Potential build - Cassegrain or Newt?

[Sfarndell]

While figuring an 18"x25mm blank (F/4.1), I came across a bargain on a 16" x 40mm (F/4.5) completed mirror with an 18 pt mirror cell. As a result, the 18" was deprioritised for a long time while I built a 16" newtonian.

Now that the 16" is complete and working well, my eye is wandering back to the 18". The only benefit to me would be to make a more compact scope. Plugging in the details into myCassegrain, to get a reasonably sized secondary of +-200mm (8"), I need to make the scope about 1.7m long (1.5m between secondary and system focus). This is almost the same as for a Newt.

Hopefully, I just entered the equations wrong, but if not then apart from the fun challenge of building a cassegrain, are there any benefits to doing it vs building another truss newt? (I'm 6"3 so height of eyepiece not really an issue)

Any help/thoughts appreciated

Scott

[Sfarndell]

...meant a 6" secondary (157mm)...!!

[cassiewoofer]

That sounds too long to me, the focal length will be 18" times 4 or 6ft.....or 2 metres, making your scope less than a metre in total with that folded light path. It will also depend on the curvature of your secondary, are you going to make this or you thinking of a flat secondary?

[Glasspusher]

In a Cassegrain telescope the convex secondary mirror is placed inside the focus of the primary mirror, so this should give some idea of the overall length of the tube. In the Dall Kirkham configuration the secondary is a convex spheroid which is not too difficult to make, the matching primary is an ellipsoid or under corrected paraboloid, again not too difficult to make. The problem with the DK is that it suffers badly from off axis coma and the best designs are approaching F20, fine for planetary use, not so good for wide field work. The design can be improved by the addition of corrector lenses which greatly reduce the off axis coma, this is the Optimised (or modified) Dall Kirkham.

The traditional Cassegrain has better coma correction and uses a paraboloidal primary mirror and a convex hyperbolic secondary which is not easy to make.

The Richey Chretien design was developed in the days when positions of stars where measured on photographic plates so round images were important right to the edge of the field. This design is well corrected for off axis coma, this comes at the price of a hyperbolic primary and secondary. Both of these surfaces are for master opticians only.

John

[Astrobits]

The Cassegrain secondary is usually placed somewhere around 20-25% inside the primary focus depending on the amplification required. So the length is not too different from a Newt but much different in F ratio.

John has described the different types very well. There is an addition he did not mention and that is the Naysmith design. This uses a conventional Cassegrain plus a small flat in front of the primary to bring the final focus to the side of the Optical Tube, possibly through the Alt/Dec axis. This suits an Alt/Az mount very well as the eyepiece remains at a set height and substantial equipment can be mounted there without the usual balance problems. It also means that the primary does not need to be perforated.

Nigel

[Astrobits]

Just a thought. As you are in Johannesburg have you examined the 12" Cassegrain in the Johannesburg centre's dome ( I assume that it is still there )? It might give you some idea of the pros and cons of the design vs a newt.

Nigel

[Sfarndell]

Thanks all for the suggestions and feedback! I think it essentially boils down to a) having a slower scope for planetary/small nebula viewing vs another big deep-sky scope (i.e. no duplication) and the satisfaction of making and building a more challenging scope which is a big draw for me.

In my research, a Cassegrain/Nasmyth did cross my mind after coming across Mike Lockwood's 12" as I am worried about accuracy/breaking my mirror when coring it. Plus, if the scope is going to be the almost same length as a Newt, I would need a monster EQ/fork mount with resultant eyepiece placement issues.

Re: the Dall-Kirkham/Classic Cassegrain option...I intend to make my own secondary so the hyperbolic secondary seems much easier than making a 10-12cm flat anyway so a Classic Cassegrain would be my choice for the better performance over an uncorrected DK. The extra work/difficulty is probably worth it otherwise I'll most likely regret taking the short cut as my experience grows. I have no idea how easy it is to make corrector lenses (perhaps regrind a binocular lens??) which would also be an option, but that or a Ritchey Chretien is likely way beyond me for now - the 18" is my 1st attempt at making a mirror so far so I don't want to overextend too far just yet.

Nigel - great suggestion about the 12" at the Jhb Observatory.. it was mentioned at the last meeting on the 14th, but I don't recall the speaker saying if it was there or not. Worth asking though.

I'll ponder a bit more until the primary is parabolised, but I'm definitely leaning more towards a Classic-Cass-Nasmyth just for the fun of the challenge.

Scott

[FraserClarke]

You may also want to consider a Gregorian, which has a concave secondary mirror (placed after the primary focus). It's longer, but has the advantage that the secondary is a lot easier to test.

You can use stock lenses for a corrected Dall-Kirkham by tuning the curvature of the primary to match the lenses. Suitable stock lenses typically cost a few hundred pounds from an optics company. You need a bit of optical design software to then workout the correct parameters for the primary -- but that's probably just a case of posting a message on here Corrected Dall-Kirkhams typically give you better widefield performance than a classical cassegrain.

[Glasspusher]

You can use stock lenses for a corrected Dall-Kirkham by tuning the curvature of the primary to match the lenses. Suitable stock lenses typically cost a few hundred pounds from an optics company. You need a bit of optical design software to then workout the correct parameters for the primary -- but that's probably just a case of posting a message on here Corrected Dall-Kirkhams typically give you better widefield performance than a classical cassegrain.

Good call Fraser, but bear in mind that the lenses have to be positioned VERY accurately in the optical train demanding a very well engineered OTA.

John

[Sfarndell]

Having done some more research it also looks like the corrector lenses have very small tolerances in their figure, as well as position and are also hard to test.

After continued pondering, the classic-Cass-Nasmyth stil appeals, but I've got some time to think: I'm currently still rebuilding a test stand and foucault/ronchi tester so *if* I haven't reached a parabola before the little voice of unreason (egged on by the enthusiasm of the old regulars at my local ATM club) convinces me that a RC is worth the extra pain of figuring the primary, then the final scope might end up as a RC-Nasmyth. Or an attempt at a RC-Nasmyth ...

[Glasspusher]

Scott, presumably you have considered test methods for hyperbolic surfaces (concave and convex) if you attempt a RC. I would be interested to know which you intend to use for your project? Thanks, John.

[Sfarndell]

John - the only methods that I have come across are using interferometry. There are a number of sites I looked at, but Mike Lockwood's build is the most explicit: http://www.cloudynights.com/item.php?item_id=1127

I'm sure you know the details better than me, but it would involve generating a concave test surface by using a glass tool to grind the secondary, polishing tool/mirror to a sphere at the same time, then figuring the tool to the correct concave surface followed by figuring the mirror to the correct convex surface and testing using an interferometer (which I understand the local ATM school has).

If you were referring to the Primary hyperbolic, I was going to use the traditional a concave test for a parabola - foucault tester with couder mask and a high number of zones - and map the surface until I reached any suitable hyperbolic surface then work out the details for the secondary before grinding. Do you have any other/better suggestions?

I'm still finding my way, but am I correct in assuming that the only real challenge between an RC and a Classic Cas is that 1) figuring the RC primary is more difficult and 2) tolerances on the RC mirror-separation are tighter for a given focal length of a finished primary? From my understanding, as both types require a hyperbolic secondary which is figured for a certain mirror separation, the dificulty in making the secondaries shouldn't differ by much.

[Glasspusher]

Hi Scott,

Thanks for your comments, very interesting. I think the methods that you describe would work OK. I think the route that I would take would be to take the primary or secondary to completion and then figure the other element against it using a flat in an autocollimation set-up. The difficult thing is to get one of the elements correct in the first place. I would be tempted to figure the primary using a Ross Null lens set-up; the tolerances are very tight indeed in terms of spacings between the mirror and Ross lens.

You will be aware that there is very little on the web about making RC optics, this it because it is extremely difficult to do and very few people have done it successfully. Also bear in mind that collimation is difficult and the centres of both optical elements need to be exactly on the optical axis demanding a very well engineered OTA.

A traditional cassegrain primary has the enormous advantage of being parabolic, it can be Null tested against a flat mirror. The hyperbolic secondary can then be figured in an autocollimation set-up using the primary and a flat mirror.

Looking forward with great interest to hearing how you progress with this project!

John

[Sfarndell]

Thanks John - some more excellent advice.

I do get nervous when experts use the words "extremely difficult" . I've never paid attention to null-testing until now (lack of experience/understanding) and have only ever used foucault/ronchi and star tests. Would the Ross test work on a RC? I've done some more reading this afternoon and it does look useful with the foucault as check. It also seems that any lens can be used as long as the ROC of both surfaces as well as centre thickness are known (and using Cervalo formulas to determine distance).

Regarding the optical alignment, that seems as difficult for a RC as for a Classic Cass.

[Sfarndell]

Lol...that smiley was meant to be a " "

[Glasspusher]

Scott, I did run some figures through for Null testing an RC primary using a 100mm diameter lens (made by Ceravolo) and the figures gave very little latitude for spacing errors which was a cause for concern. From memory that was for an F8 system, things get a little easier for F10 and longer. Regarding the Ross lens the surfaces must be high quality and the chances of picking up a surplus lens that is good are small. When testing an RC primary most of Ross lens will be used which is not the case with longer FL paraboloids in which case a carefully selected surplus lens can be used as the middle area is likely to be of good quality. Yes, you are right the same alignment issues apply to RC and traditional Cassegrains due to there surfaces having an axis of rotation. This is not the case with a DK secondary, which is spherical, as a result alignment is less critical. John.

[Glasspusher]

After a little more work it would appear that a Ross Null test on the primary is not really viable due to spacing issues. However, the glass disk used to grind the secondary could be used as an interferometric test plate. This concave disk is figured to the correct hyperbolic figure using a Ross null lens, the convex secondary mirror is then figured and tested against this reference (under monochromatic light) until straight fringes are seen. The next logical step is to figure the primary against the finished secondary using an autocollimation set-up. If a large flat mirror is not available for this set up a bright star could be used to star test the optics. This means that the OTA has to be made in advance.

John

[Sfarndell]

I ran some numbers too and with a radius of curvature of 3.7m I wasn't sure how I was going to measure the mirror-lens distance to the required accuracy on the primary !

Building the mount first and making the secondary first should work out nicely! The mount needs to be built before finishing the mirror set anyway to get the COG/axes correct for accurate eyepiece placement. I have a spare unused Meade LX200 goto 10" mount which I'd like to modify to accommodate this scope (at least, utilise the electronics for GoTo) so if I can't get the focus to go through the Alt axis, it'd have to be along the Az axis on the top of the mirror box. Having a 3rd mirror may complicate matters somewhat, but the testing could be done with the mirrors in a traditional cassegrain setup first after coring the primary (provided of course that the final focus point is behind the primary).

I'm off to the ATM class tomorrow afternoon to order the secondary blanks and more pitch and will start calculating from there as the final F-ratio will depend to some extent on the size of the secondary available. Hopefully they have a suitable lens/ross-null-tester already otherwise I'll have to make another plan.

Progress so far is that the new mirror testing stand is complete. Over the weekend the plan is to re-do my foucault tester so that a webcam can be mounted which will hopefuly help with improved accuracy on testing the mirror (less shaking) as well as for documenting the progress.

[Sfarndell]

The class was really interesting. The guys who started it 20 years ago have just about everything you'd need including lathes, artificial stars and interferometers. Just not a 19" flat!

The ATM class's blanks for the secondary are also a bit thick and heavy at 25mm so I scouted around and was given 2 pieces of plate glass 300mm x 300mm x 12mm by a local glass-maufacturer from which to cut out the secondary and tool. This might be on the thin side for a 6" mirror, but with a 22mm thick (after all the grinding) primary, weight at the cage end needs to be kept to a minimum. It'll take a few weeks to get around to making the coring tool and cutting the blanks, but I'll post regular updates.

Almost finished the Foucault tester using the Stellafane templates - just need to fix the camera. I retested my primary and the Focal Length is closer to 2.0 metres (F4.37) . There was a minor amount of stand-induced astigmatism so I'll need to make an 18 point mirror cell before I get to the final figuring.

I've also been investigating carbon fibre from which to make the truss tubes and secondary ring which might help with the weight, at extra cost. Does anyone have input on using CF for a secondary cage/holder?

Scott

[Sfarndell]

Some progress has been made...sort of. My mirror picked up a number of small scratches after 4 years in the cupboard and a poorly made lap so I had to return to fine grinding with 400 grit. This wasn't too bad as had previously ignored a deep scratch from before so it was an opportunity to get the mirror into a better state. I also took the opportunity to core the back of the mirror to within 1mm of the face, which I filled with cement and sealed with a 2mm layer of fibreglass resin.

The guru who opperated the machine accidentally punched through on 1/4 of the circumference so I sealed that hole with resin too then used a razor to shave off most of the "bump" and ground the residual to the curve of the rest as part of normal grinding. So far there have been no issues.

After attempting to finish the fine grinding on 1000 grit and picking up some deep scratches...twice, I've given up on this grit size and will finish on 800 grit later today (1 more session should do it!). I've also cut out the secondary mirror & tool and have hogged out to 80% of the sagitta. I'll wait on this until the primary is polished enough to see if the focal length has changed  

here are some photos of the mirrors as they currently stand