After the arrival of the C5 Bob's knob, I installed them right away. During the process, I found that the retaining ring of the corrector was loosen!
Anyway, I installed the new collimation knobs one by one, and then remove the corrector altogether for cleaning. To my surprise, the secondary was a bit fogged, and thank God that after some cleaning, it is shinny as new! I removed the corrector entirely in the hope to find the original marker at the edge so that I could correct the (relative) orientation of the corrector. Luckily, I found it!
Collimation with terrestrial target is not going to be accurate, so I just very roughly collimated by looking at the front of the scope, wishing that the correct was in proper location at the same time.
The sky cleared suddenly tonight, and I pulled out the Nexstar mount to do the collimation with tracking. It is windy so I don't expect to do a full collimation but just a rough one, so that I could fine tune it within a few seconds before I do any imaging.
Pointing at the sky randomly, no star was shown basically no matter how I adjusted the focuser, and it indicates really bad collimation. Swing to Sirius reveal a very ugly out the focus pattern, fine tuning the focus does not help but it shows an very asymmetric pattern, added with the turbulence generated by the poor seeing, I was a bit shocked.
Then I turned the focuser knob to a very out-of-focus position until the pattern was stable. Adjust the collimation screw one-by-one, to know which screw corresponding to which direction. After a couple of minutes, rough collimation was done.
A good collimation is not just about correctness, but it is also about the even tightnesses of the screws. The reason is that, if all three collimation screws were tight, it could hold much better than loose screws during transportation or even storage. Therefore, I keep adjusting all three screws carefully, in order to tighten (not over tightening) all those screws. Things were done within another couple of minutes.
I ended up using over 15 minutes in doing that. Re-adjusting the focuser, the diffraction pattern is now much better! Due to the poor seeing, for the most of the time, I couldn't see a nice diffraction pattern but then during moments of brief stableness, I could see it was quite nice. Maybe even enough for imaging too. I also found that background dim stars are shown.
When I was about to call an end to the task of tonight, I do a swing to M41, and it was shown very nicely. I guess only a very minor touch will be required next time when I want to do imaging. Last time the Jupiter imaging session was in vain both due to poor collimation (but I got no screw driver with me), and mis-orientated corrector, and now they were all fixed.