Been thinking.. electronic collimation?

[Rob]

I've been thinking for a while now about a new way to collimate, and how this could be controlled electronically. I'm sure this could be done. The ultimate would be a self collimating scope by a push of a button...

Kind of thought about motors & gears in the mirror cell etc.. also thought about having a hole in the primary with a laser sat behind it that could fire and hit a sensor placed in the secondary to move the motors to self collimate.

Ideas welcome?.. who knows this could be a next big step in Astronomy equipment??..

Just think.. mount out. Tube on mount... press collimation button.. unpack kit bag. Ready to go!!!

Rob

[Catanonia]

Wow, now that would be nice and complicated...

[Bazzaar]

All the big multi cell scopes like the Keck do this already.

So the technology is already sorted, you just need a few gillion pounds!

Barry

[weareborg]

This is normally done using stepper motors attached to the collimation bolts and controller from a box at the eyepiece. I have seen it fitted to a few scopes, but never actually been able to find any detailed plans on it. Although I can't imagine it being too difficult to do

Andy

[Revs]

I'd be more worried about making it work on the secondary

It would be a difficult but interesting project. But worth all the effort? I dunno. I rarely have to adjust my Dob and it's not exactly a long process when I do.

EDIT: You've got my brain going now. I was just thinking about how to do the secondary and I was thinking three stepper motors round the outer front of the tube... tiny shafts running along the spider arms to a small worm gear setup acting on the secondary col. screws.

I've no idea how to go about the automation of the whole thing, though.

[rowan46]

probably a stupid idea, but would one of those car mirror controllers work? save you fiddling about with a screwdriver

[Rob]

Nice to hear you thoughts all.. its an interesting one isnt it. And I know things move quite quick in the astro world these days. I'm sure it must be in dev state somewhere in the synta factory for 10" scopes and above??..

Who would have thought 5 years 1:10 crayfords would come as standard on most scopes + leaps in goto!

I like the idea of the car wing mirror controllers... humm, yes.

True also about the large scopes having collimation systems.. it would be pain to climb up to the secondary then realise you had left the allen key on the desk!!

[Psychobilly]

Meade have been doing it for a while on their higher end scopes...

Meade Instruments Corporation - LX400-ACF

A snippet from the linked PDF...

Precision Electronic Optical Collimation:

Collimating a Cassegrain telescope has never been easier. LX400-ACF owners will make precision collimation adjustments to the secondary mirror of the telescope by using the arrow keys of AutoStar II handbox, allowing a single person to simultaneously make adjustments and see the results. In addition, Meade precision collimates the optics at the factory and then sets that position as the default setting. So in the case where one may make a mistake in making a collimation adjustment, the default setting can always be used.

[rowan46]

thinking about it don't high end motor cars have user preference settings where you push a switch and all the seats and mirrors move to a preset position. no lasers just a positional encoder and some computer memory

[NickK]

This could be achieved in a few ways.

Very small stepper motors could be used to gear drive worms to alter the alignment of the secondary (tilt). Motors outside of the scope could the be used to planar shifting X,Y of the secondary core.

By using a cable on the seconday (think small USB) it would be possible to align then disconnect with the colimation being held without the need of routing power/data through the secondary vanes and possibly causing issues with the image path.

Analysis by CCD would be possible within a feedback loop. The loop stops when the colimation is within an acceptable devation range. You can then unplug the USB.. or if you use a vane routed cable, then you could colimate between each imaging set by using a processing step to account for temperature or mirror shift.

The large scopes have active optics where they use voice coils to bend the secondary mirror and a splitting in the image path before the cameras. The split then enters a wavefront analyser. This then effectively both colimates and compensates for atmospheric churn (ie the distortion changes over the image and the corrective action is then relayed to the voice coils - not assuming it's a flat constant distortion). Cost for this sort of area is about 10-20K for the budget end with low number of voice coils mirror and wavefront analyser.

[George]

So the technology is already sorted, you just need a few gillion pounds!

Barry

Nah just a good Blue Peter junk box and a good forum like SGL, its amazing what can be done with gaffer tape and some empty toilet roll tubes

[yesyes]

Not a bad idea. I think the biggest challenge is to make it all small enough on the secondary mirror side as to not obstruct the light path.

On the eyepiece side you could put a laser collimator and attach a light sensor to it (where you would normally see the laser dot). Then, depending on where the sensor sees the dot, move the mirrors to compensate and get the dot in the middle (a bit like auto-guiding). It might be a bit of a challenge to decide which mirror to move though as both would cause the dot to move. Also, if the collimation is way off, the dot would not end up on the sensor at all. It might take a while to move all 6 axis in order to try to get the dot on the sensor.

[themos]

I found that collimating an SCT with live video showing the out-of-focus Doughnut was fairly quick (~ 4 minutes).

[Peter Drew]

Hmm, sounds like a lot more that can go wrong to me..

[sbooder]

I suppose you could use some form of airtight bladders at three or more points and have them attached to electric pumps so you can inflate/deflate on the push of a set of buttons on a handset.

[sbooder]

Scrub bladders and replace with air pistons, much firmer!

[NickK]

I think the most important bit is the software.

You'll have a contour map/3D map of the mirror surface and the code will have to both deduce the amount of collimation required - the fun part is coping with the multitude of aberations (dust, mirror differences) as each scope will be different.

The good part is that you can quantify the quality of the alignment very easily by the quality of the circles and their spacing.

A backward propagation neural net could analyse the 2D input but with a large CCD image (detailed enough to make the system accurate) that's going to be a large network!