PDA

View Full Version : Astro tracker revisited - build pix


snaarman
10th December 2015, 06:57 PM
So, here are the build pictures of my home made astro tracker (http://e-group.uk.net/forum/showthread.php?t=40447). I re-used some of the previous hardware so there is a similar look. There is a rectangular base which I sanded flat.

On that there is a turntable that I routed into a circle. There are four ball bearing wheels round the edge (mistake, you are better off with three) and they are glued to the turntable on 8mm wood dowels.

"A" is the quick release top tripod mount which can tilt vertically (see below)
"B" is the turntable with its new toothed belt drive

http://e-group.uk.net/gallery/data/622/P1010630.jpg (http://e-group.uk.net/gallery/showphoto.php/photo/87776)

Closer in we can see that the motor "C" sits in a 4mm well so that the belt works properly with the turntable.

http://e-group.uk.net/gallery/data/622/P1010631.jpg (http://e-group.uk.net/gallery/showphoto.php/photo/87771)

Note that I mounted the motor in this position "D" so that the tension of the belt is supported by the motor body (the drive cog is offset).

http://e-group.uk.net/gallery/data/622/P1010635.jpg (http://e-group.uk.net/gallery/showphoto.php/photo/87775)

Underneath we see that there is a ball bearing acting as the central rotation point "E". This is in a deep hole and the central shaft is in fact an M8 hex bolt screwed into the turntable bottom. This allows me to pull the turntable down on the base and eliminate any backlash.

The whole assembly sits on another Manfrotto quick release mount "F". The tripod is used to get the pole star alignment...

http://e-group.uk.net/gallery/data/622/P1010633.jpg (http://e-group.uk.net/gallery/showphoto.php/photo/87773)

...

More to follow

Pete

snaarman
10th December 2015, 07:08 PM
On the top you can see a small wood block that supports the camera quick release mount. There is a brass hinge "G" for tilt adjustment and out of sight underneath an adjustment screw to set the tilt. This is the least satisfactory part of the whole thing.

Note that the toothed belt is kept in tension with a spring "H" One end of the belt is screwed to the turntable and the motor pulls on that when rotating in the clockwise direction. The spring operates on the take-up side of the belt.

http://e-group.uk.net/gallery/data/622/P1010632.jpg (http://e-group.uk.net/gallery/showphoto.php/photo/87772)

I did try to use a section of belt inside out as a sort of cog but this doesn't work. The belt is so accurate that the teeth don't mesh if you do that. Hence the screw and spring arrangement.

Finally, because the drive is a geared stepper motor you need a microcontroller. Here is my 11 8051 in a nasty plastic box that does the honours.

http://e-group.uk.net/gallery/data/622/P1010634.jpg (http://e-group.uk.net/gallery/showphoto.php/photo/87774)

Typically, I only had one push button switch available so I had to write a sequence system to operate it. If you press and hold the switch two LEDs indicate a sequence of four options.

Stop
Synchronous track forwards
Fast track backwards
Fast track forwards

When I explained this control philosophy to a fellow engineer he described my method as "disgusting"

:-)

Pete

crimbo
10th December 2015, 07:10 PM
So... how well does it track? Any periodic drift? What is the max carrying weight. How does it handle CG changes with the camera in different positions?...
....are you going to be building kits or selling plans:D...
...await next instalments hopefully with an image of comet Catalina

snaarman
10th December 2015, 07:18 PM
So... how well does it track? Any periodic drift? What is the max carrying weight. How does it handle CG changes with the camera in different positions?...
....are you going to be building kits or selling plans:D...
...await next instalments hopefully with an image of comet Catalina

He he :-)

Well the equatorial rate is adjustable in the software, so I should be able to get it spot on with some trial and error. Over 60 seconds with my 75mm I don't see any real error. I did a two hour test track in the lab using a protractor and convinced myself it was well within a degree.

On a shorter time scale, the toothed cog might not be fully coaxial with the shaft and the motor gearing might be less than perfect but I don't think I am going to pursue that sort of precision.

Backlash is always the problem with this sort of thing. That is why I ditched the worm drive idea. Strictly I should include some sort of back tension on the turntable so the drive belt is kept tight on the pulling side. I will see how the pictures turn out.

I just got a cable release for the camera, so I don't have to touch the rig to set off the shutter any more. That should be a big improvement.

Jim Ford
10th December 2015, 07:22 PM
I know you have a lot of experience with the 8051, but isn't it a bit primitive? (Crikey, is that an RS232 interface I see?)

Personally, I'd go for one of the Arduinos - the basic ones are cheaper than your 8051, well supported and a delight to use.

More recently I bought a micropython board, which runs python programs natively. It shows a lot of promise, but I've not got around to playing with it.

https://micropython.org/

Jim

snaarman
10th December 2015, 07:27 PM
I know you have a lot of experience with the 8051, but isn't it a bit primitive?

Personally, I'd go for one of the Arduinos - the basic ones are cheaper than your 8051, well supported and a delight to use.

More recently I bought a micropython board, which runs python programs natively. It shows a lot of promise, but I've not got around to playing with it.

https://micropython.org/

Jim

Yes, I have decades under my belt with the 8051 in assembler and (for this project - a novelty) in the dreaded "C" This is a fast Silicon Labs 100 pin 8051, and the ones we use at work run at 50Mhz so they are very different to the clockwork items of the past. I quite like the easy bit manipulation commands, I have all the programming tools so it's my preferred starting point really.

Having said that, we have shifted our products to 32 bit ARM chips and I can see why programmers (C programmers I mean) prefer that sort of thing to this 8 bit rubbish..

PS. Yes. RS232. I ditched that and put a USB interface on there for debug purposes ;_
I am quite tempted by the new Raspberry Pi - or even the older ones - but I do like my assembler. Mmmm.

Pete

Graham_of_Rainham
10th December 2015, 11:04 PM
I've got a nice clockwork motor in the garage, I may give this a go... ;)

Jim Ford
11th December 2015, 09:22 AM
I am quite tempted by the new Raspberry Pi - or even the older ones - but I do like my assembler. Mmmm.

Maybe Forth is the way to go - after all, it was designed for controlling a telescope. I dabbled with it some years ago and quite liked it.

Jim

OM USer
11th December 2015, 01:18 PM
It does not look like the brass hinge has 180 degree movement so it must be quite a headache to line the camera up on a particular star, eg the pole star for doing circular star trails. Maybe a ball and socket head screwed direct into the turntable would give the best flexibility in a compact solution.

As the stars only rotate in one direction I would also prefer to see the tension spring positioned so that it is in the gap between the turntable and the drive pully on the "slack" side. Having both ends of the belt touching the turntable where there is no slip will not give a constant tension as the thing is moved about and relocated.

Having said all this, a marvellous piece of work.

snaarman
11th December 2015, 01:35 PM
Maybe Forth is the way to go - after all, it was designed for controlling a telescope. I dabbled with it some years ago and quite liked it.

Jim

Mmm. Strong the Forth is with this one, yes.

snaarman
11th December 2015, 01:44 PM
It does not look like the brass hinge has 180 degree movement so it must be quite a headache to line the camera up on a particular star, eg the pole star for doing circular star trails. Maybe a ball and socket head screwed direct into the turntable would give the best flexibility in a compact solution.

As the stars only rotate in one direction I would also prefer to see the tension spring positioned so that it is in the gap between the turntable and the drive pully on the "slack" side. Having both ends of the belt touching the turntable where there is no slip will not give a constant tension as the thing is moved about and relocated.

Having said all this, a marvellous piece of work.

Yes the elevation part needs some further thought, I wanted something quite compact to preserve a sensible C of G. I think it can point almost down to the horizon, and also almost vertical from E via S to W, but it can't really point to the North or at the pole star. A solution would be to dispense with the wood and hinge and install a tripod head as you say, but it would start to be ungainly then.

For circumpolar star trails I would just fix the camera on a conventional tripod and wait as the earth rotates :-)

What it does need is a method to line the base plate up normal to the pole star. I tried using a 90 degree star prism held on the base, but ended up with a plastic set square to get the alignment. Maybe a pointy laser would be smarter.

P

birdboy
11th December 2015, 02:54 PM
Great barn door mount Pete. Just getting into astro tracking myself. Thought about doing something similar then I spotted the Skywatcher star adventurer at 285 you get such a lot for your money.

Polar alignment is critical for astrophotography Polaris alignment is not accurate enough you need the North Celestial Point. The Star Adventurer has a built in polar scope to do just such an alignment.

OM USer
11th December 2015, 10:16 PM
What it does need is a method to line the base plate up normal to the pole star.

Bit of old copper tube and some thick fusewire to give some cross hairs?