Arie's Dobsonian Telescopes

Miscollimation to merge images

One of the most dreaded aspects of building a binoscope is that the two independent optical systems have to produce two images that are well collimated and accurately merged at the same time. I tried different procedures, such as moving the primary mirrors laterally on their mirror cells and simultaneous collimation by traditional tilting of the primary mirrors. This worked, but I found it rather tedious to perform this procedure in the dark. Also, by definition it moves the primary mirror away from its ideal position on the six resting pads as was calculated by PLOP. Of course, these resting pads can be moved too, but it doesn't make the procedure more elegant! Hence I kept trying to find another way that is more simple. I found a merging/ co-collimation procedure that is a rather unexpected modification of the normal collimation procedure. And here it is.

Step 1. Separately collimate each of the two telescopes as is normally done with a Newtonian telescope
Hereby I REMOVE the diagonals containing the tertiary mirrors.
1) collimate the secondary mirror, for instance by aiming a single beam laser at the center of the primary mirror and
2) then collimate the primary mirror by tilting it. I collimate the primary mirrors with a Howie Glatter Barlowed laser. Collimation by tilting the primary mirrors is done "from above" as I described before. Long aluminium rods with handgrips perform the task of turning the collimation bolts (picture below).

Step 2. Merge the star images by tilting the primary mirrors.
Re-insert the two diagonals with tertiary mirrors and eyepieces. The images of any star or starfield can very simply be merged by adjusting the collimation bolts for the primary mirrors. That is: I slightly turn the long collimation rods that tilt the right and left mirrors untill the images are merged (picture below). The grips of the aluminium rods are close to the observers shoulders for easy operations. There is a catch in in the order how to move the respective mirrors, but that will be discussed below. This merging is naturally being done while I look through both eyepieces.

The images will now be fused, but the tilting of the mirrors inevitably leads to de-collimation of both primary mirrors! This can easily be monitored with a single beam laser that is placed in the so-called Howie Glatter tuBLUG (picture below).

The movement of the "Barlow donut", indicating de-collimation, will not be as extreme as shown below, but this is to explain the principle. The de-collimation of the primary mirrors will result in unsharp star images, so in a next step these sharp images need to be recovered.

Step 3. Re-collimate the primary mirrors.
It turns out that an effective way to re-collimate the two primary mirrors is to simply adjust the screws on the secondary mirror holders that are normally used to collimate the secondary mirror (picture below).

Step 4. Collimate the secondary/tertiairy mirrors
Hence the fourth and last step is to re-collimate the secondary/ tertiary mirror systems. This can be monitored by inserting the single beam laser directly in the focuser (picture below). The laserlight will miss the centers of the primary mirrors somewhat. I re-aim the single laser beam back to the centers of the primary mirrors, by adjusting the tertiairy mirrors. This is executed by a push-pull system on the plates on which the tertiary mirrors rest and that moves the tertiairy mirrors (picture below).

I now insert the eyepieces again and look at the stars. They probably won't be merged completely, but pretty close! I therefore repeat steps 2 through 4 at most one or two times.
By following this strict order, star images will now be both well-collimated and accurately merged.

Different collimation bolt positions

Note that the two used collimation bolts of the two mirror cells (see the black arrows in picture below) are not placed at the same positions. The right mirror is collimated by two bolts closest to the mirror support, the left mirror is collimated by one bolt that has been placed between the mirror support, and another bolt at the left upper side of the mirror cell.

What does this imply? By turning only one of the two bolts of the RIGHT mirror cell, the right mirror will be tilted in a 'left-right' fashion, parallel to the long side of the mirror box. In contrast, by turning the one bolt of the LEFT mirror cell that is placed between the mirror support, the mirror will be tilted in an 'up-down' fashion, square to the long side of the mirror box. The large red arrows show these respective movements of both mirror cells that in principle must be perpendicular to each other. Why this seemingly complex procedure.

The answer lies in the way image merging is done (actually HAS to be done) in practice. With only two collimation bolts, placed at the bottom of the right mirror cell, the images can in principle be merged. Turning both collimation bolts simultaneously in the same direction, will move the images vertically up or down. But in the end, with only these two collimation bolts available, the images will have to be merged in a horizontal movement. I noted that when the images come close enough together, the brain will interfere and snap the images together instantaneously. This while you know that 'mechanically' they were not fused yet by the turning of the collimation bolts. It may become a cause of constant eye strain, and a bad headache as the end result of a night observing!

Now this can be avoided by independently tilting the left mirror to move the left image vertically up and down.
The merging procedure becomes as follows:
1) keep the vertical axis of the images slightly misaligned on purpose (A), by turning the left mirror cell collimation bolt
2) now merge the images horizontally with one of the two right mirror cell collimation bolts (B)
3) when you see that horizontal convergence of the images has been reached (C), finalize by vertical merging (D).

Why does this work? The brain can only snap together close images when they lie on the same axis as the eyes: horizontal. Not vertical misaligned images!!! So by purposely keeping the images vertically misaligned at first, you can now see where the 'real' horizontal merging point is, without the interference of the brain.

David Moorhouse explains the above with fewer words and he solved the problem with three radio-controlled servomotors that can turn all three collimation bolts on one mirror cell. Hereby he achieves horizontal and vertical merging of the images by tilting one mirror only. Unfortunately this elegant solution is beyond my technical capabilities. And it is obvious that I can not use a long rod to turn the third collimation bolt on the right mirror cell, since that rod would be placed behind the entire binoscope. Hence I tilt the two mirrors independently, which explains the need for making two different mirror cells, on which the collimation bolts are placed differentially. But it leads to the same result. And at least my solution has the virtue of simplicity, that is, it is easier to built.....

How does this image merging hold up in practice when cruising the sky? Not perfect, but pretty stable. Tiny adjustments have to be made now and then in different parts of the sky, to keep the images merged, but with the long rods this is extremely easy and far from bothersome. It's comparable with the constant tweaking of the focusers when moving to other objects. As the title of this section suggests, the constant tweeking of the collimation bolts inevitably leads to a slight miscollimation of the primary mirrors. But this miscollimation is so tiny that you hardly see it with the tuBLUG, or more importantly, on the images in the eyepiece! Of course this may become a problem though with more demanding optical systems, faster than f/5.0-f/4.5. And this is another reason to stick to a relatively slow f/5.0 system.

So in the end the part that I dreaded the most is not so bad. The precise construction from the beginning was essential though. That led to almost well-merged pictures from the start and made the rest easy. So, as a concluding remark, the ´difficult´ co-collimation and merging steps shouldn´t keep people away from considering building a binoscope.....

next page