jmi
Senior Member
Posts: 60
|
OM2n
Jul 23, 2006 13:05:04 GMT -5
Post by jmi on Jul 23, 2006 13:05:04 GMT -5
My oldest (and first) SLR. I love this thing apart from one snag - it seems to have prodigious amounts of mirror slap. Don't really know why but it's more than a bit irritating, especially since Olympus removed the mirror lockup on the 2. Took this because I was planning on selling it(!) at the time. Obviously I didn't Been used in the interim so it's slightly more grubby now.
|
|
jmi
Senior Member
Posts: 60
|
Clouds
Apr 20, 2007 15:11:20 GMT -5
Post by jmi on Apr 20, 2007 15:11:20 GMT -5
Of course clouds here are a Bad Thing...
|
|
jmi
Senior Member
Posts: 60
|
Post by jmi on Sept 2, 2006 5:31:58 GMT -5
That's a very nice shot Bob! A fairly common sight in the Cam. About 10 years ago the council had a free bikes scheme. They were painted green and the idea was that they were just left unlocked around town. Anybody could pick one up and bike somewhere, and just leave it. Of course, what actually happened... (a) the bike thieves from London came and loaded large numbers of green bikes into their vans, and (b) vandals threw the rest in the river. Naturally the scheme was abandoned Quite a lot of the bikes are still there, tend to find them when punting... What were they thinking indeed?
|
|
jmi
Senior Member
Posts: 60
|
Post by jmi on Sept 7, 2006 13:25:07 GMT -5
The CCD is so small that the curvature is not normally noticeable, especially since the collimation on the optics is not too good right now. For a larger chip we would have to install a lens to flatten it out.
|
|
jmi
Senior Member
Posts: 60
|
Post by jmi on Sept 5, 2006 15:13:39 GMT -5
This particular three mirror design is optimised to give a very wide field of view.
For a simple paraboloidal shape of mirror, the images of a source at infinity are perfect on-axis, but off-axis suffer from coma and astigmatism. This typically limits the usable field with images of acceptable quality to 0.5 degree or less. The problem can be fixed with corrector lens systems, placed close to the focus. Typically a three-element system is used, which needs aspherics - ie. expensive, and since lenses were used this introduces chromatic aberration. The three element lens will give a 1 degree field or thereabouts.
The next level up is the Ritchey-Chretien design, which uses two hyperboloids. This is corrected for coma, but astigmatism limits the usable field to about 0.75 degree or so. The primary mirror can also be used in isolation, which gives aberrated images all over the field, but with corrector lenses (which are easier than the ones for the paraboloid) a 2 degree field is obtainable (eg. the Anglo Australian Telescope prime focus corrector).
The three-mirror design corrects both aberrations to an acceptable level (0.33 arcsec or so - atmospheric seeing means this limit is rarely reached) over a 5 degree field. No corrector lenses meaning no chromatic aberration. It is essentially an all mirrors version of the Schmidt camera, which is a spherical mirror and full-aperture aspheric corrector plate system.
There are a few snags with the three-mirror telescope: (a) the mirrors are very difficult to figure, and align correctly due to the strongly aspheric shapes and the speed of the system, (b) the focus is inside the telescope tube, a disadvantage shared with the Schmidt but not a conventional prime or cassegrain as in the first two cases, however here it's even more inaccessible being in the middle of the primary mirror, (c) the focal plane is curved, also a problem with the Schmidt.
|
|
jmi
Senior Member
Posts: 60
|
Post by jmi on Sept 4, 2006 16:45:30 GMT -5
Just put together a quick one (looking from the same direction): I'll have to do a mount diagram later (or google on fork equatorial mountings might turn up something I suppose). Rough diameters of the mirrors are: M1 20 inches M2 12 inches M3 14 inches (on a 16 inch blank) ie. a lot of glass here (well, actually it's Zerodur, which has much lower thermal expansion).
|
|
jmi
Senior Member
Posts: 60
|
Post by jmi on Sept 4, 2006 15:33:42 GMT -5
Okay, as promised... these are really quite bad, sorry - did them very (too) quickly during my coffee break, on the last few frames of a roll I nearly finished over the weekend. View from the back of the building, the primary mirror is in the middle of the rectangular section (which is attached to the mounting), you can see the secondary at the top, and the third mirror is at the bottom. There seems to be a lot of lens flare here (yes, I was using a lens hood). The CCD, mounted at the focus, we're looking up from the third mirror. The focus is roughly in the middle of the primary mirror, looking down. I must have wobbled here, rather slow shutter speed... This is the "polar wedge" part of the mounting, with the drive motor visible (this is new - what I was doing), along with a couple of boards where the various wiring is collected together. Wonder what the milk bottle is doing down there. I'll try to get some more (and better) pictures later.
|
|
jmi
Senior Member
Posts: 60
|
Post by jmi on Sept 2, 2006 12:47:32 GMT -5
Anybody just joining us, see GeneW's "New Computer" thread. Hopefully we can move the discussion out to here... I will endeavour to get some pictures ASAP, but I have no digital here so will have to be on film - hopefully by next weekend with the wind behind me, or sooner in B+W. The roof has to be off unless I had a really wide lens, and it's wind and intermittent rain here. Pity I didn't take the opportunity last week but I was testing the drive system then. I would say there is a lot more in the way of images taken with telescopes on my web page, but unfortunately the disk that most of it lives on, and in particular all the images, is not working right now (died in the middle of backups too, as usual). This is the trouble with being a part-time system admin, and being cursed this week
|
|
jmi
Senior Member
Posts: 60
|
Post by jmi on Sept 2, 2006 15:20:48 GMT -5
I may have misremembered. A 1600 would be an upgrade in a way, since the sampling would still be acceptable at 9 microns (2.3 arcsec) but the field of view would be larger. I suppose buying one of those DSLRs to steal the CCD would be sacrilege... What I'd really like is one of these: www.ing.iac.es/PR/archive/wht/simon2chip.jpg2x 2048x4096 thinned EEV CCDs, would cost about 100k UKP!!! = no way
|
|
jmi
Senior Member
Posts: 60
|
Post by jmi on Sept 2, 2006 13:01:34 GMT -5
I think (correct me if I'm wrong) this one is a KAF1400, that's the same but with 6.8 micron rather than 9 micron pixels.
The CCD system dates back to 1992, as in the entire CCD system dates back to 1992 - it's run by a 486!
I never knew they were used in DSLRs but I guess that would be why Kodak made the chips. This one sits in a sealed head of about 4x4x2 inches filled with dry air, and is thermally coupled to a Peltier effect device which takes the temperature down to about -30C. Since this is all mounted in the middle of the tube the heat is removed by a heat exchanger arrangement running a water/glycol mixture to a pump unit situated near floor level.
The water cooling and cold CCD are the bane of my life when dealing with this thing - water cooling = air locks and leaks, cold = ice (on the window at the front of the head) on the usual damp Cambridge nights.
|
|
jmi
Senior Member
Posts: 60
|
Post by jmi on Sept 2, 2006 12:25:24 GMT -5
Apologies for the confusion - Andromeda is 8 CCD images, that lot cover about 1.5 degrees out of the total 5, but I did it by moving the telescope around. I have some older film images of the same object taken years ago by Roderick but I'd have to borrow the prints. It hasn't had any film through for years, sadly. CCDs are great for getting around the horrendous light pollution we now have to contend with. The site up at Madingley Rise used to be outside Cambridge in the middle of a bunch of fields, but the town has grown out past us. If getting the CCD out wasn't such a rigmerole I would have some film images of my own by now. I would really like to try colour at some point, but that would be much easier and cheaper if I could somehow adapt it for roll film - the cost and inconvenience of 4x5 has been putting me off a bit (3 inch circles have to be tray developed AFAIK), and getting it in focus at f/1.6 will require a few hours and sheets of film too... Will take a camera down next time we get any decent weather here (could be a while). Here's another image: Messier 81 12x 300s exposure = 1 hour total (stacked CCD images) This has serious reduction problems as you can see... another problem I need to solve (long term project, this). The reason why it looks like a light leak is because it is
|
|
jmi
Senior Member
Posts: 60
|
Post by jmi on Sept 2, 2006 9:40:57 GMT -5
Here is one example. Well actually it's 8 stitched together, and I've not done a particularly good job so you can see the joins Messier 31. 8x 300s exposures with the Three Mirror Telescope at the Institute of Astronomy.
|
|
jmi
Senior Member
Posts: 60
|
Post by jmi on Sept 2, 2006 9:25:22 GMT -5
Hmm, come to think of it my router runs OpenBSD. Great setup - I forget it's even there, virtually no fiddling required. I really must write up some more of the telescope project. The telescope itself is owned by the Institute of Astronomy, but it's funny you mention image capture -- it's really a camera in a way, since it was designed for exposing film. The aperture is 20 inches as I said, f/1.6 (ie. focal length 32 inches) which as I'm sure you know is exceptionally fast for a telescope of any kind. For the non-astronomers, imagine what an 800mm f/1.6 mirror lens might look like It's a weird three mirror design and is one of only two in the world, the other is a 4 inch prototype that lives in the same building. Both were designed and built by Roderick Willstrop and the workshop at the Institute. He's still there but retired. It records a field of view of 5 degrees diameter (!) onto 3 inch circular film (cut out of normal 4x5 using a punch). We used to run it with hypersensitised Kodak tech pan film. However currently we have a cooled ccd installed, it's only a tiny 1536x1024 Kodak chip, which throws away the field of view advantage - the ccd only covers 0.5 degrees on the short side. I have been restoring it to operation, and rebuilt the drive system so the mount is now controlled through the sparc. Only one axis is automated so far, but I intend to do the other one next. I seem to be making an attempt to hijack this thread, sorry! Once I obtain some pictures (not today, the weather is foul and the building is too small to do it with the roof shut) I'll post something elsewhere. Images taken with it are easy since I have loads.
|
|
jmi
Senior Member
Posts: 60
|
Post by jmi on Sept 2, 2006 7:00:53 GMT -5
Wow, you really have been using it a long time! I wasn't even born then I'm still using BSD derivatives too - FreeBSD on a few PCs and NetBSD is running the 20 inch telescope that has been the subject of my weekend project for the last 6 months (running on an old Sparcstation 4 that used to be my workstation). Since it's been keeping me away from photography I really must actually get round to taking some pictures of it. Sound cards, argh! Linux hardware support has improved a lot in recent years but it's still not perfect... You have a pretty neat setup with shared bookmarks, etc. I'm typing this from firefox on Windows (scanning), but had to remember the URL for the board. d'oh.
|
|
jmi
Senior Member
Posts: 60
|
Post by jmi on Sept 2, 2006 5:51:53 GMT -5
Hey Gene, Glad to find more Linux users around here, and it's interesting to hear you are writing about it too, good luck with the column (and the computer)! I don't really use anything else, I only have Windows now for scanning, and that's just until I can get that sorted on Linux (not had time to have a proper hack at it yet). ("Professional") Astronomy is pretty much 100% run on UNIX derivatives so that's even more true during the week. GIMP raises an irritation for me - it's fine for most things but the lack of support for editing in 16bits/channel is quite frustrating when trying to do scanning. Mind you I spend all week writing image processing software, so may I should just sit down and write something to do the required adjustments to levels/curves before downsampling to 8 Come to think of it I'm also an FM2n user, it's a very nice finder, although I've clearly been using the 'blad too much recently because it seems rather small right now!
|
|