Well, since you just corrected the tilt error, any residual secondary mirror placement error is assumed to be rotation and/or fore and aft (offset)--not tilt.
That's a great point. Makes total sense. Tilt puts the cross hair on the center mark, it has to be (or assumed to be) right. That's what focuser axial alignment is. The rest is secondary placement.
This usually happens when the focuser axis is pointing above or below the OTA axis, instead of somewhere along the OTA axis--which means either the secondary mirror needs to be moved or the focuser axis needs to be moved.
Yep. And you're correct about keeping the spider vanes straight as possible, which means not to really use tension. Maybe a little bit, but I am hesitant to recommend it, too, for that reason.
It dawned on me one frustrating day as I was tilting the secondary laterally down toward the "bottom" of the OTA to bring the cross hairs downward onto the center mark, I was loosing the centered placement of the secondary. So, I heard you can use spider vane tension to pull it back to center. But, the tension required, after continuing to tilt the secondary, became so excessive I ran out of spider vane travel and the collimation signatures were just not there. At that point, I knew the spider vanes could not be kept parallel. And there should be no reason to use that much tension on the spider vanes. Something was obviously not right, so I decided to reset the spider and try again. And again. This is where better secondary rotation came into it's own. Not just relying on the face of the secondary looking circular under the focuser when using a colored mask to block reflections and show only the face of the secondary. Turns out, it may look good, but it's not necessarily that good. We can actually use the reflection of the primary instead of masking it. We 'need' to see the primary reflection.
I don't mean to rain on your parade, but you're not the first person to think that secondary mirror tilt would be better if the secondary mirror had only one adjustment axis.
Ah, darn it. No, of course not, no rain on my parade. Just sharing the epiphany...the epiphany of a noob is my parade.
With Don's help suggesting to toss the colored mask and use the primary reflection as a reference, it became immediately apparent why and how to ensure better secondary rotation. I absolutely know I am not the first to think of single axis alignment, but I am not sure how common that knowledge is among noobs and what significance it - single axis collimation - has on making life so much simpler. If my noob experience is any indicator, I think the tendency is to begin tilt collimation right away (after removing any colored mask) thinking the secondary placement step is done and assuming the secondary is well rotated because it looks like it. This is where many of us start to get into trouble almost immediately. The natural tendency is to look down the focuser and see the primary center at any random point and try to tilt the secondary to put the cross hairs on it. If we do not compensate for induced rotation, our signatures fall apart and we feel bad about it even though we are collimated. Hence the OP posting his picture above.
It became so easy to do the single directional focuser axial alignment along the secondary major axis, that I'd hope so many of us who post pictures like the OP (and me) could see the light (pun intended). Use a colored mask to make the secondary appear circular, then remove it and use the primary reflection as a reference to make sure rotation is, indeed, accurate. You almost cannot avoid being in a great starting position to begin single direction focuser axis collimation (with reiterations and refinements) until you hit the primary center within tolerance or better. And our collimation signatures hold nicely, too, so we feel good about it instead of frustrated. Dare I say different tools, then, tend to agree with each other, as well. And, as a bonus, normally we finish with the cross hairs moving onto the center mark adding some tension to the one inline secondary adjustment screw. This final bit of tension helps hold our secondary and collimation in place.
Yea, I make no claim to inventing the idea (wish I could, it's a great idea), only sharing one of a few ways to skin the same cat. Only hope to make it more common knowledge, even a noob (like me) can understand it easily enough, so much so it quickly becomes intuitive, and we almost cannot avoid it if we actually finish the placement step by using the primary reflection as a reference for more precise secondary rotation. And of course, as you say, some iterations to nail the primary center mark and hold our good looking collimation signatures. With good initial secondary rotation, it's so much easier to approach the primary center from one planned direction rather than from three random directions. It's so easy, it's almost automatic. Automatic is a good thing for a noob. Or at least to minimize the amount of orthogonal tilt we need so we're not fighting with collimation.
...(precise shimming is not and easy task). But without a leveling focuser, most users will need all three tilt adjustment screws, because the build precision in most commercially available "economy" Dobsonians is just not that precise.
Totally agree shimming to the OTA longitudinal axis is not easy, I spent wasted hours trying it years ago. However, shimming the focuser onto the secondary in it's natural location is not hard, and that's what is important, anyway. It's easier to square the moving parts to each other than to square everything to the tube's axis and hope they are squared to each other, too. Provided, that is as you say, the mass produced mechanics are good enough to make it easy to "chase" the secondary's location with a focuser shim (after secondary fore and aft placement). If the secondary is well placed fore and aft (toward or away from the primary along the X axis), really the focuser shim only has to be "up and down" across the tube along the Y axis.
I realize that is an assumption I make, that we can place a thin piece of plastic to shim the focuser high or low of the tube longitudinal axis to catch and center the secondary pretty well. I believe you about the other screws should not be totally "forbidden", we can use some tilt to help center the secondary, too, if needed. But, with a shim, the amount of tilt needed to center the secondary is much less. And if the single axis collimation brings us close to the primary center mark, we can "cheat" a little with minimal secondary tilt to hit the primary center without much damage to the concentric signatures. All is well that ends well.
IF all of the potential front end geometries are correct to significant precision, then the secondary mirror tilt adjustment will end up on the long axis and there's no need to allow for other tilt adjustments.
Interesting. So, in a sense, to use a 'true' single directional collimation, it's more critical to be mechanically centered and squared? It kind of makes sense, though I don't bother will all that mechanical squaring (dare I say mumbo jumbo?). Just square the moving parts (focuser and secondary) to each other and simply iterate through collimation, rotation, and placement until axial alignment is within tolerance and it looks good. I guess I can see where precise mechanical alignment is needed for precise single axis focuser alignment. But, in practice, I find such mechanical precision is not necessary. Just shim the focuser to chase the secondary location (high or low of the tube axis), rotate the secondary so the primary reflection and center mark are on the "long" axis, and go from there.
I mentioned above getting a piggy back ride on the back of giants, rather than standing on your shoulders. The latter implies I can see further than giants, which is not the case. I might be able to only see some of what you see, the view from a piggy back ride. Just thought that was cute.
My coffee is cold. I think it's a great discussion, though, and I hope the OP is following along without pulling his hair out. Collimation can be difficult to explain, more difficult than actually doing it.
Edited by Asbytec, 26 February 2020 - 06:36 PM.