well until the next storm!
PJG
. 
