Jump to content

  •  

CNers have asked about a donation box for Cloudy Nights over the years, so here you go. Donation is not required by any means, so please enjoy your stay.

Photo

14" f/4 Hubble Hexapod build

  • Please log in to reply
21 replies to this topic

#1 Adam Long

Adam Long

    Vostok 1

  • -----
  • topic starter
  • Posts: 158
  • Joined: 19 Sep 2019
  • Loc: Sheffield, UK

Posted 25 November 2020 - 09:58 AM

Well I've got a delivery date for the mirror so will share some of my thinking for discussion. Previous discussion on scope size here: https://www.cloudyni...4-16-f4-or-f45/

 

Basically I'd like a bigger scope but have a small house, and it needs to be portable. My existing main scope (link in sig) is a lightweight 10" hexapod. I'm very happy with the performance and portability, but it is only balanced with eyepieces up to ~350gm. So, along with a step up in light grasp and resolution, I want this scope to be able to handle heavy eyepieces/ binoviewer plus coma corrector. I'd like the option of a finder scope too.

 

So, for starters here's the mirror support plot from PLOP-lite via CBScope:

 

Screenshot 2020-11-24 at 13.31.12.jpg

 

Looks like a six-point cell will be more than sufficient (this is based on modelling the full-thickness in quartz as per https://www.cloudyni...ject/?p=4825659). Intending a triangle of steel tube to support, with the ends connected direct to the six-leg truss. The rest of the LTA will be laminated from birch ply similar to my 10", likewise rocker box and ground board but need to draw proper plans up.

 

Edge support - not sure here. With the Hubble Mirror being a sandwich, some extra complexity is involved. I think I'll try two-points at 90 deg, with short vertical whiffletrees to share the load to both plates. But a ladder sling is the other obvious option.

 

UTA - Merope's looks like a good starting point - two rings of 9mm ply, ~50mm wide. Moonlite focuser, on a 12mm board. Other bits of ply or tube to join, 2mm ply to line. Inner diameter mirror + 44mm.

 

Secondary - played a lot with NEWT and Mel's calculator. Choices from Hubble were 70mm or 80mm but I'm probably going for a 75mm from Orion UK which gives 11.8mm field at 100% and 45mm for the 75% - which should cover any field stop I'm likely to use. At this position the paracorr will impinge on the light path a little, but only at the biggest field.

 

Spider - not sure. Brass sheet offset like my 10" seems simplest, but maybe wire? Not built a secondary support before so need to think about that too.

 

Truss - again I'll be using female steel rod ends at the bottom and male igus at the top, all M8. UTA fixing will probably be the same with eyebolt and thumb nut, LTA I've got some ideas - my current fixing works well but is awkward to access with the bearings on.

 

Just pricing up carbon tubing - looks to be about £120 vs £30 for aluminium + heatshrink. Inclined to go for carbon as it's a nicer finish, warmer to touch and I can create slightly tapered tubes by nesting smaller sections at the ends. That will also allow the tube lengths to be finely tuned rather than having along length of threaded rod, and might even be a bit stiffer than Al.

 

Any input welcome.



#2 happylimpet

happylimpet

    Fly Me to the Moon

  • *****
  • Posts: 5,365
  • Joined: 29 Sep 2013
  • Loc: Southampton, UK

Posted 25 November 2020 - 10:37 AM

 

Secondary - played a lot with NEWT and Mel's calculator. Choices from Hubble were 70mm or 80mm but I'm probably going for a 75mm from Orion UK which gives 11.8mm field at 100% and 45mm for the 75% - which should cover any field stop I'm likely to use. At this position the paracorr will impinge on the light path a little, but only at the biggest field.

 

Dont buy your secondary from Orion UK. My problems with my 16" primary are documented elsewhere, but more relevant to your situation, my ~72mm (i forget exact size) secondary was very astigmatic. I swapped it out for an 80mm GSO and got perfect star tests. When Orion tested the mirror set, they reported that this secondary 'had been bent'.  Mirrors dont get 'bent'. Bunch of shysters.


Edited by happylimpet, 25 November 2020 - 10:40 AM.

  • clivemilne, Augustus and Adam Long like this

#3 stargazer193857

stargazer193857

    Voyager 1

  • *****
  • Posts: 10,573
  • Joined: 01 Dec 2013
  • Loc: Southern Idaho

Posted 25 November 2020 - 10:47 AM

The thickness ratio on the primary is 9:1. I'm surprised a 6-point does not perform even better than 7.8mm. I thought 10.5:1 was close to 4nm, and 13.5:1 and even 16:1 still good by some people's standards.




...

A portable, fast cooling 14" sounds nice.

Edited by stargazer193857, 25 November 2020 - 10:48 AM.

  • Adam Long likes this

#4 Adam Long

Adam Long

    Vostok 1

  • -----
  • topic starter
  • Posts: 158
  • Joined: 19 Sep 2019
  • Loc: Sheffield, UK

Posted 25 November 2020 - 01:20 PM

Dont buy your secondary from Orion UK. My problems with my 16" primary are documented elsewhere, but more relevant to your situation, my ~72mm (i forget exact size) secondary was very astigmatic. I swapped it out for an 80mm GSO and got perfect star tests. When Orion tested the mirror set, they reported that this secondary 'had been bent'.  Mirrors dont get 'bent'. Bunch of shysters.

Hmm, thanks for that. I was aware they had a mixed rep but that is worrying. I am close enough to visit them but will look at other options.

 

 

The thickness ratio on the primary is 9:1. I'm surprised a 6-point does not perform even better than 7.8mm. I thought 10.5:1 was close to 4nm, and 13.5:1 and even 16:1 still good by some people's standards.

 

Must admit this is the first time I've used the software so may have made an error somewhere, and I'm not sure if the output is the same as standalone PLOP. From the what it says on the wiki the support seems very good?

 

 

 

P-V errors. This gives you the delta between the top and bottom parts of your deformed mirror (only the area that are visible are taken into account. The area hidden by the secondary will be ignored!). P-V and RMS errors are indicated both in nano-meters (green light is 555 nm) and in lambda or fraction of the green wavelength. For “Normal” folks, lam/20 is usually “good nough”… for those of you lucky enough to have l/20 Zambuto, get to at least l/40 or 80. Your mirror deserves it!
RMS error. This is a much better measure. Let me explain why. A P-V of 1/2 length wave of light is bad (it means that the top cancels the bottom). However, if 90% of your mirror is at ‘0’, 5% at -¼ and 5% at ¼ (P-V=1/2) then you are only loosing 10% of your mirror surface, which is not “that bad”… RMS measurement takes into account not only the height, but the amount of surface at these heights. So, a bad P-V with a great RMS is better than an average P-V and average RMS as RMS “removes” the small outlier areas of the scope. Anyhow, ¼ wave is what is normally considered an OK mirror. If your cell has less than 1/20 wavelength (lambda or lam here), it means that the deformation to your mirror caused by the cell will be very small compared with the overall mirror quality and that there is no need to make the cell any more complicated.


#5 MitchAlsup

MitchAlsup

    Aurora

  • -----
  • Posts: 4,987
  • Joined: 31 Aug 2009

Posted 25 November 2020 - 04:59 PM

PLOP-heavy (std PC version) says 1.89nm with the support points on the 58% radius.

Clearly well within the 4.5nm std requirement.


  • stargazer193857 and Adam Long like this

#6 Adam Long

Adam Long

    Vostok 1

  • -----
  • topic starter
  • Posts: 158
  • Joined: 19 Sep 2019
  • Loc: Sheffield, UK

Posted 25 November 2020 - 05:27 PM

Thanks Mitch, appreciate the confirmation.



#7 brebisson

brebisson

    Messenger

  • -----
  • Posts: 424
  • Joined: 22 Aug 2016

Posted 26 November 2020 - 02:25 AM

Hello,

 

CBScope uses the PLOP code to do the calculations. For some reasons, the values are very slightly different (the init is slightly different, and the iterative nature of the system causes the differences), BUT the result are inline with PLOT and also matches calculations done through other FEM software (solidworks).

 

With regard to lambda and errors.

Adam gave a good explanation of surface error on the mirror.

If, on average (RMS), the error is less than the wavelength/8, causing an error of less than lambda/4 on the wavefront (the wave front error is 2 times the surface error because the reflection doubles the distance). Then the mirror is considered OK for visual use.

To that error, build in the mirror, so to speak, you will need to add errors caused by the deformation of said mirror under it's own weight due to the fact that it is not supported uniformly. This is what you are calculating here.

With your 6 point support, the error will be smaller than lambda/350 on the surface, or lambda/175. So negligible compared with the mirror surface error.

In addition, this error is a worse case scenario! it applies when the mirror is horizontal. But since most observation is done at an angle, the weight on the back of the mirror will be less and the deformation less!

 

 

Secondary - played a lot with NEWT and Mel's calculator. Choices from Hubble were 70mm or 80mm but I'm probably going for a 75mm from Orion UK which gives 11.8mm field at 100% and 45mm for the 75% - which should cover any field stop I'm likely to use. At this position the paracorr will impinge on the light path a little, but only at the biggest field.

Note that the "secondary" tab in CB scope is designed to cover JUST these calculations.

I created it on purpose to avoid having to re-enter all the scope data in the 4 programs: Newt, Mel's calculator, PLOP and FigureXP...

The help for that page is quite detailed and you can really get a lot from it. Especially if you link it with the EP page so that you can see your field stops.

The only thing that is not handled much by that page is vignetting.. Mostly because my scopes have an open nature (trusses, mono-truss)... But even then, once you put EP in the previous page, you will see the light cone coming in the EP field, which should be of great use.

 

Cyrille


  • Adam Long likes this

#8 Adam Long

Adam Long

    Vostok 1

  • -----
  • topic starter
  • Posts: 158
  • Joined: 19 Sep 2019
  • Loc: Sheffield, UK

Posted 26 November 2020 - 04:33 AM

Thanks Cyrille, both for the software and the tips. My explanation of the surface error is good because it was copied and pasted from your support pages. I didn't discover your program until I'd already calculated the secondary, but it will be a one-stop shop next time, great work!

 

On your EP page what are the units for the field radius? Seems like the field stop diameter would be a useful data point to enter for the most precise calculation of the illuminated field?

 

One thing I'm not clear about is the effect of the paracorr. It feels like the end in the light path must be the controlling element of the illuminated field - because it has to fit inside the draw tube like the field stop but is much closer to the primary. I haven't got one to measure. Found this in a previous discussion, but I don't think it's quite the same thing:

 

the fully illuminated field for the Paracorr 2, assuming a 60" focal length:   f/4.0  -  27.9 mm


Edited by Adam Long, 26 November 2020 - 04:38 AM.


#9 Jakke

Jakke

    Lift Off

  • -----
  • Posts: 24
  • Joined: 25 May 2019
  • Loc: Blauvac - South of France

Posted 26 November 2020 - 09:00 AM

Hi Adam,

 

I bought a small secondary from Orion via Telescope Express.

I am really happy with it, works fine in my 12" F4 Hexapod.

 

I really liked your 10", good luck with the build !

 

spiff1.jpg

uta-4.JPG


  • Don H and Adam Long like this

#10 brebisson

brebisson

    Messenger

  • -----
  • Posts: 424
  • Joined: 22 Aug 2016

Posted 27 November 2020 - 01:28 AM

Hello,

 

 


On your EP page what are the units for the field radius? Seems like the field stop diameter would be a useful data point to enter for the most precise calculation of the illuminated field?

 

One thing I'm not clear about is the effect of the paracorr. It feels like the end in the light path must be the controlling element of the illuminated field - because it has to fit inside the draw tube like the field stop but is much closer to the primary. I haven't got one to measure. Found this in a previous discussion, but I don't think it's quite the same thing:

 

 

The field stop/radius is in mm like the whole program (I made the whole program metric only, first because it is easier, and in some way to annoy the americans :-) it's my cheesy side...)

 

The field stop is calculated using the formula focal*2*tan(angle/2)

The paracor will limit the light cone in some way. And might slightly change it's angle (focal multiplier/divider). However, this cone angle modification, since it will be very close to the focal plan will not change much to the actual focal plan position.

You can normally do all the calculations as if you did NOT have it. The only thing that will change is the data at the bottom of the secondary graph will be multiplied/divided by that factor.

 

Cyrille



#11 Adam Long

Adam Long

    Vostok 1

  • -----
  • topic starter
  • Posts: 158
  • Joined: 19 Sep 2019
  • Loc: Sheffield, UK

Posted 27 November 2020 - 05:01 AM


The field stop/radius is in mm like the whole program

 

Thanks Cyrille. Metric works great for me too, but guessing english is not your first language as radius means half the diameter - so not the same as the field stop, which is what confused me. Perhaps change it to diameter or field stop? (eyepiece is also spelt wrong).

 

The other point is entering the data for my eyepieces did not produce the correct field stop. I guess this is because wide-angle eyepieces only achieve the specified AFOV with distortion, so the field stop is typically smaller than the maths predicts. I would prefer to be able to enter the field stop as measured or specified. But please note these are small quibbles with a great program!


Edited by Adam Long, 27 November 2020 - 12:19 PM.


#12 brebisson

brebisson

    Messenger

  • -----
  • Posts: 424
  • Joined: 22 Aug 2016

Posted 27 November 2020 - 11:54 AM

Hello,

 

it it is very possible that I messed up and typed one while meaning the other! I need to check.

 

yeha, not a native English speaker here (French)... but my spelling is terrible in French also :)

 

cyrille


  • happylimpet and Adam Long like this

#13 jtsenghas

jtsenghas

    Fly Me to the Moon

  • *****
  • Posts: 5,326
  • Joined: 14 Sep 2014
  • Loc: The flatlands of Northwest Ohio 41.11N --Bloomdale

Posted 30 November 2020 - 10:13 PM

Regarding edge supports on a Hubble Optics sandwich mirror, I'd recommend you set up vertical whiffle tree rollers like I did here:

 

https://www.cloudyni...b/#entry9060168

 

I converted that build to a monolithic mirror before finishing that scope, but I can vouch for the fact that it exhibited no visible signs of astigmatism during testing for focal length with temporary struts for low altitude objects. 


  • Adam Long likes this

#14 brebisson

brebisson

    Messenger

  • -----
  • Posts: 424
  • Joined: 22 Aug 2016

Posted 01 December 2020 - 12:54 AM

Hello,

 

Another option for the edge support would be to use linear bearings, potentially on a rocker.

 

https://www.amazon.f...aps,158&sr=8-31

 

Cyrille


  • Adam Long likes this

#15 Adam Long

Adam Long

    Vostok 1

  • -----
  • topic starter
  • Posts: 158
  • Joined: 19 Sep 2019
  • Loc: Sheffield, UK

Posted 01 December 2020 - 04:16 AM

Thanks guys. I was already converging on something like that but good to know it works.



#16 Adam Long

Adam Long

    Vostok 1

  • -----
  • topic starter
  • Posts: 158
  • Joined: 19 Sep 2019
  • Loc: Sheffield, UK

Posted 02 December 2020 - 01:08 PM

Mirror has arrived! Nicely packaged and no problems with shipping from China. No test cert but the label on the box has the following info:

 

IMG_4044.jpg

 

Carbon truss tubes also arrived today. Very stiff for the weight, and a really nice finish. Time to get building!

 

 


  • jtsenghas likes this

#17 Adam Long

Adam Long

    Vostok 1

  • -----
  • topic starter
  • Posts: 158
  • Joined: 19 Sep 2019
  • Loc: Sheffield, UK

Posted 02 January 2021 - 06:28 AM

Well that month went quick, time for a progress update!

 

IMG_4087.jpg

 

Mainly been building a kit of parts up, will need a few more bolts but not much else. Have both mirrors and focuser, UTA design is done and have made the first cuts.

 

Truss more or less done apart from LTA attachments. I was keen to use carbon tubes but on the initial sketch plans I realised they'd be a little over a metre long which would mean paying much more for 2m long tubes then cutting them down. Instead I came up with a much cheaper method using 800mm lengths with short sections of thinner tube at the ends. So far it looks like it will work well, and with these and the threaded rod I'll have a ton of adjustment for length before committing to gluing up the tubes. I was a little concerned that 1mm wall thickness tubing might be too small but the tubes feel very stiff and plenty robust.

 

IMG_4088.jpg

Cut all the short sections of carbon tube yesterday using a borrowed tile cutter as recommended by Oberon. Worked really well, quick, fairly precise and no dust. The cutters are only about £40 but all I paid for was a new diamond disc. A simple plywood jig held the tubes square and ensured consistent length. I also bought a small sheet of 1mm CF to stiffen parts of the UTA. This cuts fine with a router and an attached vacuum takes care of the dust.

 

Cutting all the tube inserts - delrin for the truss, ply for the UTA  - is done thankfully as they are a painful chore on my £25 'bead machine' for which the best accessory is a lot of gumption.

 

The piece of aluminium is a 135o extrusion and will form the spider centre and secondary mount, hopefully centering the forces as per Mitch Alsup's design.

 

Got a great deal on a load of Baltic Birch offcuts from the local timber merchants so have plenty of wood. Haven't finalised the LTA and cell design yet but it is sketched out in my head and likely to have some resemblance to my 10".


Edited by Adam Long, 02 January 2021 - 06:31 AM.

  • jtsenghas likes this

#18 Adam Long

Adam Long

    Vostok 1

  • -----
  • topic starter
  • Posts: 158
  • Joined: 19 Sep 2019
  • Loc: Sheffield, UK

Posted 03 January 2021 - 01:12 PM

Encouraging progress on the UTA today - dry fit of parts:

 

AL-5282.jpg

 

Some tweaks and additions required but looking like it will work. Glad I didn't go for the 16" now!


  • brave_ulysses and stargazer193857 like this

#19 Adam Long

Adam Long

    Vostok 1

  • -----
  • topic starter
  • Posts: 158
  • Joined: 19 Sep 2019
  • Loc: Sheffield, UK

Posted 04 January 2021 - 12:41 PM

IMG_5627.JPG

 

Couldn't resist putting the truss together. Looks like the stepped poles should work well, and should fold neatly if ia get the spacing right. And as a bonus it looks like the vertices of the triangles will be close to the UTA cog.

 

 


  • erick86 likes this

#20 Adam Long

Adam Long

    Vostok 1

  • -----
  • topic starter
  • Posts: 158
  • Joined: 19 Sep 2019
  • Loc: Sheffield, UK

Posted 20 January 2021 - 05:50 AM

Further progress...

 

Secondary holder/ spider: Having given the cellar a good clear out during Lockdown #1, and built a workbench, I've got a much nicer workspace than for my last build. But I'm well aware of the limits of my skills and tools when it comes to precise work. So I bought a short section of 135o Al extrusion to use as the spider hub, and after a fair bit of cutting drilling and filing had it down to a light but rigid configuration with some adjustment built in.

 

For the holder plate, I started with a triangle of 9mm ply, chiselled a few plys off so it was nearer 6mm, then drilled the centre out and turned a delrin insert to a tight hammer-in fit.

 

IMG_4111.jpg

 

This was then laminated front and back with 1mm carbon fibre, then drilled and tapped so the support bolt (and the hub axis, though less precisely) are aligned with the secondary's CoG.

 

IMG_4113.jpg

 

I attached the secondary with 3M VHB pads. I appreciate that isn't the usual choice but I'm happy that it is plenty strong, and flexible enough though I don't anticipate much differential expansion. The GSO 74mm secondary is also 15mm thick (the 75mm OO one I was going to buy is 10mm) so I think support issues are unlikely here. If anyone has any horror stories of VHB failing I'll add a safety string.

 

I then cut some spider vanes from 0.2mm titanium sheet with scissors. Pretty quickly I realised that soldering them to brass rod, as I did last time, isn't going to work - the solder just doesn't stick to it. It would also leave me with no height adjustment. As a trial I've made a little Al interface block to thread onto the rod and grip the end of the vane with a grub screw, but although it solves both problems it doesn't look great and adds some obstructions to the light path. Plus the Ti sheet has some nice properties but I'm not sure it will ever sit very flat. So I've now parked the UTA build while I reconsider a wire spider.

 

While I mull that over I've made some progress on the LTA. Steel frame and mirror support bars all cut with a hacksaw. Sketching out the LTA shape and trialling quick release lower truss attachment.

 

IMG_4129.jpg


  • Oberon likes this

#21 Adam Long

Adam Long

    Vostok 1

  • -----
  • topic starter
  • Posts: 158
  • Joined: 19 Sep 2019
  • Loc: Sheffield, UK

Posted 20 January 2021 - 03:10 PM

Few more pics, secondary holder/ spider hub (three grub screws will allow some adjustment, though I expect routine collimation to all be done with the truss):

 

IMG_4136.jpg

 

 

Plan for quick release LTA truss connection:

 

Stainless clevis bolted through to steel triangle mirror support, holds short alu arm:

 

IMG_4134.jpg

 

 

Rod ends locate against clevis, arm pulls down and in and over-centre clamp to hold down into recesses (not yet cut).

 

IMG_4133.jpg

 

I had envisaged the lever being longer so less force required on the clamp, but I don't want to make the UTA too wide and I've got adjustment at both ends so hopefully will work nicely as it is, but it needs to clear the bearings... we'll see. It also stops the rod end body twisting while allowing sufficient movement.


Edited by Adam Long, 20 January 2021 - 03:15 PM.

  • brave_ulysses, Oberon, starcanoe and 1 other like this

#22 Oberon

Oberon

    Soyuz

  • -----
  • Posts: 3,570
  • Joined: 24 Feb 2013
  • Loc: Hunter Valley NSW Australia

Posted 21 January 2021 - 05:41 AM

waytogo.gif

 

 Very nicely done!


  • Adam Long likes this


CNers have asked about a donation box for Cloudy Nights over the years, so here you go. Donation is not required by any means, so please enjoy your stay.


Recent Topics






Cloudy Nights LLC
Cloudy Nights Sponsor: Astronomics