I just have hands on experience. Since most received transponders in the CONUS are linear. Circular polarization for me is just when experimenting with the fringe satellites over the Atlantic.
But. I see a couple of things that lit up the old light bulb. Your installer got you a signal. Installed the dish the way it was boxed up. Bolted everything up and tuned it. Which is all good. And the Harvard ortho is top notch.
The one LNB getting better signal than the other. Why not concentrate on the weaker polarization by turning the feed a little. In and out. Tiny adjustments make big differences. And of course the hardware that fixes position in the scalar should never be tightened to the point it crushes the feed assy. Fine point Sharpie pens are a great thing to have here.
Perhaps masking tape to move your reference marks around.
I could almost guarantee your dish wasn't assembeld at the factory and support arm lengths tweaked for best scalar position. With those facts. And if you were to at least look at other examples in an Internet search. Some feeds are more close to the scalar plane facing the dish. And others do in fact exhibit my so called "stick out" (I believe I've been MIG welding a bit much recently).
But. With fixed geometry. As it stands. Your only option is to adjust your Harvard feed assembly.
If it clarifies anything. And it's worth a go. The 3 support arms that the plate the scalar is attached to.
Thinking back to when I installed the feed assembly I think the arms might have only fit on top but I'll see if I cant mount them on the bottom of that plate to get it a bit further. The focal distance measurement we got today was just shy of the 61.95 inches so that will probably get it closer to the correct value. The scalar ring is bolted to the support arm plate and is effectively fixed there. I could use spacers to move it closer to the dish but I don't see any way to get it further (minus moving the support arm mounting point like you mentioned).
The feedhorn itself does both slide in/out and twist within the scalar ring mount. We tried swapping in another feedhorn assembly that had fd ratio settings on it but had no luck. We'll throw the HS-200 back on tomorrow and continue to see if moving the feedhorn in and out improves the signal. It is most likely a little misaligned so I'll work to get it more square to the dish as well.
As far as the support arms go I don't see any way to adjust them as they are solid aluminum poles bolted to the dish frame. Could I be missing something here? For what its worth, measurements from the scalar ring to the dish rim matched in 3 places so I think its pretty well centered.
This all is just in an experimenters world mind you. My dish has a robust buttonhook feed mount and gives me a ton of adjustments. So support arm guys really don't have that advantage. The 2 aluminium pipes on the dish center mount have been changed to thick wall and so it is very sturdy.
But. Back to you. Have not see your entire dish front. If trimming the upper support brace lengths at the flattened section of them, close to the bolt holes, a touch. So that they could fit on the bottom of the plate w/o interfering with the scalar is not an option. Would spacing the braces from the dish face be an option? Perhaps 5-10 mm?
That would be far enough to let you see if any signal signal increase is to be had. Even IF .2-.3 dB.
But we haven't established if your feed assembly can be inserted deeper in the scalar. To compensate for your current focus you have for strongest signal.
You move the scalar away from the dish front, so you must move the feed throat an equal distance closer to to regain focus and signal strength. Make sense?
The f/D markings surely will get you in the ballpark. And is how my initial setup was. Adhering to all of the calcs.
End results differed after simply coming to the conclusion that my signals and finding transponders on different satellites sucked.
I could throw in an example. On 103 W there were a few transponders that gave me signals of 9.5-10.3 dB at most.
That was when initially setting up the dish from being mothballed for years. Measurements and calculations. placing everything where it should be with scalar at the correct distance from the dish and feed markings at the right place.
Took it for granted for months that it is-what-it-is. I just looked and signals on the same transponders are bouncing between an honest 15.1-15.3 dB.
So a few hours here and there of tweaks paid off with phenomenal results.
Mind you. One Enigma2 image returns different baseline signal strengths. It's in the software programming.
Was taking the time to be a different drummer and veer from the cut in stone calculations worth it? Hellyeah!
On a dish from the 80's.
a33 makes a great point. If using default satellites.xml files where a transponder is listed at....oh....4020 MHz.
Scanning it using the LNB local oscillator frequency of 5150 MHz (5.150 GHz). Could give a lower signal if the automatic frequency control of the electronics was at its limits.
A blindscan will tell you if that may be the fact, ignoring the frequencies and symbol rates in the .xml file.
The 4020 MHz may report as much as several Megahertz off. 4016, 4018, 4019. But I rarely see it drift above.
Norsat should have the drift to a minimum. But if you wanted to try. Like a33 says. Although keep the setting lower.
Like 1,2 MHz max.
Of course also if you did a manual scan using the .xml file. And suspect that signal could be a touch better.
You could download the channel list in an editor. Edit the transponder frequency. And blow it back in the receiver.
I often edit my satellites.xml file to show the actual blindscan frequencies and symbol rates. Even the modulation and stuff. Or just copy/paste the blindscan results into the file.
....bla bla bla. i know. "won't this guy shut-up?!!!"
