96 replies to this topic

[Mark Harry]

Thanks, understood. Actually, the method of a double bounce as I outlined, only works with an unperforated Newt mirror. Not a Cass.
But I will add, mis-collimation is perhaps the biggest killer of really good performance; even with average to mediocre optics. It pays to play around a bit, and eak every bit of precision out of whichever method is used. I don't know if anyone else has 'discovered' the trail of multi-reflection dots that I use.
******
I have also used another trick with several of the last scopes I've made (Newts) A small 1/8" hole, drilled just to the side of the focuser, helps in preliminary alignment and positioning of the secondary by observing where the laser dot hits it. Really handy if an angle other than 90* is chosen from the optical axis. On my 14", I have approx 75-80* angle so as to facilitate standing flat-footed when the scope is pointed straight up. Working with the hole, offset for the faster scope can be easily seen/made. (again, food for thought)
M.

[Sean Puett]

I watched a YouTube video on collimation. It got the basic idea down. Then a had some problems with consistent collimation. I found that i was over tightening the laser in the focuser. I just snug it now to where i could spin it if i wanted to and i get pretty consistent collimation. Sometimes i end up adjusting it once i start using the scope and see that stars are not points. For me, there is no "good enough" it is collimated, or it is not. Yes or no. I wasn't as picky before the paracorr.:-)

[howard929]

Mark,

While this has been a most informative thread to a newb like myself I'd like to jump in with a change of direction to this discussion. So much has been mentioned about HOW to collimate a newt while a major cornerstone of that entire process hasn't been mentioned. I notice that no comment or question about the placement/location of the primary mirror center spot has been discussed in regards to the problem that the original poster is having.

You make telescope optics and if I may ask, how do you go about center spotting the primary mirrors, on the ones you sell if you do center spot them or even better (since you do seem to be an exacting type of guy) on the ones you use yourself?

Howard

[Mark Harry]

When polishing a mirror, I make darn sure that the centration is within .010" or less (better). Also, I might be considered ****, but last couple years, the mirrors I've made, wedge is minimized to around .001-.002". This pretty much assures that the center of the mirror is exactly where I scribe a small ring with a sharpie.
Some might think these tolerances are a bit too tight. I don't think so. When making a lens, the tolerances are much tighter than what's mentioned here.
Ones I use myself are the same as what's been sent out to customers.
A good question; all this talk relies on the pre-existing accuracy that you've addressed to make it all worthwhile.
M.

[FirstSight]

I always judge the centration of the laser by rotating it in the focuser. (I also rack the focuser in and out to see what slop is there too)
If the laser spot wanders, just keep rotating and watching, and the dot will circumscribe a circle of sorts. The center of the circle is where the laser dot ought to be pointed. Now, you can index the thing for reference, and the laser can be collimated. Simple.

Unfortunately, this is not reliably true with most of the less expensive lasers, especially the 1.25" format versions, because the machining of the laser body, the focuser and the 2"=>1.25" adapter are insufficiently precise to allow a snugly aligned fit of the laser in the focuser. (Technically, this is called a "registration" problem). However, even when the focuser and adapter are precisely machined, chances are the laser body is not, and it doesn't take much imprecision of fit at all to destroy accurate axial alignment of the laser in the focuser. When rotated in the focuser, even if a laser beam is perfectly aligned (collimated) with the laser's own axis, the projection of the beam onto the primary mirror can wander in a ovular path around the primary mirror's center due to the registration issue alone, and the "ovular" path traced will mimic a misaligned (miscollimated) laser beam, even though it is not. If instead the laser beam IS indeed misaligned, the effect of this error on the path of the rotated laser beam will be superimposed on that induced by the registration error, making it impossible to separate the extent to which either source of error is causing the rotated laser beam path to be off-center.

You can determine whether you have enough of a registration problem to prevent successful testing/collimation of the laser beam by placing the laser in the focuser and tightening it enough to minimize as much slop as possible without tightening it so much that you lose the ability to freely rotate the laser beam. If you have any difficulty grasping the laser in your fingers sufficient to test-rotate it without inducing deflection in the beam's projection on the primary mirror, if you have to be overly delicate in your attempt to rotate it to attempt to avoid inducing this kind of artificial deflection, it's going to be impossible to accurately align the beam using the suggested method.

FOR ILLUSTRATIVE EXAMPLE:
- I tried this rotation-test just yesterday with a 1.25" Orion Lasermate Deluxe I still keep around for outreach telescope tune-up clinics, placing it in my Moonlite CR-2 focuser + 2"=>1.25" adapter. Moonlite focusers and adapters are accurately machined, so they are not the problem; alas, Orion lasermates are not so precisely machined for size. The result was that it was impossible, despite my best efforts, to delicately rotate the laser in the focuser without inducing artificial lateral deflection. It was thus impossible to determine how accurately my beam was truly aligned by using this method.

BY CONTRAST, I next inserted my 2" Glatter laser in the focuser (sans adapter of course). The fit of laser body to focuser was snug and precise, and I could rotate the laser freely and the beam stayed exactly on-center of my primary mirror. (I had the advantage of knowing that both scope and Glatter laser were precisely collimated before I began this experiment). BTW: the value of a Glatter laser isn't just that they come very precisely aligned (collimated) out of the box, but also that the laser body is machined to extremely exact specifications. When inserted into an equally precisely machined focuser such as the Moonlite, you can feel (and hear) the resistance of the trapped air that resists instant displacement due to the snug fit. A Glatter laser makes a satisfying "whoosh!" sound when inserted into the focuser.

THERE IS A WAY AROUND THE PROBLEM of how to align (collimate) the beam of one of these less-expensive lasers, e.g. Orion Lasermate Deluxe. Construct a V-block in which to lay the laser body, secure the V-block so it can't move when a laser laid in it is rotated, and observe the behavior of the beam against a white wall fifteen or so feet away. The V-block method eliminates the registration problem. However, even after you've accurately aligned your laser using this method, it does nothing to solve any registration problems resulting from lack of snug fit of laser to focuser or adapter.

[Vic Menard]

...a major cornerstone of that entire process hasn't been mentioned. I notice that no comment or question about the placement/location of the primary mirror center spot has been discussed in regards to the problem that the original poster is having.

Actually, it was an assumption I made in my first comment in my first post in this thread. After that, I agree, this thread has wandered off on various tangents. But then, that's often the way it goes when discussing collimation...

[Vic Menard]

When polishing a mirror, I make darn sure that the centration is within .010" or less (better). Also, I might be considered ****, but last couple years, the mirrors I've made, wedge is minimized to around .001-.002". This pretty much assures that the center of the mirror is exactly where I scribe a small ring with a sharpie.

To clarify for the non-ATM readers, the measurements you're describing are mechanically derived--the centration/placement of the mirror or tool on the spindle, correct? I think a lot of Newtonian enthusiasts think the optician uses an optical test (like an interferometer), or some sort of digital assessment of the optical performance/testing (like Roddier or other digital image assessment), to precisely locate the optical center of the parabola.

Some might think these tolerances are a bit too tight. I don't think so.

I agree. One half of any residual error in centration of the optical axis (relative to the mechanically measured center of the mirror) would have to be added to the final primary mirror axial alignment to assess the total error. But since this error is a vector quantity (and is rarely added as such) and is typically very small for professionally figured optics (one or two hundredths of an inch, or less--often a threshold mechanical measurement unattainable by many amateurs), the user is (rightly) left to trusting the expertise of the optician.

Nils Olof and I have discussed this potential discrepancy at length, to the point of attempting to address the error by manipulating the centration of the reference spot to match the visual alignment of the high magnification "sweet spot" in the eyepiece. We agreed this could not be a "one-time" experiment, but should be the summation of several (many) observing sessions to eliminate the many variables that could inhibit the precision of the visual assessments. To date, I am unaware (and I believe the same is true for Nils Olof) of any Newtonian enthusiast who has used this process to improve the axial centration of the primary mirror for a professionally made mirror that is an on axis, figure of revolution, parabola (not an intentional off axis surface). In all cases, the measured center of the mirror has always been the optimal reference for alignment.

[Vic Menard]

...But I will add, mis-collimation is perhaps the biggest killer of really good performance; even with average to mediocre optics.

Agreed!

I don't know if anyone else has 'discovered' the trail of multi-reflection dots that I use.

I know several users who prefer to watch the alignment of the laser dots on the face of the secondary mirror over the (somewhat less precise) alignment of the return beam dot to the laser emitter.

I still prefer using the 1mm aperture stop. The laser diffraction pattern caused by the aperture stop means I get all the alignment accuracy of a good, thin beam laser along with the accuracy of a Barlowed laser, since I can see the silhouette of the center spot in the diffraction pattern. Even better, because all this occurs on the aperture stop target located near the bottom edge of my focuser drawtube, I can use my reading glasses since the alignments are directly visible, about ten inches in front of my nose! Using this accessory, I can align both axes to a very high precision, about +/-0.01-inch, or less. I've been able to verify this precision with my Infinity XLK autocollimator.

[Starman1]

One caution I'll give beginners:
My mirror came with a small X inscribed in the coating of the mirror. That X ostensibly marked the center of the mirror.
The Catseye template showed it was at least 2mm off-center!.
I could verify that with a small ruler.
So was the mirror ground off center? No, of course not. The mark was off-center.
I placed the center marker for collimation using the Catseye template and received the reward of tiny pinpoint star images with equally-illuminated diffraction rings. The Catseye template was right, the scratched X was not.
So, where placement of the center markers on mirrors by the manufacturer are concerned: Until you've ascertained the marker is correctly placed, assume it isn't.

[Mark Harry]

Ah, yes, quite true for locating on the spindle. My .010" is relatively comparative to your 1-2 hundredths, sounds like we're on the same page.
*****
I had an 8" F/5.9 for several years. Just for reference, I kept track of the displaced dot; the stray one that wasn't in line with the row of multiple bounce dots. With that particular scope, when that dot strayed more than about 1/8" from the laser aperture (glatter) very fine delicate planetary detail began to suffer. If it was on the edge of the laser ap, no degradation was noticed.
Another 8" F/4.72 with the same variance would be noted as being definitely a little "sour". It had to be dead-nuts right on.
Just what I've noticed.
****
The biggest reason I keep the wedge down, is that the primary can be removed for cleaning, and doesn't have to be indexed to be properly registered in the mirror cell. This idea worked well for the F/5.9. Just a very minor tweak, if any at all was necessary on a checkout.
Perhaps the deliberate non-concentric condition of centration could be addressed by looking at zonal positional difference in readings that would equate to certain specified PVW errors. Personally, if I was to do this, I would not use the Rayleigh standard. Perhaps half that amount of error would make sense.....another subject altogether.
M.

[SKYGZR]

Watch This

[Jason D]

Watch This

What did you find interesting about the above video?

[calibos]

I find it interesting that he invested money on Catseye tools as evidenced by the hotspot on his primary but never invested in some bobs knobs for his secondary . I wince when ever I see someone adjust their secondary with a screwdriver or allen key

[howard929]

I'm still using a screw driver to adjust the secondary mirror on my 8" newt but then again, once it's set correctly it doesn't require much adjusting.

Howard

[Jason D]

I find it interesting that he invested money on Catseye tools as evidenced by the hotspot on his primary but never invested in some bobs knobs for his secondary .

I have reviewed his other videos and none included any Catseye tools. I do not believe he owns any Catseye tools -- just the Hotspot sticker.

But his message in the video about how "laser collimators lie" seems to be a common theme in his videos. The issue is not with the laser collimators but rather with his misunderstanding of laser collimators usage. Laser collimators (unless used with a holographic attachment) are not meant to align the secondary mirror under the focuser.

Videos like this do more harm than good especially when the presenter comes across as someone with an authority on the subject of collimation. There are many beginners out there who will see that video and conclude that all laser collimators “lie” and should be avoided – far from the truth.

By the way, in his 9 series videos, he fails to cover focuser axial alignment. He basically centers/rounds the secondary mirror under the focuser then moves to barlowed laser. He completely skips the step about directing the forward laser beam to the center of the primary.

Lastly, it would be a good idea to give credit to the right individuals when such videos are made. For example, he could have mentioned Nils Olof Carlin as the inventor for the barlowed laser technique.

Jason

[Mark Harry]

I watched less than half that video, and I spotted 2 gross errors. First when he backs off on the setscew adjustments, and tightened the central retaining screw, he made the completely erroneous assumption that "everything was square and correct" (NOT SO!) Even with careful construction and assembly there will never be a perfect scenario of alignment.
Second, just prior to that, he grabbed and rotated the secondary assy as a whole, and introduced rotational error. He then proceeded to adjust the setscrews to take out the error. Again falsly assuming that a rotational error, and setscrew adjustment can be combined to effect "proper" collimation. To pop this theory, all he has to do, is look thru the focuser, with either cheshire, or naked eyeball to see how screwed up the centration of the secondary is!
I would certainly take that video with a grain of salt. Better information can be found right here on CN, or from something like Newtonian Notes (Novak)
M.

[Nils Olof Carlin]

That video again:
yes, one should check rotation, but I don't think it is more critical than looking through a centered peephole and turning the sec holder until things look symmetric, and leave the rotation there.
*Then* the obvious thing is center the laser beam on the primary spot by tilting the sec holder (it is not difficult to verify that the fully illuminated field is centered and reasonably circular - I think I have outlined how somewhere within the vast CN archives).

But it is important to understand what is the criterion for each step - and perhaps even more what isn't.

What you want by placing/rotating the secondary is to have the fully illuminated field centered, and circular. The criterion for this: after the tilt is done with the laser, use a peephole of some kind, illuminate the inside of the telescope enough to see some light in the secondary even outside the reflection of the primary - then check that the primary appears well centered in the well circular secondary. You may have to move the peephole outside the focal plane to make them look nearly the same size. If not well enough, move the secondary as indicated and start over.

The criterion for secondary tilt is obvious with a laser (well collimated in itself, but don't go overboard here).

Then, the criterion for the primary is in the barlowed laser (if that's what you use - otherwise a Cheshire, BlackCat or other collimation cap), centering the spot shadow. As Vic wrote, the optical center tends to be close enough to the geometric center of the mirror blank - at least if it is circular (it should be on commercial mirrors). The factory centering of the spot is something else - trust but verify, and replace when necessary.

Nothing else. The going back and forth as shown in the video is, I trust, trying to apply incorrect (and likely conflicting) criteria. If the collimation is grossly off (like if you have taken the telescope apart for a major overhaul), you may have to start with coarse collimation and do a second turn of finer collimation.

I see little need for that video camera. But it is quite doable to put a webcam on a Cheshire tool, even if you have to change the lens to one with longer focal length - this would let you twiddle the collimation knobs on the primary while watching the laptop...

Nils Olof

[FirstSight]

I see little need for that video camera. But it is quite doable to put a webcam on a Cheshire tool, even if you have to change the lens to one with longer focal length - this would let you twiddle the collimation knobs on the primary while watching the laptop...

Actually, this sounds like a terrific idea in need of practical implementation adapted to the Catseye Cheshire (aka Blackcat). The main design issue to solve is the mount adapter for the webcam: how to implement a mount that securely attaches to the Blackcat and centers the webcam lens over the peephole. The two other issues involved are choosing a webcam with a lens design/focal length appropriate for the application, and whether it's possible to design the mount so it accepts certain type(s) of off-the-shelf generic webcams, or whether it would be necessary to instead obtain the raw internal camera elements and build the mount around them.

The reason this seems like such a good idea to me is that it sure is handy whenever my scope is significantly out of collimation to begin the primary mirror alignment phase with the Glatter TuBlug/Laser and get the collimation close enough that what's left for the Catseye Blackcat and Autocollimators is the final tweaking to insure that alignment is dead spot-on and not merely close. It would be nice to be able to elect the same convenience using the BlackCat alone. A further reason this would be handy is that when the mirror is center-spotted with the Catseye triangle or hotspot, though it's easy to confidently get primary collimation close with the TuBlug, it's nearly impossible to determine when it's sufficiently spot-on due to the fact that the TuBlug projection of the triangular spot asymmetrically warps as you adjust the knobs, making it challenging to determine when the triangle is actually centered in the image. Occasionally, the Blackcat will confirm that I guessed right where spot-on collimation was using the TuBlug, but more often than not, it's off by enough to detract from the image without a little further tweaking using the BlackCat.

[Jason D]

I see little need for that video camera. But it is quite doable to put a webcam on a Cheshire tool, even if you have to change the lens to one with longer focal length - this would let you twiddle the collimation knobs on the primary while watching the laptop...

Actually, this sounds like a terrific idea in need of practical implementation adapted to the Catseye Cheshire (aka Blackcat). The main design issue to solve is the mount adapter for the webcam: how to implement a mount that securely attaches to the Blackcat and centers the webcam lens over the peephole. The two other issues involved are choosing a webcam with a lens design/focal length appropriate for the application, and whether it's possible to design the mount so it accepts certain type(s) of off-the-shelf generic webcams, or whether it would be necessary to instead obtain the raw internal camera elements and build the mount around them.

It sure is handy whenever my scope is significantly out of collimation to begin the primary mirror alignment phase with the Glatter TuBlug/Laser and get the collimation close enough that what's left for the Catseye Blackcat and Autocollimators is the final tweaking to insure that alignment is dead spot-on and not merely close.

Hint hint http://www.youtube.c...h?v=Vj12cx3tnsM

The problem is that reflections are at different virtual/real distances and it is hard to bring all to focus simultaneously.

[FirstSight]

I've watched that video before, and it's a very good one. Nevertheless, unless I overlooked something in one of the segments, nowhere does it show how you mounted the (web?)cam in position to view through the peep-hole of the cheshire or autocollimator. In order to get collimation fairly close, so long as a somewhat soft but still reasonably viewable image was presented of all involved elements, it wouldn't need to be a sharply focused image of every involved element. You'd only need that perhaps for accurate final tweaking to insure that you'd achieved dead spot-on status.

[Nils Olof Carlin]

The problem is that reflections are at different virtual/real distances and it is hard to bring all to focus simultaneously.

This is true - whether you use your eye or any camera. The spot is one focal length away, but the reflection of the Cheshire or equivalent is projected at infinity. What you can do is stop down the peephole for better depth of focus, and set the camera focus to split the difference.

The barlowed laser, in this respect, works as if having a very small "peephole".
But maybe a Krupa type collimator would be even better to mount a webcam on - you would have to make the collimator (not a major project), but you could use the webcam as is. (The photo was taken with a standard camera, not cropped but scaled down to 10%).

There is a similar problem, only twice worse, with the autocollimator. Two of the reflections appear one focal length away - the other two, not at infinity but one focal length the other side of infinity That does not prevent you from taking some very nice images, Jason...

Attached Thumbnails

• 

[BluewaterObserva]

I use oversized secondary mirrors. So exacting precise secondary alignment is not so critical at all.

I have gotten real good at tweaking on a star. But then again? Eq mounted permanently in the obs? I don't find I have to adjust collimation all to much if ever.

[Starman1]

I use oversized secondary mirrors. So exacting precise secondary alignment is not so critical at all.

I have gotten real good at tweaking on a star. But then again? Eq mounted permanently in the obs? I don't find I have to adjust collimation all to much if ever.

My own secondary is oversized to the point where the telescoping sight tube (the TeleTube) of Catseye has to be set to f/4.3 on my f/5 scope to allow me to see the outline of the secondary mirror.
When I used that setting of the TeleTube, I have the TeleTube fully inserted into the focuser, the focuser approximately in the right place for focusing eyepieces, the reflected image of the primary *just* smaller than the outline of the reflective secondary surface which is *just* smaller than the inside diameter of the front edge of the TeleTube.

When I adjust the length of the TeleTube to the f/5 length (which is longer) to match my scope, I cannot see the edge of the secondary until I pull the TeleTube several inches out of the focuser, and then the reflected image of the primary's edge cannot be seen in the secondary. And at that point the peephole is nowhere near the focal plane of the scope.

The reason I mention this? Using the TeleTube, and adjusting it thus, I can get the slightly oversized secondary mirror accurately centered under the focuser and the reflection of the primary accurately centered in the secondary. So though it may not be critical, it can still easily be done.
And since you have an EQ scope permanently mounted, you may want to do astrophotography with a large chip camera wherein that centering may become more important from the stanpoint of vignetting.

[Jason D]

whether you use your eye or any camera. The spot is one focal length away, but the reflection of the Cheshire or equivalent is projected at infinity.

That does not prevent you from taking some very nice images, Jason...

Tough but doable

Attached Thumbnails

• 

[Nils Olof Carlin]

Indeed, the centering of the secondary is not highly critical, but it is a one-time job (until you take the telescope apart next time). Theoretically, you should have the peephole a bit outside the focus, at the point where the primary and secondary look exactly the same size. This is not really critical, but it may feel good to understand why.
So, to be so critical, you can start by checking the rotation of the secondary and adjut if necessary. This done, adjust the tilt of the secondary (using a laser is easiest). Then move the focuser or/and the sight tube (teletube) to make both mirrors look the same size. This done, you can adjust the teletube to frame the secondary, but it is not really necessary - you see anyway if the secondary is centered on the primary's reflection or if you have to move the secondary, and if so what direction and at least approximately how far.

Even if you have the peephole at the focal plane, the error is totally negligible. BTW if you use the hologram centering (with the Glatter laser), you should pull the focuser out as far as possible to minimize the geometric error.

Nils Olof