Contents

• Introduction
• Initial Assembly
• Mechanics
• Optics Modelling
• First Optical Trials
• First Mounting
• Optics Testing
• Lessons Learnt
• First Light
• Collimation

Introduction

I have just completed my 12" Cassegrain ATM build so I'm fairly chuffed.

This is a diary post of photos of the build.

I must admit that the scope is built from existing optics from George Hole and Sons of Brighton. The primary is a 12.5" f/2.7 and the secondary is 92mm diameter and its CoC is unknown. I bought these second hand from a chap on Astrobuysell for £250 two years ago.

So this has been a bit of a shot in the dark since the only way to test the optics is really to make a workbench to hold the optics, which is as near as counts a telescope.

I took the idea for a truss tube scope from a chap on the internet who had built his Newtonian in a similar way.

A picture of the original telescope:

i'm fairly sure this shows a secondary mirror using a slider at the upper end on a single arm. The focuser is a fixed tube and I received the mirror mount with the mirrors and it was a 2" thick piece of wooden board with 1/2 bolts, it was massive and the eyepiece tube was fixed in the rear of this board.....

back to top

Initial Assembly

You can see the 1" diameter 14SWG aluminium pipe being used as the trusses in a hexagon straight configuration. The whole point is to allow the stiffener sections carrying the mirror, the saddle and the secondary respectively to move relative to each other so I can setup the system.

A picture showing first trial fit of the saddle parts

The saddle parts carry a dovetail either side across the scope. This one is fitted with inserts on the inside surface (to prevent pull-out) and drilled for a Losmandy dovetail. There is a second stiffener to go on the top edge which is all rebated using a router and glued only, no  screws. 

For a sense of scale, the overall length is 85cm, ID is 13" and OD 17".

Back to top

Mechanics

For the attachment of the secondary mirror I took a strip of 1mm Alloy sheet and beat a narrow lip on it around the edge off an alloy pipe section and screwed it to the wooden disk to hold the mirror itself. The sheet had a thick plastic coat on the inside to lay against the glass and the lip only lies over the unsilvered rim. I've since flocked the outside of this with flocking sheet to preent stray reflections.

The primary mirror mount is fairly standard. I used plop to calculate the dimensions for a 9-point support in three centrally-supported triangles. Each triangle rested on a round-topped coach bolt which I centre-tapped for a screw and spring washer. The triangle was screwed to the top and thread-locked, with a small bit of play availbale for tip-tilt. This resulted in low profile supports. I even had to file the head of the screws flat to get the clearance below the mirror. Each triangle point was topped with self-adhesive rubber feet whihc seem to work well.

The mount board is attached to the mount frame by a backing board. THe backing board is pierced for ventilation and provides 3-point collimation, tightening against sprung bolts through the board, using wing nuts on risers to not interfere with the adjacent set screws provided for locking collimation. The lock nuts, since they bear on a pywood disk, have acorn locknuts thread-locked on top.

A similar arrangemnt is in place for the focuser mount. The mount plate is fastened to to the forward face of the primary mirror carrier frame. I turned a flat annulus from 6mm aluminium sheet to which I screwed the primary baffle mount on the mirror side and a focuser interfacce on teh user side. The focuser is a Revelation Schmitt/Cass focuser which uses a 78mm diameter dovetail lip and clamping screw system to clamp to the interface and provide rotational adjustment. The interface piece normally threads to the back of Schmitt/Cass. I turned>the skyward face of this to hold a 63.5 mm OD Aluminium pipe as the primary baffle as an interference fit and also tapped three holes near the edge to mount it to the plate. The plate uses a 3 pairs of push-pull screws system to allow the focuser and primary baffle with it to be aligned to the secondary. The primary baffle is fitted by hand and has no further adjustment so it better be right !

I did have to come up with a fairly complex way of centering the mirror and clamping it to the backing board. This was due to narrow clearance through the spacers and the need for a strong retaining mechanism to hold an 8 Kg mirror. The clips are IKEA cupboard wall clips with felt glued to the inside lip edge. The entire L-bracket slides radially along a nylon t-bolt track I milled from Nylon stock, using a screw to pull it in toward the centre. Small nylon set screws are used to adjust the proper spacing between bracket and mirror edge so the clips don't overlap the edge of the mirror too far.

Back to top

Modelling the optics

Both baffle dimensions, inter-optics distances and tube sizes were calculated together in a spreadsheet to keep track of the Mass, Centre of mass and relative positions of components around the saddle plate. I used the goal-seek capability to adjust component distances from the saddle plate centre as the system fixed centre of mass to work out where everything needed to be and still function optically. The primary variable for this is the distance the secondary mirror lies forward of the saddle, move that mirror and you move the entire optical system with relation to the saddle. That might explain why everything optical is fixed onto the tube frame and the optics on their frames can move independently up and down the tubes or as a unit by pushing the tubes through the saddle to balance the overall CofM. The primary mirror for instance was almost inside the saddle due to its large mass.

Assembly

Back to top

Mirror frame with cutouts.

Each plywood frame was washed in thinned varnish to penetrate the wood for preservation. The mirror carrier plate was also washed in black stain on top of this varnish.

The nylon mirror clamp t-tracks were screwed directly to the wood. I could have through-bolted and may yet change this. Through bolting allows me to tighten them up after setup with the mirror in place. Screws don't. The collimation bolts are coach bolts with low profile heads, through from the front of the mirror carrier, through 8mm thread inserts, through 1" springs and finally through the carrier frame to stand-offs and wingnuts. The collimation clamp screws are simple set screws through the carrier using M4 inserts on the mirror side to prevent pull-through.

Assembly of the mirror end has been more difficult than expected. First I realised that the mirror clamps line up perfectly with the tubes, making it very difficult to tighten the radial clamp screws and mirror clamp screws. Next the focuser mount and collimation plate is very difficult to put in place with the mirror carrier in place, so to add and remove it means removing the mirror carrier. Finally the mirror had too be manouevered through the tube on to the mirror carrier since I had to tap the mirror carrier frame on with a soft mallet to get it where I wanted it and didn't want to go through the pain of removing it.

First assembly - view down the scope without mirrors onto mirror carrier and focuser plate showing tube clamps:

So, some changes to make :

1. Ensure that the mirror support triangles centres are offset from the hexagonal tube pitch to ensure that all the mirror clamps are accessible - the space betwewen the triangles is where the clamps must live, hence the relationship. This means rotating the mirror backing plate by 15 degrees.
2. Make the frame fit to the tubes more accurately. I cut and drilled all frames separately using geometry for markup. It didn't work well enough to get all holes and centres perfectly aligned. The mallet provides gentle persuasion. They need drilling all together as a single unit and all the other piercings marked out before the centres became inaccessible by removal.
3. The centres of the focuser and baffle are guaranteed to be aligned, but adjustments have had to be made to ensure that the centre of the mirror can be lined up with the centre of the baffle.
4. The lack of clearance of the mirror and carrier clamps means the mirror needed to be threaded through the tube and lowered over the baffle onto the carrier. Not a fun job.

   However its done, the clamps are on, the mirror all works and is relatively centred, roughly collimated and now the optics can be checked.

All assembled:, without secondary

Back to top

First optical trials

So I moved the secondary mirror frame towards the primary from 625mm ( modelled spacing required) to 585mm in increments of 5 mm or so until I got the image where I wanted it. A change of about 6%. This means the focal plane is now outside the focuser and eyepieces need a 40mm extension or so to come to focus on objects at infinity.

Focuser on rear frame:

My worry here is that Cass optics are designed to work at a specific spacing. I might have been able to specify the spacing had I the optical measurements for the secondary mirror. Since I didn't the next step is to measure them. Measuring a parabola or hyperbolic with convex surface without its mating surface is hard unless you directly measure the surface using Foucault through the back of the mirror. This normally requires the secondary mirror back been polished. But its not.( aside: how clear does it really need to be to do this ?) The test remaining to me then is to check the image quality with a Ronchi grating against a real or synthetic star and adjust the secondary-primary distance until the lines are as straight as they can be. If that means that the image disappears back inside the focuser then I will have to come up with a different focuser solution.
On the other hand, the view of the distant trees so far has been great, not showing any obvious distortion. I should really get a foucault-gram of the primary. I have been trying but I need to put a webcam on it to get my eye close enough.
I will mention that the mirrors also have a star etched into their sides, so I have also made sure that these line up in case of this marking the optimal rotational alignment. You never know. Another thing to check - whether rotating the secondary will improve the image by cancelling wavefront errors or make them worse...

Scope all flocked with secondary and baffles:

First Mounting

And from the user end:

I then took it into the observatory and tried it on the static mount. I found a couple of things: the Altair Losmandy dovetail bolted to the OTA saddle does not sit well on the Altair dovetail receptacle on the mount - the mounting holes in the dovetail are not counterbored deep enough for "standard" Allen cap screws. I even turned down the heads until they did lie flush and then there wasn't enough hex shape broached into the metal to take a hex key. I shall have to increase the counterbore depth on those holes. Bit of a design flaw there or I am getting something wrong ?.
It did look good though.
<no picture available>

Final design flaw.
It doesn't fit !
Even the short length of a Cassegrain is such that it interferes with the dome roof on transit and on rotation around the dome. I shall have to increase the dome height or reduce the pier height. I was so enjoying have a tall pier so I can stand observing comfortably underneath the existing Vixen 8" cass and 5" refractor combination. Its just as well I am looking for a new dome too, that way I have options to explore.

Back to top

Optics testing

Back to top

Lessons Learnt

For the flocking material I used the velvet self-adhesive material from Wilkinsons. Outsides of objects are easy and very trimmable but to get the inside layer to stick I had to roll the velvet back on itself taped to a water pipe down down the center of the roll to ensure even adhesion over the 42cm length. Still it worked and now the internal reflections are gone. The final thing will be to turn some internal baffles to prevent glancing reflections and define the image aperture. Once I have adjusted the optical model to reflect reality, I should be able to calculate the size and spacing of those using a bit of geometry. I'll just turn some tight fitting disks and bore them out to a knife edge lip.

Collimation for this beast means aligning the focuser and secondary boss together along the optical axis, aligning the secondary optics to return an image centrally to the focuser and then adjusting the primary to be normal to the secondary. Even using a synthetic star, it will need to be 80 yards or so away to achieve acceptable focus distance without extraordinary extension barrels.

Back to top

First Light

I had to build an observatory to fit mind. That is in this page...

First light was to get the scope on the mount in a vaguely balanced way and get initial collimation and focal point measurements down.

1. First result: distant trees are not at the same focus as stars.

2. 2nd result: the revelation focuser with its 50mm extension is annoying if the focuser only has 35mm of travel. ie there is 15mm for which you cant get an eyepiece in focus in either combination.

3. 3rd result. inside and outside stars were rotten. I had done an initial alignment using a Cheshire eyepiece but there was significant alteration required to get a collimated star image.The first activity was to ensure that there was a sliver of blue sky around the secondary from all vantage points in the focuser. The next was to ensure the secondary reflection still lined up concentric, the last to adjust the primary.

Finally I got round stars inside and outside of focus using a 40mm eyepiece. I seem to get full field illumination and I don't think I can go to a longer eyepiece. The field of view appears to be about 40' of arc. Plenty big for a low power full moon. So I tried it on M32 while it was up there. It completely filled the field of view with a deep core and companion and wings spilling out to either side. Very satisfying !

Last activities are getting a guide scope on the side. I need to reverse mount on the the existing Vixen dovetail there but that's not too hard. In pursuit of this I have aquired an adapter from FLO to mount to the dovetail.

I'll get some pictures to finish this off with and then describe the observatory build .

Back to top

Telescope Ronchigrams - first try

3mm inside focus on a real star

At focus on a real star

3mm outside focus on a real star

Back to top

Collimation

1. Align the optics centres quite carefully. My mirror carrier plate, as well as having its components interfere with the OTA longitudinal tubes, was not as well centered as it should be, primarily due to cutting out the centre before fully marking out the circumference leading to having to estimate the centre again later. So I have re-made this from a new piece of Ply, having spent time carefully marking out all the components before I start and checking and re-checking constantly to ensure alignment.

2. I have replaced the original Crayford focuser with a non-rotating helical focuser from Borg. This drops the weight by almost a kilo, reduces the height, provides a greater focuser extension distance and perhaps crucially, provides a direct measurement of the distance moved by the focuser in 0.1mm increments on the focuser barrel. This is important later. To do this I had to take off the mirror carrier frame to get access to the focuser mounting plate. This I replaced with a new one with a 60mm diameter x 0.75mm pitch Vixen female thread cut into it. I can now use all my Vixen parts, including extension tubes and 2" e.p. adapters. I had to cut a 57mmx0.75mm to 60mm Male thread adapter to fit the Borg focuser upside down to the plate. Fitting upside down meant the largest aperture on the focuser is available to the user and the measurement scale on the barrel circumference (since it's a helical) is more accessible. On the inside of the focuser mounting plate on the mirror side is screwed the base fitting for the primary baffle, using the end of the adapter thread screwed through the focuser plate, like a very large nut. The primary baffle pushes into this and will be glued later.

3. Re-assemble and mount on the telescope mount. Mount the secondary as far away as possible from the primary such that an image is still accessible using an eyepiece in the racked-in focuser (i.e. its not within the primary hole). Using a star and Ronchi screen, get a three-bar Ronchi image as straight as possible by moving the secondary toward the primary and re-focusing each time. For each mm of travel of the secondary, the image moves (secondary magnification squared) mm. Which seems a handy way of determining the secondary magnification if you don't know it.

4. Take out the ronchi and focus on the star to measure the focuser distance. Use an eyepiece with focal length that is close to the F-ratio of the telescope i.e. most likely a 12 or 15mm eyepiece. Rack out the focuser until the secondary obstruction is visible. Record the focuser distance for a repeatable apparent size of the secondary obstruction. Rack the same distance to the other side of focus and compare the apparent size of the secondary obstruction. It should be equal on both side of focus. If not, if its smaller on the inside of focus, increase the separation. If its smaller on the outside of focus reduce the separation.
   Mirror spacing too close - under corrected - move secondary further away
   Mirror spacing too far apart - over corrected - move secondary closer.
   Cassegrain mirror spacing Or use an oil flat to create an autocollimator arrangement that allows you to judge the best optical spacing combination using the knife-edge test. Auto collimation knife-edge test The OPD grows more or less symmetrically either side of nominal spacing, at the rate of about 0.006 waves RMS per millimeter of mirror spacing error. The corresponding P-V WFE increases at the rate of about 0.023 waves per millimeter of mirror spacing error.

Back to top

2nd light on 12.5" cassegrain.

(Note to self - Need to find foucaultgram of mirror for evidence. )

I have re-assembled this telescope having made significant changes to the entire ota. During the summer I have:

• Re-bored the holes in the skeleton frames to ensure alignment of the frame tubes by stacking the frames and re-drilling together.
• Re-made the mirror support plate with the mirror clips rotated by 30 degrees relative to the tubes so they don't interfere with the hexagonal-layout tube truss structure
I
• n re-making the mirror support plate i have also paid particular attention to keeping the central bored hole concentric with the edges and the mirror stops.
• I have re-made the mirror mount skeleton truss frame. This frame carries the focuser and the mirror mounting plate and they were not sufficiently concentric on the first attempt. The focuser plate is now threaded for 60mm Vixen threaded attachments and specifically the Borg low profile helical focuser, which I spent quite a length of time re-fitting with m48/2" threads and a 60mm male telescope-side thread.
• I have finished all the aluminium tube clamps and fitted them to the skeleton frames. They are necessary now as the frames are now loose on the tubes.

Re-assembly

I re-assembled the scope frame and mirrors and did first trials in the garage: I used a laser collimator to align the focuser on its independent mount to the centre of the secondary holder. then I fitted the secondary mirror, ensuring that the focus point for near objects was a good few inches outside the focuser, having learnt from the first assembly.

Collimation was performed using a cheshire eyepiece to ensure that the secondary reflections were concentric before adjusting the primary to re-centre.Basic adjustments completed, I mounted the scope on the observatory mount and re-collimated using the 2" auto-collimator which I find quite sensitive at indicating good secondary concentricity. The reflections of the vanes for the secondary were aligned in one axis but not the other - odd. I did a first view on Vega. It was high at the zenith and bright in the dusk sky. The out of focus rings at 40mm were asymmetric so I adjusted the primary to even them up. It required a lot more adjustment than expected to do. I changed to the 25mm Meade 4000 plossl eyepiece ( mag: 3200/25= 128 ) and nipped up the star image to be concentric either side of focus again using the primary collimation bolts. I find that I can either have the star image condense into the centre of the rings as you approach from either side of focus or you can have good concentric out of focus rings but the star centre moves that tiny but important bit off to the edge of the shrinking ring just as you come to fine focus, but not both. (This I think means something is not quite aligned but I can't see what). I moved to Deneb, for mount alignment purposes. Smack in the middle and no real change in the image due to movement of the OTA. I moved to the central star (Gamma?) and then to beta to view the double. The double was wide open with lovely bright yellow and blue colour but the image was soft. A bit cometary or squashed spider even. The legs of the spider wafted as I watched and no amount of focusing would get crystal round stars. I then moved to the double double E Lyra. There it was, again, E1 and E2 wide open with clearly discernable doubles but the close doubles were not clean; they were fuzzy, even using the 9mm eyepiece (3120/9 ~ x345) Finally, making use of what dark sky was left in the face of the rising full moon I moved over to M57, M27, M29 and then closed up, leaving the scope pointed at 10 deg alt and 90 deg az.

Situation:

The focuser works fine. Visual focus needs to be moved to the full extended length of the focuser to enable mounting a DSLR at focus when drawn fully in. So the secondary needs to come towards the primary by about 20/sqrt(M-1). The primary baffle is too long. I need to determine the required length compared to actual. Currently it's 37 cm from the front of the mirror. I cannot see all the secondary from the inside edges of the focuser, or even a little light around the edges. Holding of collimation between slews looks good but the initial collimation change i had to make might be due to the swing from horizontal to vertical - I shall have to look into that and mount the mirror with the tube vertical. Underlying collimation looks flawed. I need to resolve why the vanes aren't symmetric in the reflections ( maybe they aren't mounted symmetrically - is this really something to worry about ?) and also why an auto-collimator collimation had to be adjusted so far on a star. Finally, why does the star collimation do that trick of disppearing into the edge of the rings as I come to fine focus and how do I go about tracking down the cause ? Will that solve my soft and squidgy stars issues ? I should note that the background stars are tight and round more or less to the edge of the field in the 25mm eyepiece and its the brighter stars that are soft and troubling me. Having seen a foucaultgram of the primary which to me looked fine even for such a fast primary (f/2.7), I wonder if its roughness I am seeing in the mirrors. I have no way of testing the secondary other than by making a matching nulling mirror. TBH I can't see that happening. I shall try to collect some video evidence next time, now I have a first view. I shall also try to use the gauge on the helical focuser to judge the symmetry of the out of focus rings on either side of focus for the purpose of spherical aberration measurement and validating that the secondary is at the designed separation. Its about the only means I have for doing this test.

Cloudy nights input:

Assuming you know system's effective focal ratio and primary's focal ratio, it gives you secondary magnification "m". If you can measure secondary's r.o.c. R2, then you have all you need to get the system parameters. If R2/R1-r (R1 being the primary's r.o.c.), then the mirror spacing in units of primary's f.l. is

 s=1-(m-1)r/m 

 b=(m+1)(1-s)-1 

 W=[1-m^2-(K2)(m-1)^3]d/512(F1)F^3 

Update - third pass at this using DPAC testing

The next step was creating the joint light source/ronchi grating eyepiece. This was easy enough - I used an off-axis guider with the light source coming through the pick-off prism and the grating across the aperture. While this worked for a small scope, the angle between the prism source and the drect on-axis viewing line was felt to be too much and difficult to get the telescope aligned to get the reflections back. So I moved to a turned piece of nylon and fitted a single PWM controlled LED light source inside it, much closer to the viewing line. This works a treat.

Having a working ronchi/light source eyepiece and proved the liquid mirror concept worked, I moved to mounting the telescope inverted on a box above the mirror. Of course this means I need a box with a 113" hole in it and a tip-tilt-plate with a similar size hole in it to align the inverted scope with the liquid mirror. Next job... Good job I have wood, handwheels and a router with transom bar for cutting large circles. Another day later.

the imageshows the eyepiece with a small usb camera taped to the eye hole for capturing images of the ronchi patterns.

I use the same camera on my foucault tester. Its essentially the guts of a MS Lifecam 3000

Finally, with telescope inverted on the tip-tilt plate, on the box above the liquid mirror I could begin Ronchi double-pass testing. This is suposed to be a really sensitive test of mirror optics because the light passes through the optics twice. Noting that securing the primary mirror ready for telescope inversion is scary of itself, the first job was to colllimate the scope upside down. The laser had already been used for centering the secondary mount compared to the primary mount and focuser. Centering of the primary mirror was done by eye. I suspect a custom tool should be made for this for the future to remove faffing with clearance bars between the focuser/baffle mount and the mirror central hole.

The black knobs on the lower wooden plate are 2 from a set of three, screwing these in and out aligns the scope with the oil batch underneath.

The box underneath that is a simple rectangle with the top surface having a large diameter hole routed in it, the size of the largest aperture. I propped up a board in front of the end opening , as seen here, in order to prevent air currents across the surface of the oil disturbing the view.

At last I had fringes. One of the changes I had made was to remove the large focuser requiring a large back-focus distance and replaced it with an eyepiece mount that srews directly into the vixen thread on the focuser plate. That removes about 80mm or so back-focus requirement but makes consistent focusing harder without a movement mechanism to move the eyepiecce back and forth. So I used the Baader micro-focusing 1.25" eyepiece mount to hold the ronchi eyepiece.

The photo below is a montage of the ronchi patterns obtained by moving the focuser through inside focus to outside focus. Each re-focus required the tip-tilt plate to be slightly adjusted to best centre the fringes. Each image was taken with a stripped-down board camera focused at infinity, taped over the eye-hole in the ronchi eyepiece.

What this shows is what the original star-test showed, the inter-mirror separation is not enough, indicated by the curved lines )( inside focus and there are concentric zones. What it doesn't indicate is which mirror has the zones. I suspect the secondary and the only way to tell is to go back and get a foucaultgram of the primary to rule it out. Once I know which of these it is, I shall make a mirror box and send it off to John Nichol to re-figure.