Jump to content

Stargazers Lounge Uses Cookies

Like most websites, SGL uses cookies in order to deliver a secure, personalised service, to provide social media functions and to analyse our traffic. Continued use of SGL indicates your acceptance of our cookie policy.

sgl_imaging_challenge_banner_lunar_landings.thumb.jpg.b50378d0845690d8a03305a49923eb40.jpg

tonyowens_uk

Members
  • Content Count

    80
  • Joined

  • Last visited

Community Reputation

76 Excellent

About tonyowens_uk

  • Rank
    Nebula

Profile Information

  • Gender
    Male
  • Interests
    Hillwalking, cycling, engineering, travel, wine, politics, telescope-making, model-making, invention, history
  • Location
    Wicklow, Ireland
  1. I think your design and design approach are fantastic Chris! Like you I had to change my whole thinking about pathways to structural stiffness as a result of commissioning 'cheap plastic' prototypes in SLS or FDM from service bureaus. The apparent stiffness and damping with the parts perplexed me until I discovered the routine use of low density infill to cut material use. Hey presto, cellular lightweight plastic 'castings', with resonant frequencies comparable to homogenous metal parts!. My emerging practice now is to use like you a mixture of materials and processes in which FDM and SLS fit. Where ultra-stable 'mechanical grounding' of parts in the face of temperature humudty and load variations is required (as in some optics and some grinding operations) I specify waterjetted granite or moulded polymer concrete. For thermal stability but lower stability requirements I use steel or Permali Wood. For general structures and parts without any special strength or stability aluminium in all its forms was what I used. There has been a developing problem over the last 25 years with the gradual disappearance of jobbing machine shops in Anglo countries, presumably linked to the financialisation of their economies and a wholesale shift to Asian sourcing for consumer goods. This has made both my professional and 'hobby' work more difficult. I have responded by investing in basic machining and printing capability. I'm seeing a gradual substitution of SLS and FDM printed parts, with critical surfaces machined if necessary. I also source gravity and pressure die casting parts from India and mouldings from many sources including Poland and Spain, where previously I would have turned to UK and German firms. The main reason I posted details of my 14" planetary Newt concept (I havent had time to start building one!) was to offer you a quantitative datapoint about gravitational sags and frequencies in a similar truss-type telescope structure for high resolution use. My concept used machined and extruded aluminium for most of the structure. I am not judging your structural concept about which I know little. FWIW given the cellular parts and super-light weight at the secondary end, I would be surprised if you were troubled by structural issues. There are some unique advantages that apply to 3D printed parts for telescopes too. One of those, is high levels of insulation. For things that are within or close to the optical beamline, where metal parts like tubes and spiders if made of metal tend to sub-cool during clear nights, there is far less of an issue if cellular dimensionally-stable resins are used. Another is relative freedom from the need for coatings and finishes. Finally there is the ease of incorpating things like brackets, reinforcements, pneumatics and wiring. As you say, our profession is having to reinvent again all the established design techniques, in a world where the West now struggles to manufacture things profitably and where additive technology is finally becoming reliable and good enough for first-class products. I'd be very interested in your progress on the Gregorian once the secondary is polished! Good luck with your endeavours! Tony Owens
  2. Hi Chris and my compliments on your unorthodox monorail Gregorian. This style of 'monorail' tubeless scope reminds me of the late Horace Dall's planetary Dall-Kirkham. I had a crack at designing a 14" F4 planetary prime focus Newt a couple of years ago. At the prime focus I could interchange various coma correctors, cameras and filters. I used ITEM alloy extrusions with screwed and epoxy-bonded joints and waterjet-cut pieces of 4mm and 10mm alloy plate to construct a goniometric primary mirror mount and various reinforcements to get the gravitational sag under control. For a Newt the max tolerable decentration of the 'true' axis of the primary mirror from the eyepiece or camera to maintain no more than a 0.20 drop in Strehl is given approx by DeC = 0.005 . FR^3 [mm]. For my optic I therefore needed to assure no more than around 0.32 mm decentration of the primary from the camera, for all possible poses of the OTA and for the heaviest imaging train payload at the prime focus. This number is somewhat simplistic and debatable and little more than a guideline for acceptable structural sags, but as a target it was fine. In designing a structure I actually worked to around 25% of this target and aimed for a lowest natural frequency of around 20Hz. The thinking here was that the EQ8 mount I was planning to mount this on is marginally capable of supporting a 20Kg payload/counterweight combination with a natural frequency of around this. The natural frequency not just of the mount but of the entire scope/camera system needs to be as high as possible to allow a good percentage of quality frames when lucky imaging Moon and Jupiter at 120 Hz especially with some wind. Here is what I ended up with: I used a CFRP flat plate secondary single-stalk design which is magnetically fixed to the OTA using a kinematic mount for interchangeability. The mirror is contained in a lightweight shroud. The structural deflections due to gravity (linear and tilt displacements) of the OTA at zenith pointing looked good. This was Horizon Pointing showing less than 100um of linear sag of the prime focus group: This was Zenith Pointing. Note how tilts of Primary and primae focus group matched: This was another Horizon Pointing scenario with the OTA rolled through 90 degrees w.r.t. the gravitation direction. It proved to be the worst case but still the linear sags fell not too far outside my target: This pic shows the lowest natural frequency of the OTA assuming a rigid circular saddle support of diameter 160mm is used. I call this 'Nodding Duck'. The frequency is 34 Hz which is perfectly acceptable: This is the second lowest vibration mode, at 38 Hz. I call this 'Secondary Stalk Waggle': Finally, this is the third lowest vibration mode which I named 'Primary Waggle'. It appears at 42 Hz so is not a concern: My experience was that structural FEA was of tremendous help in arriving at something that almost met my stiffness requriements under all possible poses and imaging payload configurations. It never occurred to me to 3D print structural parts in something like PET-G using lightweight cellular infill at that time because to be honest I did not take FDM printing seriously. I since revised that silly opinion and am now gradually learning the practice of large 3D printing with a large printer. I am very encouraged to see your project and hope your practical findings about collimation stability show my own analysis to be too conservative! best Tony Owens
  3. What I'm looking for doesnt exist yet! I want a mount that has no meridian flips, no worm gears (despite all my work on refining EQ8's) no microstepped stepper motors, has a weight and size low enough to be airline cabin baggage, maintenance-free, cable-free, backlash-free, counterweight-free, capable of being rained-on without ill effects, and with at least a 20 Kg imaging payload. Perhaps combine the wireless servo controls and GUI of the AZGTi, provide differential timing belt transmissions, reinforced phenolic casework, wire race bearings and a single arm fork design... Or maybe something with delta 3d printer kinematics plus a 4th axis for image derotation? One of the best resolved commercial designs I've seen recently is the Panther line of Goto mounts. Despite the counterweights!
  4. A good gloss of that period of hope and belief in the power of technology to effect change for the better. Spoiled a bit by the creaky voice (vocal fry) now fashionable among yuppie American women.
  5. All worm gear mounts of this class are pretty similar and better than 5 arcsecs peak-to-peak periodic error is typical. It doesn't make any real difference what the exact figure is, provided sensibly tuned autoguiding is being used. All will be reduced to 0.5" RMS on RA and a bit less on Dec, provided backlash is controlled. The EC mounts are a different animal, and essentially offer more linear RA or RA/Dec motion. But this is not a useful advance on open-loop tracking (no encoder/s) unless unguided imaging is a requirement and some mount modelling can be considered.
  6. It doesn't really matter which of the three you mention you choose, if performance is your major requirement. All are good solid mounts, all autoguide well. The 2015-on EQ8's/HD110's are better than the early mounts due to lower Dec backlash. The CEM ad AP mounts have spring loaded gears unlike the EQ8 which helps worm gear mesh clearance (i.e. backlash) control between summer and winter temperatures.
  7. Wim i'm interested in your Linux setup due to the USB3. In principle I should be developing my high resolution planetary imaging, but keep getting distracted onto other pursuits! What had stumped me in putting together a good fast planetary imager was the 2-3 metre length limitation of a ZWO planetary cams like the ASI174, when running flat out. The best resolution appeared to be a USB3-capable SBC mounted on the OTA or mount saddle plate, then off via hard Ethernet or WiFi to a server. But Intel NUC's are a bit pricey and delicate for that sort of outdoor exposure. Something else was required... My understanding of USB3 on SBC's around 1 year ago was that basically very few exist and none are reliable. What has been your experience? cheers Tony Owens
  8. Sounds good Gina. That will get the scope back running by the shortest route. Best Tony
  9. OK. I was talking about how to remove and refurbish the entire cell complete with optics as a unit. Can you mark and extract the optics from the cell/tube assembly and get them out of harms way? If so, provided you photograph and carefully replace any lateral and axial shims used to centre the corrector, you could of course just decide to leave the cell alone. It would not be my choice (I dont take any nonsense from mechanical assemblies) but you could do that, and just clean the optical elements and replace them in the cell. Otherwise, with the optics out, you can warm and extract the frozen screws as I've described. How many of these radial screws are being obstinate? Do you reckon they've been threadlocked in place? If so, the usual method of removing fasteners bonded with acrylic glues is to carefully heat the joint to 200C which softens the glue enough to break out the screw. A flame cannot be used due to the tube paint. But a soldering iron on the screw head might do the trick...
  10. Not sure. But washing broadband multicoated lenses is not rocket science and is low risk. There are not many moisture sensitive glasses used in astro correctors and eyepieces that are not BBAR coated and therefore quite cleaning-resistant. You need both organic and inorganic solvents to wash the lens then lots of DI water to flush. Mechanical cleaning by wiping with soft wetted tissue is a good idea provided scratching is minimised. Do the work in a plastic bowl and warm the diluent water. Whatever about high priced wonder fluid, there is nothing wrong with good old screen wash concentrate, diluted down with clean water. This contains alcohol/s plus a little acetone, in an aqueous base. Plus cool looking blue dye. Very important to flush several times with DI water. Car accessory shops will have this too! Dry by arranging in a rack (dishwasher drawer?) and gently using clean compressed air to blow off the drops. Not rocket science... Tony
  11. Gina I have a couple of Intes-Micro Mak-Newts so I've experienced your issues. If you have time to do the job properly I'd follow Peter's advice and remove the cell and service the corrector and secondary properly. Marking the joints between every part with a small permanent Sharpie pen beforehand. Change the fasteners for stainless ones and either use silicone grease on the threads or seal them with a low strength loctite retainer. RS and Farnell have this. Warm the cell laid flat on a sheet of clean tissue in your domestic fan oven to 60 degrees, then remove and use good quality hex wrenches on the screws and they should release. If you manage to ream the head of one, carefully drill the head off using a Dremel and remove it that way. There are various ways to extract the remains, once the precious corrector and folding mirror are out of the picture. Cleaned, reassembled and collimated it should be as good as new or better. If not, we have a high resolution Shack Hartmann analyser and a certified 1/10th wave autocollimation flat big enough, over here. I'd be happy to sort out any alignment problems you encounter FOC. Tony Owens
  12. This is a different animal than a 'hypertune' or even an aftermarket worm. There are fundamental problems with the design of clutches and choice of worm and shaft bearings in the EQ8 which were addressed to some extent by the manufacturer after a mixed reception to the original product. I was one of the victims in fact... The eventual result of my efforts was a new form of antibacklash worm gearing that also includes constrained layer damping to lower the meshing noise 'floor'. Also a redesigned clutch that can be tightened without imposing radial load on the shaft and inducing backlash and an image shift. Improved environmental sealing and a precision transfer timing belt drive to RA and many other improvements are there too. In combination with some other changes it amounts to a modernisation not an overhaul. A UK patent application on the gearing was filed at the end of 2018. The effect is that there is no significant backlash on either axis, regardless of axis position, ambient temperature or wear. Nothing to fiddle with - the mount 'just works', as it should have done back in 2014 when I bought mine! Tony
  13. There will be a website at some point but finding time is an issue for me. There is a special interest Yahoo Group about EQ8's here that includes various contributions from me about the project here: https://groups.yahoo.com/neo/groups/SkywatcherEQ8/info Pictures of the mount and our shipping crate for couriered collection and return: Picture of before and after polar plots (5 test points, equally-spaced around each axis) showing the effect of the new gearing on backlash as benchtested: Dec Backlash as tested on the sky with PHD2 and a 1180mm EFL 20 Kg autoguiding payload: The autoguiding performance has been tested and shows around 4" pk-pk periodic error for the new RA gearing, which corrects down to 0.40" RMS on RA and 0.35" RMS on Dec with careful tuning of settings on PHD2. There is some work still ongoing on improved autoguiding involving different motor firmware and a new ASCOM server which offers certain advantages over EQMOD in principle, and perhaps also in practice. We shall have to see! Tony Owens
  14. Welcome David! Tony
  15. I'm late coming to this discussion - apologies. One potentially interesting avenue to consider, against the CEM120 and upcoming centre-mount competitor model from Skywatcher, is an upgraded EQ8. I need to disclose that I make these so have a vested interest. But ignoring that for a minute. Some of my spare time over the last few years has gone into sorting out the limitations of this mount and this has been quite successful. Imaging load capacity, and guiding responsiveness and accuracy are significantly improved due to use of stiffer ground steel shafts in place of the aluminium originals combined with different worm bearings, damped, diamond-lapped antibacklash gears and now a new ASCOM server named Green Swamp, specifically designed for this mount. Through-the mount cabling is relatively trivial to implement if needed as the axes are hollow. Custom saddle plates for larger instruments and cranked piers for eliminating meridian flips are straightforward too. This route makes sense for those more interested in functionality than 'bling'. Specifically, for people who: 1. already own early EQ8's with large Dec backlash issues which make good imaging difficult 2. need a reliable, economical, heavy imaging mount/s for remote use, and are unconvinced of the superiority or practicality of unguided encoder mounts 3. are willing to consider buying a cheap used EQ8 as a basis for upgrade The cost of our modernisation runs 1200-2000 euros plus VAT where applicable, depending on spec.
×
×
  • Create New...

Important Information

By using this site, you agree to our Terms of Use.