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_30_second_exp_2_winners.thumb.jpg.b5430b40547c40d344fd4493776ab99f.jpg

Ajohn

Members
  • Content Count

    783
  • Joined

  • Last visited

Everything posted by Ajohn

  1. These days a lot of OSS desktop stuff is extended or written by people who want it to do something or the other. They come and go. Personally I feel that desktop Linux is in danger long term. Server side should be safe but Microsoft keep trying to get more and more of that. People seem to forget that some Linux distro's are run by commercial organisations. I believe Redhat went bump recently. More might follow. Ubuntu used to mention company wide solutions as the main header on their web page. In some ways this is what some of them aim at and that plus selling support is what keeps them going. The pay for packages some offer are also more stable than the free stuff. While it may seem a good idea to have every one running Ubuntu it wont be long term. It's beginning to look like some producers think that has happened already. To me though INDI etc seems a bit odd. I have installed debian and other distro based packages before without any problems from source. I'm wondering if the problem is down to it mostly being run on Gnome and that it includes a KDE application - kstars. I've always run KDE and if I can't find a distro rpm for a Gnome application I have had problems installing it. Not something that has cropped up very often so pass. I just tried installing the INDI Ekos VM in virtual box on Linux to get something up and running. VB comes up with a corruption error. I think they provide the VM to allow Windows and others to run it without Linux. It's on their download page if show all downloads is clicked. The error relates to unzipping. Odd as I don't usually have any problems unzipping what ever or using VB. Ubuntu? I did try it once. Spent time making Gnome a little more kde like. Set up a separate root account and found that even though it had let me do that it didn't work. They seem to have hijacked it for some reason or the other. Then tried the KDE version more recently and was extremely unhappy about the miserable selection of software it's desktop installer offered. I do most things from the desktop including working as root at times. Rant - What was the initial comment on the INDI forum concerning OpenSuse - switch to a more modern distro ??? Actually they have an answer to bleeding edge called Tumbleweed - you will get it as it arrives if that is what some one wants. Rant over. John -
  2. I'd try asking the distro people Steve. I seem to recollect a mailing list or something similar. Sounds like bleeding needs fixing as well. That wouldn't surprise me at all. Sometimes the need by some people to change software causes them to forget usability factors and press ahead with major changes or enhancements. I'm glad it runs for you. Hope it does for me too so that I can see what it's at. I've owned various telescopes for a long time but until now have steered clear of imaging so am setting up for that from scratch. Me - I'm a confirmed Linux user and I too find the INDI area frustrating. It might make me run windows software again. From your comment the asto distro people sound like they may be fairly sane. No doubt it will update at some point. That can take rather a lot of work at times especially if the bug count needs to be low. I followed up an arduino twin channel focuser shown on indi - I dare not comment on what had been put in the arduino. John -
  3. You've hit on the reason I used the word disappointing Steve and some others. INDI is a bit juvenile compared with main stream OSS. This is unusual actually. The problem with bleeding edge all of the time is that there is a fair chance that problems will continue to occur and some will remain undiscovered before the next one comes out. The odd thing about INDI as well is that they don't seem to keep earlier versions about. I have found older source on some distro archives but it doesn't seem to be in the same form as the ones they supply. I'll probably look at that more later. To be honest INDI is the first piece of software I have tried to install from source and not been able to do it even though the dependencies are all there. Usually the only problems I have is not being able to install updated libraries without upgrading the desktop - KDE in my case. Sometimes that isn't a problem as the new libraries can co exist. It's a bit sad really. All distro's maintain stable releases for some time. These usually need upgrading not long after the next stable one comes out. The fact that they do this makes it easier for applications people to put out stable releases as well. An ex windows users needs to realise that these are near beta releases but bugs generally aren't too much of a problem. This is nothing like running beta windows releases as I have in the past or beta windows software releases. I haven't used windows at all for a very very long time now. I can't say that that has caused me many problems and bugs I have reported have been fixed. I understand KDE5 has emerged. It's a totally different approach. I'm not keen on anything that must use that until it's clear it is where it's going and is getting near bug free. Could be that INDI might need it. I put in a feature request on INDI. A web interface for the lot with the set up split onto several ip addresses - the reason for the separate IP addresses is the screen size aspect you mention also if a tab in a browser doesn't have focus there is sometimes no need for it to update which will help keep the bandwidth problems down as well. I'm after a more or less portable set up though and no PC. It seems like they are looking at a thin client approach but it doesn't appear to offer this sort of thing, KDE 4 varies according to the distro it is on. It's been pretty good from day one on OpenSuse. They sampled some of it when most were using KDE3 still but it was hopelessly buggy and easily removed. Unfortunately when I have managed to build an INDI of some sort Kstars falls over when I add it, even the original KDE version but it's ok installed on it's own. I still have a few things to try and some one is working on rpm's for it but only bleeding. For me if that means chucking KDE4 for KDE5 it's a none starter at the moment as I use my PC for all sorts of things. It will be interesting to see what's in the astro distro. Some one who is taking the trouble to do that probably wont be too keen on getting into changes every 5mins. I'm going to install it on an Atom netbook with 1gb or ram after doing a little more with what is already on it. People do not have to take the latest upgrades - if it isn't broke don't fix it, just keep what you have around some how so that you can re install if needed. Source is ideal for that or the various package formats. I'm also going to install Debian and it's INDI package at some point. From what I have seen that may not work out either. That would really be weird if that is the case. John -
  4. The INDI/KSTARS/EKOS package will control lots and lots of things. It seems to be determined to be bleeding edge all of the time though which I find a bit disappointing. I often compile from source and have had some for me unusual problems with INDI as well, compiler errors and things not working out even when all of the dependencies are about. Could be it's a bit distro incestuous. Some one is working on RPM versions though. Main reason for posting - just found this http://www.distroastro.org/?id=select They seem to have grouped the lot together to stop it disappearing up it's own back side. I've not loaded it yet and people may get a "may damage your computer" warning when the page is loaded. I have seen more of those lately - not windoze etc I suspect and has SOFTWARE on it. Another interesting couple on ARM linux. I want to get these running on QEMU 1st. http://www.cloudynights.com/topic/385710-piastrohub-raspberrypi-for-autoguidingdslrgoto/page-24#entry6490467 http://maxchen.im/tinyastro/en/ The strange devices mentioned are TV dongles. The correct ones are available as far as the processor goes from Ebay for about £35. Can't make my mind up. Their wifi range is about 5m and I need more and a USB hub is needed. Then there is the Odroid. The 'pi 2 is slower but it may be easier to fit a more powerful wifi to it and still 3 USB sockets available. These 2 can be controlled by a tablet remotely. There are various ways of extending things on the end mentioned on the INDI site if INDI is involved that will also run on a tablet. They are looking at making INDI into a sort of thin client too but I suspect they are hung up on involving a PC somewhere. Actually there is a lot of Linux software that is being ported to ARM. I've stuck with KDE right from a very early version. Only problem with 4 was the disk indexing. I had been wondering why my disks kept tinkling so turned it off. Kmail ? Not a problem as OpenSuse built KDEPIM3 into 4 and it's still available on the latest version. Kfind too and the K xml editor. These days I only know that the indexing is there because of the pregnant pause when I load up a file from disk. John -
  5. A sharp idea callump, straight to the point. I like that. John -
  6. Low breaking strain fishing line too but I've never had to do it. I wanted a cross hair eyepiece for aligning a scope so bought a Chinese 10x microscope objective with cross hairs- Ebay. I will need to turn up a sleeve to convert it to 1 1/4. For the price I was surprised by the quality. It's an eyepiece for a compound rather than a stereo microscope so the field of view is limited to around 16mm. Focal length 250mm / magnification so 25mm in my case. They have an accurate scale on one of the axis so it might be possible to use them for polar alignment. I didn't fancy the price of the astro ones for the amount of use it will have. John -
  7. I think you might need to post again - choose add attachment - select the photo and then tick add to post. John -
  8. I have the pole star problem as well - iOptron doesn't help either as it's used to set up the mount. One thought I had to align quickly once it's been done properly some how is to use a pier rather than a tripod but It's probably possible to do the same sort of thing with a tripod. Just locate the feet so it always goes back in the same place. In my case I have to make an adapter to fit the head to the pier so when it's broken down that comes off still fixed to the head. Then it's just a case of locating the adapter on the pier tube - a tapered pin etc. iOptron, guess I will have to take it somewhere and align the gears correctly. It doesn't have a clutch so the worm wheel teeth have to be aligned to get the axis pointing in the correct direction in it's home position - at the pole star. No scope level pointing east west with this one and mount aligned north south which on other mounts gives a decent start. I've wondered about trying to use a Celestron SkyScout to roughly align it. Some people it seems have used these as a sort of go to but I'd guess any metal close to it might mess it up. John -
  9. Some of my post was for people who might be thinking about making a mirror Damian. I mentioned the full sized lap because of an earlier comment about the centre polishing slowly. The usual way round that with big mirrors on the bottom but generally done with a machine is to use a smaller lap in the centre and forget the initial pure spherical mirror. It's possible to measure what will loosely be 2 radii this way as it's done to make sure it isn't too deep for the eventual parabolic shape. I have a similar problem with an F3 mirror for different reasons - the figuring depth. Trouble is I am not sure if it's suitable for hand work as the "cenral sphere" might be off centre compared with the outer one. In real terms they wont be distinct spheres but measurements can still be taken. RAC's mirror is a nice piece of work. The software he shows is probably one of the ones on here https://stellafane.org/tm/atm/test/software.html I know for a fact that I can't make measurements as accurately as RAC can. I don't do it often enough. I've tried all sorts, masks and shadows touching sticks, centre, 70% and edge correct and the rest smooth - not too bad that one and etc anything else I came across. In the end I followed Schroder's caustic test in ATM III. The others were fine for rough figuring. A lot is made of the difficulty of making Schroder's test rig but really all that is needed is a Tex' one sitting on top of another so that things can be moved in an X Y fashion and the knife is replaced with a wire. I viewed it with 10x eyecup loupe as it was easier to mount than an eyepiece. Nothing ambiguous about this one at all. It's relatively easy to position of the wire to better than 0.001in. I tried the Dall null test as well. Had problems but later found that it needs setting up pretty precisely by ray tracing it. It's probably better used to null the mirror and work back from the position the bits finish in or to null the mirror looking for zonal defects. A lot of this area depends on how accurate the mirror needs to be. There is also need to consider the source and knife / wire spacing when high accuracy is needed. Big mirrors help with that a bit. My F3 mirror is on the back burner. It took a long time to come up with a design for a flat field cassegrain. Finally did it and concluded that I couldn't hope to make some of the small glass optics accurately enough for it to be a good as it should be. It will probably turn into a Dall Kirkam at some point. Not sure. John -
  10. I'm surprised you haven't made a full sized lap for producing the sphere. It works well especially with mirror on the bottom with a random mix of over centre and mild w strokes with say 1/4 dia long strokes. Plus random steps round and try and be random with the rest as well. Mirror on top tends to bring up the centre quicker though. Tool on top tends to wipe up towards the edges of the mirror which is why it's best to finish off this way to avoid any chances of a turned down edge as most of the light in an image comes from the outer parts of the mirror, pi r^2 etc. Some one mentioned using the fringes ahead of a knife edge in front of the ROC earlier. This work rather well and is probably the best way to check right up to the edge. It needs a slit source to work well though - around 0.04mm from memory over a hole say 5mm diameter. When you get to a sphere with a source like that you will find it very difficult to cause the knife edge to grey out the mirror evenly but that too is a very sensitive test but will usually give bright diffraction rings around the rim if it really is spherical making it difficult to see what's going on there. The knife edge has to be set very precisely. Texereau did that by swinging it on an L shaped arm with M6 bolt on the end and mounting the knife edge so that it could be tilted to square it up with the slit. I did it the other way round and tilted the slit and as it was separate lost it. The diffraction effects can be reduced by widening the slit but that can also mess up accuracy. There are various ways of making a lap. Very big mirror makers cast squares and stick them on. I've followed a book by Muirden and it worked for me but it assume the mirror is transparent. The mirror and tool need to be warmed up, not too hot to hold but sort of warm and cosy style. Pitch melted and then stirred while cooling, lift the stick out and let some drip back in. It needs to take a few seconds to melt back in. Pour an even spiral over the tool. Let it cool until it takes a finger print without sticking. Cover the mirror with plenty of polishing powder mix and place it over the lap while keeping it moving around - w and centre over centre stokes etc. If the mirror is transparent lack of contact can be seen. I've managed to get full contact this way. If not and the pitch gets too cold trim up and hot press again etc. I cut the channels with a wet tennon saw. I've tried the other ideas, very coarse draining board mats when they were available, didn't really help and the channels still needed trimming up with a saw. Also a hot knife but it leaves a lip. One thing I did find which helped was pressing with a piece of fine plastic mesh over the lap. This gives a lot of very small facets and I have done this on any size of lap I've made - onions used to be sold bagged with it. A good coat of polishing solution prevents sticking. I've never done anything as big as this and suspect I would bite the bullet and cast squares or maybe cast a sheet and cut it up. I think a sheet would be easier to maintain a constant thickness. Cutting it up pass but I suspect I would give it a try. One thing that does help with pitch - it sticks well to warm surfaces. John -
  11. Does anyone know how thick their mirrors are at various sizes? I'm probably most interested in the F6 200mm. John -
  12. I couldn't stand the wait so bought an ebay 321404721059, 1st class post was one of the reasons, postage and packing will have cost the seller £1 assuming their jiffy bags are really cheap. They seem to sell a lot and it looks like a well made board to me. They are 30p cheaper than when I bought mine. John -
  13. It's better than most gaj as it mentions gashing. The actual worm form he produces isn't as wide as the blank is thick which fits in the the home made one I posted. I have made a tap like that and it was very difficult to clean up the cutting edge after it was hardened - I couldn't get it dead sharp before that, there was always a bit of a burr. I also think that making the worm out of silver steel / drill rod is just making life more difficult. It isn't the easiest of stuff to get a decent finish on but cutting oil helps. I would make the hob separately and try to finish the worm off between centres so that if the diameter needed trimming down later or the thread recut a little deeper it would be easy to do. I'd just use ordinary free cutting mild steel or free cutting stainless for that. To be honest I think better results could be obtained by replacing the slitting saw with a fly cutter carrying a small tool with the form that is needed on it. The rad of the fly cutter used ideally needs to be a bit larger than the worm's but simply cutting straight through the teeth on the worm wheel works well. Lapping will generate a more exact form anyway especially if the size relationships is similar to what I suggested. There is plenty of information on making and using single point gear cutters on the web. Either the worm has to be tilted or the blank for the wheel has to be tilted when it's cut this way but I feel it's far more likely to give accurate results. John -
  14. You might be able to find some cork mats to use for contact points on the mirror. Just cut small pieces stick them to the locations and adjust so that they grip the mirror lightly. As cork varies I suspect that the gasket material that's around 1 1/2 to 2mm thick would be best. It's cork binded with rubber. The natural stuff may turn out to be too soft unless you can find some thin dining table type cork mats but they may be too thick. Put 3 pads under the mirror as well. At this size and even larger I don't think you need to worry about tube currents especially with a wooden tube. The main thing is that the air is stable in the tube. You could simply cut a hole through the centre of the wooden part of the mirror holder to allow air to pass through the tube if your concerned about this aspect but looking at your holder there is already plenty of space for air to pass through the tube. John -
  15. Bit off topic to the thread but it struck me that people put up with long exposures on refractors with relatively small apertures but no signs of the use of smaller sizes of reflector. People tend to go big and short focus for photo/visual use but that brings down the maximum field angle that will have more or less round stars. Against that people pay a lot our for a 5 or 6in F6 APO. I agree 100% about fork mounts - No counterweight, a big advantage. John -
  16. I'm wondering when some one will do a quad 6in F6 newtonian. Could you tell us where the steppers and gears come from Boyd, say an ebay number if any came off there? John -
  17. I'll 2nd that. I have seen lots of web shots of people making worm wheels like that but always wonder how easy it is to do and get the teeth to match up once the disc has completely rotated. Not my work but an example of a home made worm wheel made in a different way if some one has the kit to do it. These are made by cutting ordinary gear teeth at a slight angle to match the worm. The width of the teeth is around 1/4 of the diameter of the worm or less and the tops of the teeth are concave. It needs a dividing head but people use a gear for indexing and make up a simple one. It also needs a hob to cut the teeth. The problem with that is buying one that can match a worm that can be cut on a lathe so people often make a hob. There is a decent run down on making one here and there will be plenty of others on the web. http://www.helicron.net/workshop/gearcutting/gear_cutter/ The other way of cutting the wheel is with a fly cutter - a single toothed cutter with the same form as is used to cut the worm making sure that the two mesh on the flanks of the form and that there is clearance at the ends of the teeth of both. Rather than cutting the teeth at an angle the worm could be angled instead. Accuracy can also be improved by lapping the 2 parts together with a very mild abrasive. It's the way I would do it but I have a lot of other things to do so cheated and bought a ZEQ25GT instead. Weather has still prevented me seeing how good it is and lack of adapters means I haven't topped out how much load it can take yet. When one off worms wheels are made professionally they part precut the teeth with a slitting saw so that there is something to ensure that the hob or tap has an accurate guide as the blank rotates. John -
  18. I've used some B&Q chalkboard paint. It's water based so easy to wash of the brush. Seems to stick to anything but may not be so good on smooth plastic. Often needs a smear 1st and then a final coat. Labelled Flunky Colours John -
  19. Ahhhhhhhh I see you have swapped the motor. Personally I hope the servo arrangement normally in them is stable enough. Some one has used the continuous rotating type where the mark space ratio sets the speed they turn at. Not sure I fancy that option as the servo positioning aspect is lost. John -
  20. I wasn't clear on the servo or stepper motors Boyd. I got the impression you were going to replace the servo's with small stepping motors. As you know I'm curious about the servo approach as well so wonder if they have problems and if you have replaced them. Must admit I'm thinking in terms of using servo's in the normal configuration -- an arm that swings through circa 90 degrees rather than converting them to rotate or buying that type which are bit thin on the ground. John -
  21. Looking at the spec of the smart guider I suspect the auto focus just performs temperature compensation. Focusing is what has put me in a bad mood. I finally found out what is in a 2 channel arduino unit on the INDI site. Best say no more but the word stupid comes to mind. John -
  22. Thank you for mentioning the smart guide Boyd. I might just go in that direction. I'd be happier if it would interface with other cameras but as Linux telescope software mostly seems to live on the wild side, latest greatest, compile errors and other problems I might look at what can be done with it easily. An arduino actuating the buttons on the controller etc - no probably only kidding but I'm feeling really cheesed off having just used Linux for over 15 years now. I'm embarassed by some of what I see so probably wont ever mention open source to anyone again not in this area anyway. John -
  23. Interesting. You will probably have to do some sort of PEC to linearise the encoders at some point based on the spec's. RS seem to have discontinued this one but do a different type that uses contact via a spring loaded plastic ball. It's the same sort of principle - the ball presses on the outside pushing internal "tracks" together. No hysteresis on these. It might be possible to attract and generate the pressure with some form of ball via a magnet working on the other side of the encoder. I came across another type of encoder. It's a hall device and detects rotation of a magnet above it. There were some pages on the web concerning "super hacks" for model servo's. This is one form http://uk.rs-online.com/web/p/mechanical-rotary-encoders/7142693/ This is the chip that does the work but 16bit http://uk.farnell.com/avago-technologies/aeat-6600-t16/encoder-magnetic-ic-16-bit-program/dp/2213639 The problem as always though is that they aren't linear so the claimed accuracy on the servo hacks is a bit misleading unless they correct it some how. There is an interesting little program here for printing optical code wheels, also plenty of information on their use http://www.mindspring.com/~tom2000/Delphi/Codewheel.html#Introduction I have sometimes wondered if 2 could be used on an axis. The 2nd one geared up to interpolate the readings of the other one. It's also possible to use more than 2 optical sensors to increase the pulse rate. The quadrature one shown has 2 displaced by 1/2 an interval on the encoder wheel. Other fractions can be used as well. John -
  24. May as well add another aspect to home made mounts. Sometimes aluminium is a better option than wood. It can be routed in the same fashion as wood even using the same cutters or milling cutters such as end mills and slot drills. Some what lighter cuts are needed though but it can be done. Really the speeds are too fast for milling cutters but if all else fails........... John -
  25. That is one of the reasons I posted a link. It's not a bad implementation of it but notice how thin the lower part of the ring is. That part will twist in use unless more bearings are added to the front side rather than just at the rear. Or maybe the ring could be made a lot more substantial - maybe fabricate it from 2 spaced discs. There is a commercial one that costs a lot and uses a circular bent tube but I would wonder just how accurate the bend is especially as it forms the drive. You could probably find pictures of this type of mount used in observatory instruments - the ring is massive The other problem is the name - this arrangement is usually called the horse shoe mount but not here. The split ring of the other sort is sometimes called a horse shoe mount as well. http://www.dalekeller.net/ATM/newtonians/keller10/keller10b.htm That one is vice free but and English mount is far easier to make - more or less the same thing but rather than having a ring on one end it just has a bearing. So really it would be best to post some links to what you have been looking at. Friction drive is likely to have problems with fast slewing. One way round that may be to attach a length of toothed belt to it to convert to a geared drive. The bearings would then need to be gears as well. Faster slewing with steppers is usually obtained by a combination of micro stepping and the speed that the motors are switched at. You would need to look at the changes in torque characteristics when micro stepping is used. You might say 1/2 step every sec for tracking and full step as fast as the motors can manage for slewing. The output from the steppers still need gearing down to obtain torque and drive accuracy. DC motors are switch mode (pulse) driven at some fraction of their max speed for tracking and driven continuously for max slew rate. The encoder is usually on the motor and the drive from it is reduced via gearing so that the encoder's pulse outputs come often enough to obtain a steady speed. In the past the encoders have been driven by the actual drive output but via gears again to up the speed. People have even used optical encoders out of mice. Direct shaft encoding mounts are available commercially but they use encoders that output fantastic pulse rates per turn and are also very accurate. In the case of your 0.2 degree servo motors gearing would be needed on the output to make that accurate in drive terms, so for say 1 arc second it would be 60^2 * 0.2 to 1. There is another style of mount that I feel could be used more - an example here, fork mounting http://strock.pi.r2.3.14159.free.fr/Ast/Art/Ouranos.html Same problem - massive parts but that could be overcome by fabricating them. It doesn't show any details of the main bearing. It will be a steel bar with a point on the end. That rests in a hole in a metal plate and the other end will be supported with 2 metal squares with a V in them, the V's making contact with the bar. Replace the plate with a hole with a plate holding a spherical bearing and it would be fairly easy to add adjustment to it. John -
×
×
  • Create New...

Important Information

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