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.


  • Content Count

  • Joined

  • Last visited

Everything posted by Ajohn

  1. He has clamped the diamond cutting blade into a cup shaped holder in the drill so that it bows the cutter and forms a slight cone. Some one on the web came up with this idea to shape glass on a milling machine some time ago but removed it all due to the risk of breathing in glass dust. Worth bearing that in mind. Years ago Sky & Telescope showed another way of doing it. Stick the grinding gizmo on the end of a rod to get the right radius, support from the roof and swing it around over the mirror to generate the radius in it. Needs some form of simple radius adjustment and care in aligning with the centre of the mirror. This used an angle grinder and a hose pipe - very high rpm so rather messy. Best way is probably on a drill press but needs a minimum size of cutting disc. It's possible to work out what angle the disk needs to be at to generate a certain radius when it's plunged into the glass. As it's not always easy to tilt the spindle on a drill press the glass disc is tilted instead and rotated by hand with the drill running at minimum speed. Breathing in the water mist is likely to have it's problems as well. Easiest way to find out more about it is to search cloudynights for diamond generating and maybe sine tables. Mostly used for generating lenses with diamond core cutters. Rather large cutting discs can be found so mirrors are equally possible. An old pro way - an attachment to fit onto a grinding and polishing machine. The grinding spindle runs across a former shaped to produce the required radius via a fine feed of some sort. Loads of water and a settling tank as glass dust is very good at blocking drains. I read an account where 3 tonnes of glass was removed. I have heard a rumour that some are playing with cnc to achieve the same thing. Must admit something knocked up along the former idea and really enclosed sounds attractive if some one wants to make a number of mirrors. Correct speed for the cutter is about 3,000 ft/min so enclosure is essential really unless some one is tired of life. John -
  2. The way to handle a Boxford is in pieces. Mine is an ME10 and that has a rear drive rather than the underdrive on the other models. The head is held on by 2 bolts on all of their models. Taking that off and the tailstock makes it a lot more manageable for 2 people. The saddle can also be taken off. The stands on most lathes if they have one are always heavy. Personally I wouldn't buy a Myford 7 unless I saw it working and it produced a nice smooth parallel finish. It was a good option when the bed and saddle could be sent back to Myford for a regrind and saddle fit if it was needed. 2 people can manage a Myford 7 off it's stand but it's easier with the tailstock and saddle removed. The headstock is best left on. The Myford ML10 and Speed 10 are better options in a number of ways. For their size they have an adequate distance between centres so there are unlikely to be problems drilling holes. I owned a Hobbymat for a while - good job it came with a lot of stub drills. I also repaired the screw cutting drive on it several times. The main gain of the Speed 10 over the ML10 is a higher maximum spindle speed. That makes it easier to hand feed when turning very small parts but really for that sort of work 4 ot 5,000rpm would be a lot better. If power feed is used that aspect doesn't really matter. On my Boxford for instance it's usually running at 500 rpm and I make M2 screws by sizing them in one cut. It's a case of buyer beware buying used lathes. Even Boxfords can be worn out. The main gain is a back gear. The Chinese lathes minimum speed is usually 100 rpm and that can be far too fast for screw cutting at times. Back gears can reduce the minimum speed a lot. For instance the min speed of my Boxford is 38 rpm. That is still fast for very coarse threads really but these seldom need cutting - in my case anyway. Boxford's pluses over a Myford are a better over all design - especially the bed and a proper power feed system. It can be a problem obtaining the gears for a slow feed speed on a Myford. The Boxford doesn't use the lead screw pitch for power feed so an imperial one for instance will feed at 0.0015 per rev as it comes if it has a screw cutting gearbox which they usually have. They also tend to wear better than Myfords. The ML10 and Speed 10's often haven't been used much. Old books - nothing has really changed in respect to manual machine tools from way way back other than there being more options available re carbide tooling. More recent copies of the Machinery's Handbook even give details of metric threads. Many other types too. All the other data in them is still valid as well. These days CNC is usually used. I started with a Peatol ( Taig in the USA). They can do very precise work until the heads bend. Unimat next for something bigger. Then a Myford ML7. Then finally the Boxford ME 10. There way also a Raglan before the ML7 but as that was exceedingly heavy it had to go in the garage and rust was a problem. I also fitted the ML7 with a Super 7 head. All over a period of over 20 years but I must have had the Boxford for over 10 now. It would probably have been better to get it right 1st time. The initial reasons were bits for small telescopes and odd bits and bobs. More recently bits for microscope and other bits and bobs. I need to modify modern tap heads shortly so that we can retain our now rather old taps on the kitchen sink - or replace the entire sink and unit it's on. The sort of work a lathe can do? Lots of components so that it could be made. Some think I polished it all or used emery cloth.. John -
  3. There are a large variety of Philips web cam mods. The best link I had which was from some one in the uk now seems to be dead,. A search bought up a cosmic shed blog. That rings a bell. The pages covering several mods. From memory various mods were numbered sc1, sc2 and so on. That may help find details on the web. It also had a link to a German company that would sell one off sony CCD sensors. Price wise and some years ago a larger black and white very sensitive ccd chip cost was about £20. I believe there is also a more sensitive direct replacement sensor but the low light cameras have sensitivities of <1 lux. There seems to be plenty of info still kicking around on long exposure mods etc. I may have already mentioned the Tucsen 3mp microscope camera. I sold mine for something better and it isn't. I know from using it on a microscope that it's more sensitive than my eyes. It uses a modern cmos 1/2 in technical camera sensor. Unless it's changed the software allows direct access to colour channel and chip gain and these really do have to be used for good results. They also do cooled ccd camera that are capable and much cheaper than some others. John -
  4. Ajohn

    DIY Lenses

    Much to my surprises HV Scan is still active in the UK as Schotts agents. I used to bother them mercilessly for glass quotes. They don't seem to mind. The glass is expensive. BK7 which is basically ophthalmic crown but "better" is the cheapest. Flints are generally dearer by a factor of 2 or more. The prices for glasses for things like F6 apo's might even work out more than buying a scope. The other problem is that some of the glass used in these is attacked by water making grinding interesting. Newport Glass in the USA used to list kits for making refractors and most people in the USA seem to use them often asking them to preform the elements so that they "just" need fine grinding and polishing. I believe Scan can organise that too. People in the USA make some huge refractors some detailed on cloudynights. Looking at spot diagrams I often wonder why but it's hard to say having never looked through one. There are a number of designs on http://www.telescope-optics.net/ Some introductions to using Oslo on here http://www.atmsite.org/ What seems to be lacking in the UK is companies such as the ones listed in the old uk atm resource list who will sell optical glass offcuts etc. Perhaps some one on here has some ideas on this score. I tried to attach one of my Oslo designs for a 300mm dia F14 Steinhill for you to play wth but it wont accept the file. John -
  5. Ajohn

    DIY Lenses

    I would second the Muirden book. He advocates a flat back design for a 1st attempt which gets round the final figuring problems as they are done in auto collimation using the flat side of the tool as a mirror. Best read the book to see how that is done. Alternatively it's possible to produce a mirror the same way with say 2 6in mirror blanks. Getting the glasses mentioned in the books can be a problem. Manly because Schott F4 was popular and is no longer made. Best thing to do is to ask some one on cloudynights in the atm section to work out the best curves for using say BK7 and F2 giving them the figures out of the book to start off with. You might like to download Oslo Edu and ray trace it yourself as well. Near the end you might find that some of the curves or thicknesses are out. Again cloudynights is the best option as some one will calc the best changes to make to one of the surfaces to correct the error. There is one person in particular that often helps with this sort of thing and that way you get access to the best software that is available and some one who knows how to use it. If you do go for a flat back some will say WHAT! etc but if that's what you want to do stick to your guns. John -
  6. There are some black and white C-Mount video cameras about that stack images internally for low light work without flooding the scene with light via a lot of led's. There are board cameras about with the same type of electronics on them as the spec you mention without the led's. These have the same lens fitting as the one on the usual web cam astro adapter and are probably a lot cheaper fitted with a high spec chip. The web cam used for the star shot is probably one of Philips's with a ccd rather than a cmos sensor. They did 2 versions one normal and one low light. I don't think anyone makes a web cam with a ccd sensor any more. All sorts of things can be done to the old Phillips cam including fitting it with a larger sensor. There is also software about to extract raw and do a number of other things. Some other manufacturers also produced cameras based round the same ccd sensor but it's important to get model numbers correct for the low light versions. Could be that old xbox cameras used ccd sensors as well but while I was interested in the subject I never saw any mention of it. Not really on topic but I have seen good moon shots taken with an Olympus E-PL1 but looking at ebay prices the days of super cheap because a new model has come out are gone. The E-P3 seems to have better low light manual focusing performance when used as a normal camera even though the sensor is the same. I took some 10 sec sky exposures with an E-P3 out of curiosity with a camera lens. It seems to stack internally so may not be of use for star type work. On the other hand these cameras have a vast number of modes. The original canon digital rebel, 300d in the UK has an amazing low light performance. To such an extent in some cases that one UK retailer was selecting them for astro work at no cost. Some microscope cameras have been used for astro work. The best bet is probably the Tucsen 3mp camera. Just some "lower" cost options. John -
  7. It isn't difficult Dave. Many people who know absolutely nothing about machine tools buy one and teach themselves with the aid of books and various web forums. If you search "workshop practice series" on ebay you will find very reasonably priced books on most aspects of things that might interest a model engineer. The usual problem people have with lathes is cutting speeds. They read that this and that material has to be cut at so many ft/sec surface speed or whatever the metric equivalent is most of my books are old. What books don't point out is that these are largely related to doing a job as quickly as possible with an acceptable rate of tool wear. They also indicate that some materials needs to be cut more slowly than others eg marine grade stainless steel, cut that too fast and it work hardens. In practice many lathes produce a better finish at much lower speeds. Low speeds are not a problem providing a lathe has a sensible power feed. Aluminium can be a problem too. People might find themselves using high speeds and the aluminium cuttings melting and sticking to the tool. This isn't obvious when it's happening but leaves tell tale smears of aluminium stuck to the tip of the tool and produces a very poor finish. Books mention a huge range of tool cutter angles. That style of carbide tip I linked too will cut all of the usual materials well. The ones that come with ebay sets are ok too. The main thing about cutters on a lathe really is setting them correctly. An external tool needs to be set as close as you can get it to the centre height of the lathe but not over it. Things are the other way round when boring holes and boring bars bend so the tool usually has to be set at a height to account for that as well. The other problem is what lathe to buy. 2 things to check on Chinese lathes. One is usually a fact of life. Spec's give the swing over the bed but often the maximum feed across the lathe isn't sufficient to turn over that size of item. Usually they will turn over the other swing that is mentioned over the saddle. Some wont turn up to a centre in the tail stock. That can be fixed by using a thing called a morse taper extension but it annoys me as it happens because the spec of the lathe looks better if it can't. People buy lathes on the basis of centre height and distance between centres so there have been instances where importers tell them to increase it even if it causes this problem. The people who sell them usually do know about any problems like this so asking about them may help. For screw cutting a lathe with a screw cutting indicator is best but people manage to cut with the screw cutting drive permanently engaged it's just a lot more difficult to work up to a shoulder as the lathe may take some time to stop when turned off. People have been known to fashion a handle to turn the lathe spindle by hand to get round that. Various books are often useful to people like me who had a very thorough toolroom training. In a toolroom the perfect machine for the job is usually available. At home people have to cope with what they happen to have. John -
  8. There are a whole series of books that can help with various aspects of "engineering" at home that go by the name of the "Workshop Practice Series". The same people also publish several related books. One thing that these cover fairly well is anodising in the booklet on plating. They give a range of current densities to use and I find that that aspect is more critical than some web pages suggest. Best experiment on some clean scrap. I used lead flashing for the anode/cathode and should remember which one it is. I cleaned it up with wire wool and then etched it a little by leaving it in white vinegar for a while. Good quality fountain pen ink can be used as a dye. I mostly use hss tooling on a lathe as I can get better results on most materials than carbide tips. This is basically because I polish them with a fine slip stone. Diamond lapping sticks can be used too. I started working for a company where everybody did several years of toolroom work before anything else and the slip stone came from that. Some one on ebay lists 5% cobalt hss tool bits and that is a decent grade to go for. Many people buy 6in grinders but these are too small to make a really good job of grinding tools especially with the wheels they come with. Axminster sell some white aluminium oxide wheels that work remarkably well on these machines. Water is still needed to cool the tool bit from time to time but the wheels cut hss nicely if a bit slowly. They also do a hammer headed wheel dresser which is very easy to use. Some one on a Schaublin group suggested trying a particular type of carbide tip. They do work rather well. Personally I would avoid buying these off ebay as I feel sure that many are part worn. Better to buy from here. These are the type of raked finishing tips that were suggested. http://www.shop-apt....um-uni-tip.html They work well on aluminium,stainless,brass and steel. A cutting oil helps on stainless and the others sometimes too. Morris Lubricants will sell 5ltrs mail order. A near life time supply as all that is needed is a very light smear with a lightly wetted brush. I use £ shop kiddies paste brushes. The wooden handled ones last longer. It's worth nosing about the site as they have offers. The same style of tips are shown in hobby tips as well. These will fit the ebay small shank tipped tool sets. Best buy those with Torx screws as the hex socket ones soon round over. This site also carries spare screws. I often use imperial threads where I can as screw cutting metric ones often involves reversing the lathe which is a bit of a pain. If a lathe has a screw cutting indicator one easy way is to disengage the drive, turn the lathe off, reverse it and re engage the screw cutting on the same index mark. The same thing can be done when the tool has traversed back to the start if needed. If buying a typical heavier duty model engineering lathe the Myford name usually crops up. Super 7's in particular. Bearings are usually worn on ML7's. These can be good and are heavily featured in just about every model engineering book. In practice however they have their problems usually associated with wear particularly in the bed and having owned several lathes I feel Boxfords are a much better bet. Usually cheaper too but one with a T slotted cross slide can be hard to find or that will have to be bought separately if needed. The Myford Speed 10 and ML10's can often be found little used and with lots of bits and pieces. The bits and pieces that come with any lathe are important. If bought later costs tend to rocket. One interesting thing about these myford xxxxxxx 10's is that they often come with imperial screw cutting gears. Buy a couple more and they can cut metric too. They also did metric lathes as well. For a variety of reasons I'm not too keen on metric machines so stick to imperial. The reasons mostly relate to divisions of an inch against what can easily be done on metric machines. John -
  9. I wont be posting any more on this topic Alan. Posts tend to be long on subjects like this and take a long time too. My "agenda" is followed by the majority other than some that use Ronchi plus a lens. How people want to go about making a mirror is purely up to them but I think it's only fair to warn them about pitfalls. That actually is the only reason I post. As some one else pointed out they hope you have achieved much better than a 90% correction. I do too. All I have done is point people at methods that will help ensure that they do better than that as long as they shoot for reasonable F ratios and not so fast that things get difficult. Probably a lot better with a little care. There are catches even with that one - chasing minuscule amounts of glass but again that is up to them. The other catch is too fast an F ratio for the techniques to be easily used but again that is up to the individual. John -
  10. You will probably be fine for wide field use Alan or not too far off it. A lot depends on what you have looked through before too. Some old books mention amateur mirrors taking on fine curves cooled down. I have my doubts about under correction being fixed. John -
  11. If some one is thinking of buying a drilling machine the 2 on the far left top line are similar to the one on the end that I have had for about 15 years or more now. http://www.chestermachinetools.com/hobby-drilling-machines-835-c.asp The same machines can be found at all sorts of prices. The 5/8in chuck on mine has lasted. I like the round table types as it easier to line up before drilling when work is clamped to the table or in a drill vice. The whole table can be swung in an arc and rotated making that aspect easy. A decent drilling vice is a good idea as well. I have an Indian made one like ebay 310337079949 except it has a handle like ebay 370371191156 by Soba. ArcEuro have some premium drill vices in stock but no pictures. They are well known for finding the right thing at the right price. I think Rotagrip may still stock the type I have. People do use those for light milling - plate etc. John -
  12. Most of the large capacity pillar drills that are cheap are aimed at wood work. There may be a 5/8 capacity chuck on it but the power is more suitable for 3/8 maybe 1/2 in in steel. Aluminium takes more effort to cut than wood but most materials can be drilled with them providing the size is taken up in steps remembering that the steps need to be smaller as the size goes up. The real answer to chuck as against drill sizes is a set of blacksmith drills. These have 1/2 in or the metric equivalent shanks and go up to 1in. Many lathe owners wouldn't be without a a set. I am not going to comment on using large drill is a hand held drill electric or otherwise. Work needs to held well and the hand drill needs to be suitable. Costs may well exceed a reasonable diy pillar drill. That may have a large chuck and no morse fitting so the chuck that comes with it has to be used. Keyless chucks are wonderful but even at the cost of a decent make like Rhone they don't hold as well as a key operated chuck and there is only one reliable make - Eclipse. The do keyless replacements for hand drills which aren't bad and a keyed one may be hard to find. The keyless ones tend to work at their best when the shank is pushed right back so that it rests against the back of the chuck. Chuck keys come in a variety of standard sizes. Unusually the wiki doesn't have a page on them but there should be info some where on the web. If some one wants a strong chuck for a lathe or drill with a 2 morse fitting the best bet is a used Eclipse off ebay. These and the 3 morse ones were largely made for use on lathes and are designed for a hard life. For a laugh, my metal workshop AFTER a tidy up. It needs another now. There is a small dore westbury miller mostly out of view on the left. John -
  13. No doubt you will enjoy it Alan and don't let this put you off. Trying to get people to look into Foucault though I knocked up a ray trace of a smooth mirror with 90% correction. I comes out with a peak to valley error measured the kind way of just over 1 wave. Strehl ratio 10%. These are the spots. The top one is a perfect 300mm F5.5 mirror. The diffraction ring size can be seen in the shot but I doubt if any one can see the actual light spot. The lower enlarged single central spot is for the 90% corrected mirror. It's interesting in some ways. The rms error as mirrors are often quoted is juts under 1/4 wave. But the other small graph which shows contrast against resolution is no where near rayliegh's limit. Looks like getting on for a 95% correction is needed for that. I haven't included the effects of the 2ndry mirror size but I suspect you have already seen the contrast improvement when it's small. John -
  14. One thought for you Alan that people have used for star testing that involves a plank. Make up an L shape. The short side of the L goes on the ground with the mirror on it. The focuser is mounted on the "far side" of the long side of the L with 2ndry mirror adjacent and on the opposite side. Then wedge the lot up so that it points at polaris. It can be left like that and the important bits fitted when you are working with a bit of thought. A plumb bob and compass should help initial alignment. Whoops never mentioned the hole in the plank beneath the focuser. John -
  15. For people that feel that they must test this way it's probably worth enlarging on Nigel's comment about matching the pattern against the one shown on a print or PC screen. What this actually means is moving the screen precise distances along the axis of the mirror and obtain a match at each of them. Using 6 positions is unusual as far as I'm aware and there isn't really any reason why this should give more accurate results. The distances are usually measured from the precise radius of curvature of the central region of the mirror. The accuracy of the movements that's needed is dependent on the mirror. The total variation in the radius of curvature from centre to edge can be worked out with the usual formulae. The main man on this style of testing is Mell Bartel's and he has a web page complete with an applet that shows the bands for various screen positions and explains how it's done. He also states at the end always test in some other manner as well. He mentions measuring with an engineers rule graduated to 1/100in or 1/4mm and a 4lpm grating as per Tex's brief mention. I wish people luck in finding a finely graduated metric rule. Imperial ones to 1/100 in can be obtained off Ebay from the US site. Bearing this in mind in some ways it pays to make up the usual tester and fit the screen to that rather than a knife. An M6 screw is adequate for measuring to 1/4 mm = 1/4 turn. I suspect his page assume a stationary light source. Best do some more searching to be sure. If the light source moves with the screen the movements will be halved. Still not a problem, 1/8 turn. Some inaccuracies can be introduced by the stationary light source. It's assumes it's dead level with the screen or knife and not too far of the axis of the mirror. Only one analysis package I am aware of account for the distance from the mirror to the light source - it's called sixtest and is for Foucault testing. I need to check but this could mainly be to allow completely on axis testing. A very small light source is used in front of the knife edge. Mirrors can be figured that way but the calculations are different and there are added complications. A mirror only produces a clear image when the source and tester are exactly at the centre of curvature. When masks are used with the Foucault test the apertures in the mask are behaving like small individual mirrors. Much more so than any other similar test, ronchi or shadow edges. The caustic test makes a lot of use of that fact but many people wont spend the time needed to make the tester. It's effectively a Texerau style sliding table with another mounted on top of that at right angles both driven via micrometer spindles. This can be so precise that on imperial testers it is worth having graduations to 0.0001in and that may as well be 1um on metric ones. This tester doesn't involve reading shadows at all. There are details in ATM III. The page is here http://www.bbastrode...com/ronchi.html Later on at the end of the Ronchi page Mr Bartel mentions taking extreme care at the end stages of figuring. It looks great from the images that are shown by the applet but the lines wont look like that in practice. See the video. Like any of this type of style testing the appearance will vary according to the size of the light source. I don't think that they could ever be arranged to be sharp edged though. Nicol's optics mention experience and that comes down to just where in the fuzzy edges the real bands are. No problem as Bartel mentions at mirror making classes as some one there will know how their tester behaves. Hopefully they will have used it a lot. Texereau - I mostly go on about the book because it puts people in very safe hands. There is no harm in reading other books and web pages etc but what he runs through is very sound. Alan's sticky tester for instance. Tex uses a screw feed because any feasible tester will be sticky to some degree or the other. The screw works against spring or even elastic band pressure. Within reason how easily the table on the tester slides around doesn't matter as the spring pulls it back against the screw. It's likely to even function with wooden V's running on copper pipe or what ever. That will just need a strong spring. All the screw need involve is 2 nuts in a hole in a block of wood. 2 to stop the screw from wobbling around ideally spaced say 20mm apart. Make sure the threads line up though. Once some one has made the sliding table they can put what ever tester they like on it. For Foucault the 2 scraper blades slitless type is very easy to make. As I mentioned personally lacking a suitable local metal supplier I would go along to a B&Q Warehouse and buy a strip of metal and a junior hacksaw and a small file off ebay if B&Q doesn't have anything suitable. The strip can be used to make the V's, a small piece for the screw to work against and a strip for the tilt screw to run along. Material steel, brass or aluminium. I recollect that they have all 3. The rod for it to run along can be tube. They sell that too. Lacking a nice long towel rail in the bathroom some might like to buy a hanging rail for a wardrobe. Some have end fixings that are like the ones on towel rails. Cut their length off it and erect the remaining part in the bathroom. Tube is easier to drill hole through than solid bar. Just cut 1/2 way across the diameter along the tube for say 10mm and then cut down to remove half of the diameter over that length. The tube wont crush when you screw it down then either. John - Doh I'm for ever dropping f's. Probably more of them too. One thing that I forgot to mention. Mel uses and indoor star test. What that means in practice is that he uses another telescope with a know perfect mirror and places a tiny pin hole exactly central on it's focal plane and illuminates it. The scope then simulates star light. How closely is debatable really but given a suitable mirror / telescope kicking around probably close enough. Any mistakes in positioning the pin hole can result in over or under correction. Longer focal lengths also help improve the accuracy of the "star light". How good it is depends on the size of the pin hole and the F ratio of the scope. Longer F ratio's will always improve it for the same size of hole. Pass on how "good" that aspect needs to be. Lacking a suitable spare scope the test doesn't really interest me. It's a little like an auto collimation test with a flat mirror and needs very careful setting up. -
  16. Sorry Nigel - saving on typing. It's a surface aluminised mirror out of a shadow graph. These are a projection instruments that engineers use to enlarge things to allow them to be measured.. Decent ones will measure to 0.0001in accurately. When projected through them 0.0001in might be enlarged to 0.1in or more. I think I have 3 of them. 6in diameter and 19in diameter however wishful thinking may have inverted the 6 to make it look like a more desireable 9 in my mind. It tried to remove the coating of the small mirror with a variety of acids and nothing touched it. When I get to a cass 2ndry the 19in one may prove handy. Engineering optics are usually made to 1/4 wave or better and as Muerden points out a dish in a flat of that sort of order doesn't make a significant difference to auto collimation testing. I bought them some time ago very cheaply of a surplus engineering items dealer. Next time I spoke to him he was a little annoyed about the price he charged me and laughed saying if he ever had any more I would have to pay a more realistic price. Just had a look. 6in yes, 19in yes, 9in no sign so I need to go on a hunt. I'm fairly sure some one has been moving things around again. Tidy up's are fine providing some one remembers where they put things. Odd items usually finish up behind other stuff. In my view these aren't odd items. John -
  17. For planetary type use you should be able to run that mirror at least at 300x Alan. A decent star test really needs more than that and often the seeing isn't good enough to allow it's effective use anyway. It reminds me of a mac newt I bought of a well known uk supplier some years ago. I set it up and thought this looks great and then pointed it out of the window. Couldn't see a thing. Why well the 2ndry mirror wasn't aligned with the focuser. I complained as there is no way that the scope could be set up correctly that way. Dealer agreed. The well known German supplier of such things basically responded do it yourself via a star test - see cloudynights. All I will add to that is the dealer decided not to sell them any more and it's very doubtful I would ever buy one again either. Personally given the next major cost of £112 plus VAT for a good coating there is basically no way I would take either of the tests you are going to use as gospel. It can probably be done for less than that but quartz over coating is very effective. I have a flat blank that came coated that way and nothing I have put on it will even touch it. One thing that you can do that does give some indication is to try some terrestrial subject that are at least a couple of hundred yards away and go into over magnification to a ridiculous degree. A good optical set up will still hold a clear image but the contrast will drop. To give you some idea when I received my 110 megrez I found I could stick a 3x barlow and a 4mm eyepiece in it and the image held up. I'm probably a little lucky for one of those. Many scopes wont even focus at the mad often quoted maximum magnification levels as the image is too fuzzy so there isn't really a clear focus point. All scopes should focus cleanly at a magnification well over their diameter in mms. I would say by at least by a factor of 2. Only problem is that this often needs a barlow and some are not very good. Old Teleview powermates are very good if some on happens upon them. There is also a need to know that the eyepiece being used is ok. Do remember that this is a pretty severe test though but a scope that really is to Rayleigh's limit should meet it easily. What often happens is that the image is so fuzzy that it's not possible to find a clean focusing setting. John -
  18. Personally I don't think a Ronchi screen is worth the cost and offers plenty of scope for an unacceptable level of errors even on a mirror that has been shown to be smooth in other ways but if people want to use it fine. I may have an extreme view. A good mirror to me is smooth and shows a max error of 1/10 wave PV. An undercorrected or over corrected mirror is probably one of the worst faults of the lot to end up with. It has to be said that if some one can manage 1/4 wave on a 300mm mirror there is a cost saving over mirrors such as the ones that are supplied by Orion. One thing I don't understand is that given that a mirror has been figured it should be relatively easy to check the correction at least roughly with a very simple set up. I would also guess that a Ronchi test just like all others needs a sensible light source so maybe in this case it isn't. That isn't a guess really. I'll use that word as I haven't tried it that way. One of the shadow analyisis software links mentions the original slit less tester where a single knife edge passes half way across a led. People may run into grief with that arrangement these days as many modern led's have the internal leads shaped to form reflectors. John -
  19. Being my usual helpful self I have managed to find a video to show people what a ronchi parabola test looks like that fits in with my memory of how it behaves. Not many about which doesn't surprise me There is also an article in one of the ATM books about making the test quantitative. In other words taking actual measurements with it. As far as I'm aware no one has made any use of it. It's a bit flawed really as it's difficult to take the actual measurements and also needs more kit. When I get round to figuring my mirror I will try and take some photo's to show that Foucault plus a mask isn't subjective at all. That is some months away mainly because I feel that I have no chance of getting an F3 mirror right without a polishing machine. I do think that the Foucault test will be ok providing a more modern testing set up is used. Lots of comments in that respect relate to old style testers with a large fixed light source. If it doesn't work out there are other tests I can use. Then it's on to trying to make a cassegrain 2ndry mirror and even worse than that coming up with a design of scope that will hold it's collimation. All in all it's a pretty ambitious project. If I find I can't make either of the mirrors then I will use the main mirror to make something else. Probably an F6 newtonian but who knows. John -
  20. 2 posts - sorry a web woopsy. My browser hung up for some reason. The distance an artificial star needs to be away from a telescope varies on size and F ratio. The size needs to be a bit on the small side as well. Ideally under 1/4 of the diameter of the diffraction disc according to some. The problem with testing a newtonian is that it is figured to produce no Spherical Aberration when the object is at infinity and moving the object nearer will introduce SA. If some one wants to do the maths they are here. There are some comments about that make it a lot simpler than this too. N* focal length etc. www.telescope-optics.net/star_testing_telescope.htm Another way that I have read about but never tried is to place a small ball bearing out in the sun so that it reflects an image of it back to the telescope. Some people have used shiny things off xmas trees etc. Personally I would focus the scope on something a long way away 1st to check that the scope is focused on infinity when the reflected sun is imaged as a check just to be sure. In practice I suspect this method is really only of any use for collimation. Texereau uses a ball bearing to get a sub diffraction disc size point source. If some one has the space for a true artificial star test this is a relatively easy way to make source. It can be lit by all sorts of things. Any type of artificial star can be used for collimation. Size within reason doesn't matter. Where pro's have used one for figuring they generally mention a friendly building often some miles away. When it comes to real stars the best on involves filling the mirror with the light from a star eye piece out and using a knife edge on it's image. Bit like the faucault test only that the mirror will grey out evenly. It's an interesting thing to try. The other way is the image inside and outside of focus. Rather subjective really. Perhaps a better test of the whole scope is splitting a double star. Just how close will it get to it's theoretical best. To me this all means get it right while making it. If some one wants an easier check of what they have these 2 software packages might help. Both measure Foucault shadows for you. Really the main thing about a Foucault test along with a mask is watching how the shadows move in adjacent holes when a reading is being taken and having a fine control of the screw that tilts the knife = big knob. The problem with these software packages is taking the photo's but some cameras will work fine held some how close to the knife edge. Web cams usually need their lens changing and the zoom ones going on one I bought recently aren't suitable. http://foucault.sourceforge.net/ That one is easy to get and mentions using a slitless type tester http://f1.grp.yahoof.../nullfinder.zip This link may not work. The file is in the atm_free yahoo group in the files section so if you want it you will probably have to join. This is another similar package http://www.sonderloe...index.php?id=14 John -
  21. I don't see it as an accurate test used that way Alan but have no problem with people using it if they want. I see it as the nichol optics links states. It can produce mirrors within Rayliegh's limit with practice. I have also read somewhere, probably in a rather famous book on testing that more has been written about Ronchi testing than any other method and then dismisses it immediately. Lot's of people use a Ronchi screen for figuring but with some form of null set up usually a lens so that the lines look straight when the mirror is correctly figured. The most used null method is the one by Ross which in real terms is just a variant of the Dall test. The latter man also used a Ross type lens but in a different way. Both of those options are too complicated / expensive for most people. I see no point in using a Ronchi screen for checking for a sphere as with a stationary source it can be done much the same way with a Foucault tester except diffraction effects are used. This can probably be done with a moving source too. This test can be very precise and to me that is important. John -
  22. I don't see it as an accurate test used that way Alan but have no problem with people using it if they want. I see it as the nichol optics links states. It can produce mirrors within Rayliegh's limit with practice. I have also read somewhere, probably in a rather famous book on testing that more has been written about Ronchi testing than any other method and then dismisses it immediately. Lot's of people use a Ronchi screen for figuring but with some form of null set up usually a lens so that the lines look straight when the mirror is correctly figured. The most used null method is the one by Ross which in real terms is just a variant of the Dall test. The latter man also used a Ross type lens but in a different way. Both of those options are too complicated / expensive for most people. I see no point in using a Ronchi screen for checking for a sphere as with a stationary source it can be done much the same way with a Foucault tester except diffraction effects are used. This can probably be done with a moving source too. This test can be very precise and to me that is important. John -
  23. I have always used straight ordinary exterior grade ply even for furniture but edged with hardwood mouldings. It can be sanded, stained and varnished just like any other timber. The quality of the external faces does vary so it's best to look around. The biggest problem with this type of ply is cutting it. Jig saws seem to be popular around here but a small circular saw is a much better option used correctly. Going too big makes these saws more difficult to handle. Personally I wouldn't buy a battery operated one either. The one I use most often has a 160 odd mm dia blade. I've had it a long long time Makita DIY range so it wont cut as deeply as most saws of this size. The blades that they come with are useless for ply. Freud do some excellent thin kerf blades. For ply a 40 tooth should be fine at this size. They are rather cleverly made blades and produce a fairly high quality finish. On hardwood for instance edges just need a light sanding. When using the saw make sure that the blade is set square or at the angle that needs to be cut and set it so that it only goes a few mms past the depth of the material being cut. 3 or 4 mm is fine. Sheets of ply need supporting when they are cut. Easiest way to do that is to clamp lengths of lath etc some way in from each end. Even though these finish up being part cut by the saw they can be used over and over again. Best thing to use to cut straight is a straight edge. Siverline do some that are intended for plastering but work well. I found a 6ft one adeqaute for a long time but eventually bought an 8ft one as well which is often a bit cumbersome to use. Trend also do something that comes with clamps. Using these is just a case of measuring the position of the blade from the edge of the saws sole plate and off setting the straight edge to accommodate it. It still pays to mark out where the cut needs to go to save making mistakes. Other straight edges can be used - the edges of ply and mdf sheets are often very straight. Say a 150mm with of that could be cut off any old way. Aluminium extrusions or even a straight piece of timber. Routers can be run along straight edges as well. Even jig saws but these often splinter ply what ever type it is. They are also handy for ripping length of timber on other types of saw. The straight edge is clamped to the work and run along the edge of the table of the saw. It's far more precise and and easier to handle than the fence on some machines. I have often managed to cut up full sized sheets of several things on top of a single DoItAll workmate. 2 wouldn't be a bad idea really but the fact that the piece being cut off remains in place makes the use of one feasible. Some thought is some times needed to avoid the saw fouling on the various clamps holding the laths and straight edge in place. I did own a well made table saw for a while but found the straight edge and circular saw adequate and more versatile even though the table saw had a panel cutting attachment so I sold it as it was hardly ever used. I still own a radial arm saw but to be honest it wont really do anything that a straight edge and circular saw can do. Next step up from exterior grade ply in my book is marine ply. John -
  24. I looked through a mak cass that hadn't been coated at some light bulbs that were on and around 200 yards away. The filaments were sharp and clear. I believe it is possible to star test on a bright star but there will be double reflections off the 2ndry mirror if it's not coated. You will need t o make sure you are looking at the right one. The other problem is that a piece of plate might just distort the image. Figuring via ronchi isn't a test I would trust by the way. There are some notes on it and a link to software that shows the images seen for any mirror. www.nicholoptical.co.uk/pdf/The%20Ronchi%20Test.pdf You should be able to pick up the usual shadows with a knife edge easily with anything under an F8 mirror and even past that with a bit of care. F6 is easy. One thing that can prevent you from seeing them is too large a light source. Easily fixed with a fine pin hole through some kitchen foil over what ever you are using. The best way to make the pin hole is on a hard smooth surface taking care to not let the pin skid. A bit of lightly ground scrap glass is ideal. With that the pin can be spun and will definitely turn out round. Rather than take measurements work out where the edge and 70% zone should be, pick up the centre, use a stick with nails or what ever to mark the shadow positions. Move the knife to positions that should put the shadows in the correct place and see how they look. The edge can be difficult so you may have problems getting to the position where the shadows are exactly on it but the 70% zone, 1/2 the total movement from the centre position should be very close and easy to see. If things come out correctly all should be ok. If the ronchi lines are clear and sharp zones should show up on that. If they aren't the knife will show them. I'm reminded of some one who despairs about people who spend a lot of time making a mirror and no time on testing gear. Mind you he would have people make a Bath interferometer. Fine if you can but they can turn out more complicates then they look to be. John -
  25. There is always here Nigel. I think most will supply as much as you want http://www.ebay.co.uk/sch/Lapidary-Materials-/3237/i.html Rock people can spend a fair amount of money and time on their hobby. There are also abrasive grit sellers but last time I found one the quantities were very large. The mixed grit shot blast type abrasives should be fine at the coarse end. John -
  • Create New...

Important Information

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