I moved the scalar about 15mm back by fixing the tripod legs to the front of it. To find Paksat 1R again, I set the elevation and pointed the dish in vaguely the right direction. 40E is very obvious so once that was found, I nudged the dish a bit and found the beacons easily this time. I don't know why I missed them before, I might have set the wrong polarisation.
View attachment 101081 View attachment 101082
I peaked the dish on 4196.5H. I then scanned using Crazyscan
View attachment 101084
Then I used EBSPro on 3818 to peak the feed and scalar positions and then re-peaked the dish and got this scan result:
View attachment 101085
Some extra TPs compared to the
first scan but some missing. 3984 was a dB or so lower as well so I re-peaked for SNR on EBS on 3984 and got:
View attachment 101086
3818 now gone but I could lock it manually. I was using 1MHz scan steps with 4MHz width. I couldn't lock anything on horizontal although I could see them on the spectrum.
I was quite surprised to get 3728V in view of the bit rate. Here's a capture of Geo News which seems to be a slide show with voice over:
View attachment 101087
Since we've talked about it, here's AJK-TV on 4184V
View attachment 101088
As CH said, there appears to be some frequency dependence associated with fine tweaks of dish and feed alignment hence losing 3818 when 4184 was peaked. So, the tin dish didn't do too badly in the end.
View attachment 101089