Prime focus camera newt

[furrysocks2]

I currently have a heavy diy dob (link in signature) which I intend to motorise but unsure I want to go to the effort of remounting as an eq it given its mass and balance.

I do however have a second primary mirror sitting idle (8.5" f/7.6). Early on, I'd been thinking binoscope but the level of engineering and cost of optics put me off.

I am interested in the idea of video astronomy and modest imaging. I'd prefer not to buy a DSLR and all the associated kit. With three young kids and the Scottish weather, observing and processing time is limited. Seeing more than you can through an EP while not testing anyone's patience feels like the right way for me to take this hobby. A bit of science on the side would be fun (occultations, variables, spectroscopy, for example)... and of course the moon - an easy target!

I nearly bought a bare 2nd hand 100mm f/5 skywatcher frac last week for a good price - shout if, given what I'm saying, you think I still should!

What I'm considering though is an ultralight, low CoG imaging newt with a camera at prime focus (no secondary). Ability to focus, focal reducer, possibly a filter drawer, etc, shouldn't weigh all that much, nor block any more light than a secondary.

Very keen to hear if anyone has any thoughts or reactions to this, or can help me to set expectations. FoV and camera sensitivity probably the limiting factors, along with stability and accuracy. Or do I get the 100mm frac?

 

Regards,

Matt

 

[hughgilhespie]

Hi Matt,

Nice idea! My main concern would be how to acheive and maintain focus. Choices are move the camera along the optical axis or move the main mirror up and down. Both seem difficult to do whilst retaining collimation (actually not quite so difficult as you won't have to worry about aligning the secondary.)

If it was me, I think I would go for a fixed camera position, easier to have it mounted rigidly like that, and think about a way to move the main mirror up and down the OA. It would only need to be a few millimetres travel as you don't have to worry about changing eyepieces, people who wear glasses, etc.

I think that you would likely get some coma - unless you somehow fix a corrector in front of the camera, but for basic video astronomy it shouldn't matter too much. In fact if you use a camera with a fairly small sensor and centred on the OA, you shouldn't see much coma. You will waste a lot of your photons though. Swings and roundabouts!

Looking forward to hearing more about this.

Regards, Hugh

 

[Dave_D]

could you not build a spider around a 2" helical focuser?

[furrysocks2]

Thanks, Dave and Hugh.

I rather like the idea of both - having a movable primary to account for gross setup changes and a non-rotating helical microfocuser at the top end, as well as adjustable truss rods to mount a secondary for visual.

 

At this point I'm still unsure exactly whether my focal ratio, sensor options and other hardware result in a "useful" aperture and FOV. I can't see that there's anything wrong with the idea in principle and I assume that however it ends up performing, it's an idea worth pursuing. I know this type of thing is not new and has been discussed on here and elsewhere before but I'm yet to sit and have a good read up.

Sensor size will likely be limited by budget - if I consider 1/1.8" as the maximum, can I get away with 1.25" optics? I'm almost certainly planning a RunCam Owl Plus, though waiting to hear a guy at the IOTA's review of the Night Eagle. In addition, likely a board cam or a cased equivalent (I know, I know, but...), or perhaps a second hand ATIK, ZWO, etc. I've got a webcam that should do for first light. I'd prefer not to have to learn too much about how to mod anything (manual gain, etc), but I suppose that's all part of the fun.

Ideally, it would hold collimation to within some tolerance through its range of motion. Should I consider or can I discount a wire spider (below, I think I've worked out an absolute maximum sensor load of 1kg)? I don't yet know how to calculate the position of the focal plane for the various configurations of the above and where that would position the CoG of the optical train relative to the spider support points - is that what a "moment of force" is? I didn't do physics past high school. It would be rather convenient if thumbscrews at the support points of the wire spider itself could be used for collimation - I've never looked into wire spiders. Otherwise, it will have to be something more rigid.

For the truss rods, I've hacked about with extendable washing line props for a few things - 4 foot twin section, painted steel tubes in various colours, diameters and styles of clamp, depending which supermarket you go to. Typically only about £3 each - the "green" ones are currently top of my list. One of the nice things in my mind is that with some careful measuring and marking or alignment holes, it should be possible to readily adjust the rods to accommodate either the sensor assembly or the secondary+focuser at the top end. One of the threads I glanced at earlier mentioned thermal expansion of truss rod material and steel may well be the worst choice - but anything else I choose would increase the cost significantly. I'll probably begin with those and determine if rigidity will be adequate (£25 down if not) and even though it might also imply additional weight behind the primary, I'd appreciate any comments on this if there are any - perhaps I'm being a little optimistic.

 

My spare primary mirror is:

• 216.5mm x 1620mm (f/7.48)
• 3.3kg

Top end weight - perhaps something like 250g min, 500g typical, 1kg max for the sensor assembly (?), plus top ring:

• 100-150g for a micro-focuser
• 100g for filter/drawer
• 100g for heatsink and fan
• up to 300g for the camera
• up to 150g for barlow/reducer/corrector
• collimation adjustment
• adapters, extensions, parfocalising rings, etc.

My current thoughts on the general design:

• collapsible/portable
• length adjustable to facilitate secondary and focuser
• position adjustable primary and microfocuser at sensor
• manual collimation, focusing and filter changes
• camera rotation independent of collimation or focusing
• rotary encoder digital setting circles and stepper drive on both axes
• horizontal alignment (alt-az) and polar alignment (eq) mount configurations
• continuous reach of the full sky

Budget... 

• first couple of sensors - £100-£200 (significantly more as and when)
• other top end hardware - min £100 (perhaps double that, but no more)
• motors, setting circles and control bits - £150 (I already have most of this) - £50 extras?
• other materials £50-£250
• mirror re-coating £100
• plus beer and mistakes.

 

An enjoyable thought exercise if nothing else, however it's an awful shame to keep 8" of glass in the dark under the workbench when it could be collecting starlight, populating a screen for all to enjoy and doing "science". 

Although it does make me wonder why I didn't just go and buy a 12" on day one instead of spending more to build two 8s. :/

 

 

[AKB]

Something like a SX Lodestar or Ultrastar would perhaps work for this?  It's the right size, being 1.25" diameter and no cooling required. Also the right weight at under 100g and just a single USB cable for signal and power.  It's a bit pricier than you're perhaps thinking, but maybe a second hand one?

[furrysocks2]

Just now, AKB said:

Something like a SX Lodestar or Ultrastar would perhaps work for this?  It's the right size, being 1.25" diameter and no cooling required. Also the right weight at under 100g and just a single USB cable for signal and power.  It's a bit pricier than you're perhaps thinking, but maybe a second hand one?

Thanks. Yes, I've seen them mentioned a fair bit.

[hughgilhespie]

Hi Matt,

For your spider design I suggest that you don't go with a wire design. Much better to throw off the shackles of conventional Newt design and go for a really heavy duty jobby. Yes - in principle you want to minimise the cross-sectional area of the spider to minimise the light loss but in doing that it is really easy to end up with a telescope that is pretty much unuseable due to the secondary mirror (or camera in your case) wobbling about.

I would look at a design that starts from a 10 mm thick circular aluminium plate with 3 or 4 segments cut out so that your spider vanes are about 10 mm WIDE - Shock horror!!! But - some numbers:

Area of main mirror       36644 mm2

Area of 50 mm diameter 'camera plate'  1964 mm2

Area of 3 off 10 mm wide vanes 2490 mm2

Total obstruction, by area, about 12.2%.

I think you could probably get a suitable plate fabricated by water jet cutting without too much trouble. Then you have a good, stable base to attach you camera and focusser.

Regards, Hugh

[furrysocks2]

So a single piece, central disk and vanes together?

A hole in the center perhaps, with collimation adjustment outside that but within the 50mm?

It may also eliminate need for a top ring if the truss rods mounted directly to the vanes, or do you think the additional rigidity afforded by having a solid perimeter would be better? It could still be done as one piece. 

[hughgilhespie]

Yes, a single piece with central disk and vanes all in one. Why not 'extend' the vanes past the outer circumference to make mounting points for the truss rods? If you are worried about flexure, use thicker plate, maybe 19 mm? I think you would have a really neat design where the 'spider' becomes the whole of the top of the Newt.

Regards, Hugh

[furrysocks2]

I can see that, I like the idea.

Found this for £16, could get an offcut locally, otherwise shouldn't be more than £20-£30, plus cutting.

With a continuous circumference, around 350g I think.

[furrysocks2]

I can't quite get my head around light angles and such - if the front aperture was exactly the same diameter as the mirror, does that mean I get 100% illumination only on a single point because anything to the sides of that only a decreasing portion of the mirror can see? Do I need to first work out my target FOV and increase the front aperture accordingly - ie so the left hand edge of the mirror can just see an item 0.5xFOV to the left of center? Calculating this for a 2 degree FOV I think I get 28mm all around but perhaps I'm just not understanding yet.

 

[hughgilhespie]

Hi Matt

Don't know if you have used this software but it's the bees knees for Newt design.

https://stellafane.org/tm/newt-web/newt-web.html

Should answer all your questions.

Regards, Hugh

 

[furrysocks2]

Thanks again, Hugh.

I did use it for my last scope, but because I'm not using a secondary and can't control how it sizes the baffles, it's been confusing me a little but I think it confirms what I was thinking about aperture size and 100% illumination.

 

Fat fingering it tells me 1 degree FOV is equivalent to about 28.2mm illumination size and comparing that with the 1/8 degree FOV I think I get with my small webcam on the other scope (very similar focal length), that would suggest it is a 1/4" sensor which I believe it is.

If a 1/1.8" sensor is 7.18mm x 5.32mm, then without a focal reducer I imagine I would get a FOV of 0.25 x 0.19 degrees - double the dimensions of my webcam. I'm assuming a 0.5x focal reducer would give me more - double that? 0.5 x 0.38? Should I therefore be try to achieve a 100% illuminated area of at least 0.5 degrees? I think that means 'd only need 7mm or 8mm extra front end aperture.

Doesn't the resultant FOV with a reducer also depend how far from the sensor the focal reducer sits? I assume one should try to keep it close to the 0.5x, as that's the scale it was designed for? If I've got it more or less right, is 0.5 degree FOV enough for what I want to do, and should I expect to start to see coma at this?

I can drill holes and make sawdust but I'm a bit out of my depth with optics - any pointers would be appreciated.

 

[furrysocks2]

I've been looking into whether the diffraction spikes introduced by the spider would be a problem.

10mm vanes would give a vane thickness of 4.6% of diameter of aperture.

I found this page which simulates diffraction patterns for a few vane types, including 3 vanes of 5% of aperture. A couple of screenshots from that page (credit to the author)

 

Personally, I don't think I mind seeing spikes. If I'm showing video to bystanders, I don't think they'll mind either. However I wonder if it might make certain analyses trickier, like measuring magnitudes or spectra, or just be a bit sub-optimal if I do lose contrast, planet surface detail, splitting binaries, etc.

I also found a page here discussing a 3d printed spider and the disadvantages of that. In particular, the benefit of tensioned vanes, where a single piece spider would be relaxed. Also a warning in reference to adjustments to center the camera on the optical axis being important in truss tubes, as opposed to rigid tubes.

It's all a bit of a stab in the dark - I love the simplicity of the idea, though.

[hughgilhespie]

Hi Matt,

Can't help you with the optical calcs I'm afraid. Everything I have read is about calculating the fully illuminated circle when you are using a secondary mirror. The site you linked to on diffraction patterns with thick vanes is very interesting. I quote

Even really massive supporting "vanes", that reach into the optical path, have very little influence on planetary conrast. The geometry of Figure 6a is often found in lightweight (robotic) telescopes, that carry an electronic camera instead of a secondary mirror. The camera (or the secondary optical component) is supported by a tripod that reaches directly from the cell of the main mirror to the first focus. The geometry of Fig. 7a would be for a tetrapod respectively.
Following the simulations, a rule of thumb for the vanes would be: a three-vane spider with vanes with diameter under 5% of the linear diameter of the mirror is uncritical for the contrast in planetary observations - that is a much higher value than normally

I assume that if you are doing some sort of video astronomy, you will be limited to fairly high brightness targets, so the extra diffraction pattern for high magnitude targets shouldn't be a problem.

This thread is inspiring me to complete the 12.5" scope I started 4 years ago. I really ought to finish it as I bought a really nice mirror from R F Royce and it's a sin not to use it!

Regards, Hugh

[furrysocks2]

Aye, I read that too - he makes it sound not too bad to be fair. In an earlier post, I wrote "however it ends up performing"... so I'll stick with the elegant, single-piece design. As I intend to build a scope with a modular top end, I could revisit spider design at a later date.

 

8 hours ago, hughgilhespie said:

This thread is inspiring me to complete the 12.5" scope I started 4 years ago. I really ought to finish it as I bought a really nice mirror from R F Royce and it's a sin not to use it!

Nice mirror to have sitting ready to go and I'm happy that this thread is inspiring you - I wish you the best of luck! 

 

I had a thought yesterday that rather than having a spider waterjet cut, I could either get a template laser cut in acrylic or take the time to make a clean-edged wooden pattern - I've used a wood router successfully on aluminium before and could potentially flush trim my own aluminium spider, having first cut the majority of waste off using bandsaw and jigsaw.

I've mocked up the spider in 3/4" ply - not particularly good ply so I'll mix up some epoxy and sawdust tonight to fill the voids (sorry for the poor quality webcam photo). I can scratch around for pieces to begin making a temporary sliding focuser with some degree of collimation adjustment.

 

I also have a holder for the primary that I got with the mirrors and have never used - the weight should help to keep the CoG back. The brass insert with stand-offs will likely have been for use with the other primary mirror - they were both for the same scope but the one in my other scope is much thinner glass.

 

I have six lengths of 1/2" aluminium channel left over from building the previous scope though while extremely light, I'm less certain they would be rigid enough than I am the steel props and it's less clear how I could readily adjust length to substitute a focuser/secondary for the imaging assembly - I could simply have two sets of trusses. In any case, having the twist-lock adjustability of the washing line props would probably aid in sizing any others and worth testing out because it's relatively cheap and easy. My local green supermarket had no props in stock today. 

My dob mount uses soil pipe end caps and a couple of kitchen chopping board pads for the alt bearings and it shouldn't take too much to have the new scope sit in that for first light.

 

A few things to be getting on with:

• clean up of ply spider
• drill and epoxy threaded inserts for mounting
• construct temporary focuser
• prepare the truss rods
• adjustable mount and collimation for primary
• estimate distribution of weight to size bottom end for temporary bearing placement

Despite initially intending it to sit on the existing dob mount, it may help to have an idea of the intended mount(s) as I'm building the bottom end - it's perhaps the one part that I should not allow to simply evolve, but actually design it early on.

I have spare 1.25" and 3/4" aluminium T-track left over from a workbench build and I am wondering if they may be of use for manual adjustment of the position of the primary, or even perhaps bearing position for gross balance adjustment.

 

[furrysocks2]

I believe I can, to some degree, hide other hardware directly behind the wide vanes.

Collimation adjustment needn't necessarily be constrained to fit axially within the 50mm central disc. For example, a suitably sized finder ring or similar could be mounted on standoffs from the main spider, rotated so as to have three narrow, knurled thumbscrews sit directly behind the vanes, using them to adjust the tilt and position of one end of the optical assembly. I should sketch out what that all looks like and work out how far in I can go and still fit the image within whatever diameter of tube is at that end.

I may be able to mount a secondary from the top spider with triangular supports extending back towards the primary, again hidden behind thick vanes. This might be asking a lot of the top spider in terms of rigidity, but the supports could be taken right out to the edges. I could conceivably tie a couple of narrow spiders from the secondary mount to the truss rods, if necessary and if flexure of the truss rods themselves didn't then push and pull the secondary around. The focuser could be mounted onto a plate or similar, being attached both to the outside edge of the top spider and a small clamp around two of the truss rods, somewhere below the position of the secondary.

I don't know how much space I can play with though - perhaps by mounting additional stuff some distance behind front end aperture, despite all being within the footprint of the vanes, there would be additional diffraction visible for certain light paths. In other words, does the "depth" of the obstruction matter wrt diffraction, not just the 2D "area" and shape of the obstruction as viewed from the front?

I appreciate this all depends on rigidity and load etc. Having a secondary assembly and focuser attach without having to dismantle the trusses would be a huge benefit. A restricted on the weight of EPs for example may therefore be tolerated.

All this, unless the belief that I can hide stuff behind wide vanes in the first place is wrong.

 

 

[hughgilhespie]

Hi Matt,

Great start!

And just to put a cat amongst the pigeons, have a look at this site http://www.astro-vr.de/

Spiders - who needs them!!

He is also doing prime focus video astronomy.

Regards, Hugh

[furrysocks2]

Thanks - while out doing the shopping, I've got my washing line props for the truss tubes and a few plumbing parts for a temporary focuser.

 

1 hour ago, hughgilhespie said:

And just to put a cat amongst the pigeons, have a look at this site http://www.astro-vr.de/

Spiders - who needs them!!

Indeed! I hadn't seen that, thanks for the link. He points to a lengthy forum thread to go with it (here). No doubt a heavy piece of steel but it gets the job done, I like it.

I will nevertheless stick with the more conventional (more aesthetically pleasing to me) single truss design. Not that I am any longer considering this an "ultralight" scope as I suggested in my first post - I'm not sure I ever was, other than aiming for lighter than my current scope which is somewhere >10kg - I'd need to weight it.

[furrysocks2]

Example calculations for deformation of various telescope tube constructions:

http://bossanova9.org/astro/ATM/5-Tubes&Trusses.pdf

[furrysocks2]

More maths, it all seems right in my head but who knows. I'm sure it's nothing new but I enjoy working things out from first principles.

 

If I choose a three vane spider and six truss rods, then each truss rod from primary to prime focus will cut a chord across an imaginary circle, resulting in a hexagon, with the mid-point for each side being mid-way between along the truss - the minimum distance across the light path between truss rods has to be wide enough to allow all light to reach the mirror unobstructed.

The mirror is 216.5 mm, so radius is half of that. The height (or apothem) an equilateral triangle is calculated as "S/2 * sqrt(3)", where S is the length of a side, ie the radius of the mirror. And twice the calculated value is the minimum separation of truss rods.

If the truss rods were mounted directly to the perimeter of the mirror, unlikely but illustrative, then I get the height of the hexagon or separation between rods of 187.5mm - narrower than my mirror and hence obstructing light.

Increasing the imaginary circle to which the truss rods are attached to 250mm, I get a separation of 216.5mm which is the diameter of the mirror, but the front end aperture is larger than this so I need to go wider with my truss rods.

From the pdf in the previous post, in reference to wind induced vibration in truss tubes:

Quote

... many of the truss tubes use foam pipe insulation around the tubes. This acts to enlarge the diameter that affects this vortex formation.

Therefore, I will allow for 15mm pipe lagging - the pieces I have are approx 45mm in diameter when put round the 19mm steel tube, so I need to further separate truss rods by twice (45-19)/2, or 26mm.

If prime focus aperture were 230, then the clearance over and above the mirror diameter is approx 13.5mm. Plus the 26mm is 39.5mm. So I need a minimum, bare (uninsulated) truss rod separation of 216.5+39.5 = 256mm.

To achieve this, the imaginary circle to which the bare truss rods are attached needs to be approximately 296mm. My plywood spider from last night just happens to give me approximately 295mm, so all is well - how about that?! 

 

Another consideration is a dew/light shield for the primary, the green supermarket from which I got my washing line props sells a 12L pedal bin for £10, one in matt black. It has a quoted diameter of 25.1cm. My primary mirror cover is approximately 24cm, so it looks like I'll be incorporating a rubbish bin into the build, flocking the inside with some self-adhesive black velour.

[furrysocks2]

A quick ten minutes at the drill press to remove the clips at the top end of the washing line props.

When I know the maximum length I need, I can decide on a minimum insertion length and cut these down - I don't need 8 foot anyway. That'll save a little weight, but overall these are heavy.

The collar on the green sections is a friction fit - they should be secured either with epoxy or a short screw to prevent them slipping off.

I intend to begin without flattening the ends for simplicity's sake, rather simply bolt each individually to a small bracket.

[furrysocks2]

First thoughts for a bottom end - two rings sandwiching the pedal bin, lower of which carries the mirror.

• a copy of the spider for mirror cell mount, but removing less waste to allow for collimation screws while also allowing for fan
• a shallow rebate to accept bottom rim of the pedal bin dew/light shield
• a second copy of spider ring but with no obstructions, also with shallow rebate to accept the top rim of the pedal bin (quoted height 41.5)
• possibly tie the two together with threaded rod through the flanges (120 degrees between)
• two lengths of 1.25" T-track fastened to sides of each ring from which to mount bearings (180 degrees between)
• attach truss rods to flanges of top ring

 

 

Flanges are tapered each side to 30 degrees, meaning no bending of truss rod tubes, though reduces the base of the truss rod triangular and actually makes it a trapezium, but might be ok.

 

[furrysocks2]

Rushed into it tonight.

Laser point deflected 6-7mm from vertical to horizontal. Not bad for all the faults it has, but not good enough.

 

It may be a while before I get to have another crack at it, so a summary of changes to make:

• cut all truss rod sections to 800mm
• giving collapsed length of just over 4 feet, extending to 1700 max overall length, retaining 300mm insertion
• flatten ends of truss rods
• use cross dowels and cross dowel bolts to edge mount both truss rods on a single bolt
• ring in the picture and bottom ring to have larger outside diameter (either remake, or glue addition under)
• meaning the truss rods will taper inwards slightly up towards the spider
• possibly add a middle ring
• threaded rod and wood stretchers for bearings to bottom end

 

[furrysocks2]

Second mock up - didn't take much to do the rods.

No measurable deflection of the laser point down in the bottom from vertical to horizontal.

I have no doubt that I have underestimated many things so far but a few such things became apparent this evening:

• I had to flatten a fair length of the rod ends, more than I'd anticipated - paint flaked off, looks a bit scratty
• the rods obviously therefore won't collapse fully - may no longer fit on the back seat of the car
• with such a degree of freedom, accurate alignment of the spider is probably much more interesting than I gave it credit for
• the wooden rings, even for a mock up are significantly weaker than I imagined
• six truss rods as opposed to eight looks weird to me

It's something to think about, at least.