As Vic said, your scope looks very well collimated to me in the one image I downloaded. Trying to figure out where that residual tilt came from. Now, I do defer to Vic on the problems with single axis collimation, because there are some small issues. But we can deal with them.
Here's what I am thinking. You mentioned getting the primary reflection (hotspot) onto the secondary major axis, then tilt a single screw moving along the major axis until the hot spot was centered on the minor axis, too. Ideally that would be where the site tube crosshairs appear to intersect, too, with a perfectly placed secondary. But, what we're looking for at this stage is focuser axial alignment defined by the site tube crosshairs, not the secondary axes. So, I might suggest moving the hotspot onto the cross hairs instead of eyeballing the major and minor axes. As your are moving the hotspot along (or parallel to) the major axis, if you miss the hotspot you back off and refine your rotation until that single inline tilt brings the hot spot onto the crosshair as close as you like. That is focuser axial alignment regardless of secondary position which should be close because you put it there.
Now your focuser axis is collimated, but not necessarily the best secondary position you can achieve. However, interestingly and importantly, now that the focuser axis is pointing at the well centered hotspot, then the reflected edge of the primary should (necessarily) be concentric with the focuser draw tube regardless of the secondary position. This is true because the center of the site tube (cross hair) is on the center of the primary (hotspot). These are two of our collimation signatures, as well: draw tube and reflected primary edge. So we can use these signatures to refine the secondary position (also a signature) to them from its initial position set above. This is the beauty of refining your collimation at the apex, any error becomes very apparent.
With the focuser axis collimated, focus out to the apex where the reflected edge of the primary appears the same size (or slightly smaller) than the secondary mirror. Any deviation in your secondary position will become very apparent against the - now collimated - reflected mirror edge (and focuser draw tube). This is why we sometimes mention seeing the edge clips. They are not a collimation signature, but they are a proxy for the reflected edge of the primary. While your focuser axis is still collimated to the hotspot, you can see any deviation between the secondary (one signature) and the primary edge which is also concentric with the focuser draw tube (the other two signatures). Refine your secondary position using fore and aft and rotation - and even a little tilt - as needed until the edge of the secondary coincides with the edge of primary reflection when the focuser axis is, again, collimated to the hotspot. Your mirror clips will be easily visible all around and you should have better secondary position.
Of course adjusting the secondary position will mess up your initial collimation. That's okay. We'll bring it back after refining the secondary position (as above) to the other two signatures. This time with a little better secondary position rotate the hotspot back onto the secondary centerline. Loosen the one screw, refine rotation, and repeat the single axis secondary alignment at the apex. Tilt the hotspot onto the crosshair along the secondary major axis. Tighten one screw, loosen the other two or visa versa. If you miss it. Back off, refine your rotation a little, then tilt the secondary along that axis, again, until you hit the center mark with the crosshair.
This time, with better secondary position (relative to the other two signatures) you will notice a beautiful thing as you bring the hotspot back onto the crosshair. All three collimation signatures will come together at the same time. Draw tube, (refined position of the) secondary, and lastly the reflected primary edge as the focuser axis is collimated. It's like magic when you see it happen and your secondary offset will point at the primary mirror, too. It'll look as good as it performs.
So, yea, I think the one step you are missing (probably my fault) is refining your signatures at the apex. When the focuser axis is collimated to the primary center, both the draw tube and the primary reflection will be concentric with each other. We can use those concentric signatures to refine the secondary position so it is also concentric with both (when the focuser axis is collimated). I think that is an important step you missed above. I only briefly mentioned earlier. This is the magic part...and how I refine collimation in 10 minutes prior to each observing session. You can too. It really is easier than it sounds.
Step 4. Insert laser and use the adjustment screws to fine tune laser within hotspot.
The other thing that catches my eye, after all the trouble holding onto your signatures above, is using the secondary collimation screws again to adjust the tilt of the secondary in step 4 may have induced some unwanted tilt and rotation. A small amount. If you used the three screws independently, then I am sure of it. The two orthogonal screws are responsible as they induce off axis tilt of the secondary and they also tend to rotate the secondary a little bit with off axis torque applied to the secondary.
This was my biggest problem using a site tube, too, I kept retweaking the secondary as I was trying to collimate the focuser axis. I could see the tilt and rotation looking down the site tube, then I'd try to correct it in a seemingly endless battle fighting to hold the secondary position, rotation, and signature while trying to achieve collimation. You cannot see this happening with a laser, but you can with a site tube. The result is frustration.
This is why I prefer the single axis collimation, so much easier. The single screw on axis greatly reduces those errors. You can do it with a laser, too, but it's hard to avoid the temptation of using all three screws independently coming at collimation form three different directions. One direction is actually easier and it really helps hold onto your collimation signatures we worked so hard to achieve.
Sorry for the long read...
Edited by Asbytec, 24 May 2020 - 08:29 AM.