Home and Links
 Your PC and Security
 Server NAS
 DVD making
 Raspberry Pi
 PIC projects
 Other projects
 Next >>

Introduction to basic star tracking

Start track Intro

Basic motor drive

When observing at low magnifications a motor drive is largely a convenience - it is quite easy to keep the object in the field of view manually. However at high magnifications it's almost impossible to avoid 'overshooting' and loosing the object, especially when using an Alt-Az or non-aligned EQ mount. Further, whilst planetary imaging is (just about) possible without a motor drive, for sure you need a driven mount if you intend to do any deep sky imaging.

Are decent motorised mounts expensive ?

They don't have to be. Many perfectly good telescope mounts come without motors, especially at the mid to lower end of the market (where price counts the most). An unmotorised version of a mid-high end mount is often the manufacturers 'loss leader'. This is because they know that the purchaser will soon discover that motors are needed for serious use and thus expect them to come back and pay hundreds of $$'s for their add-on 'customised' motors and optional 'goto' electronics.

Which mounts benefit the most ?

It's mounts at the lower end of the price range where motorised star tracking is most vital. The cheaper the mount, the less sturdy it is, and the more prone it is to wobble at the slightest touch. At high magnifications the slightest disturbance - such as any manual 'knob turning' in an effort to track the stars - can 'kick' the object out of the field of view. Attempting to manually 'track' stars at any decent magnification with a low end mount is all but impossible.

What about 'off the shelf' motor kits ?

Whilst many 'generic' or 'after market' retro-fittable motor kits are available, these can still be expensive and often consist of little more than a simple geared servo motor with some simple 'open loop' manual speed controller (which you could build for yourself at half the cost).

The other thing wrong with such kits is that they are usually designed to fit the widest possible range of mounts = so will not be 'mechanically optimised' (i.e. specifically geared) for your mount. This often makes them (very) difficult to fit and (even more) harder to adjust to work properly. The combination of difficult adjustment plus the often very poor (to unacceptable) 'drift' can mean such kits will never deliver reasonable tracking. Needless to say, none of these problems will be discovered until after you have parted with your hard-earned cash.

How do different types of mount effect motorisation ?

The two basic types are the 'EQ' (Equatorial) and 'Alt-Az' (Altitude/Azimuth) mount. To track the stars, the EQ type needs one motor (as the head only needs to be moved in one axis, so long as the polar alignment is good), the Alt-Az two (as the head must be moved in both axis).

Driving a 2 axis motorised head really requires computer control (and is best left to the experts).

It is possible to 'convert' an Alt-Az head into an EQ head by using a 'wedge'. This is rather expensive and you would typically find it cheaper to purchase the manufacturers own (dual) motor controls. Note, howewver that dual axis (Alt Az) suffers from (unwanted) 'field rotation'

Setting up an EQ mount to track the stars

Before you can expect your mount to track the stars, the whole mount + telescope must set-up correctly and meet 3 vital criteria (which must be achieved in the order shown) :-

a1. The 'polar axis' of the mount head has to be set to the correct ANGLE for your Latitude. For example, in London, England you would set it to 52 degrees .. but in Aberdeen, Scotland you would set it to 57 degrees. This only has to be done once (unless you take your telescope on holiday a lot :-) ). If you have an Alt-Az head, you set the wedge to the correct angle instead.

a2. The base of the mount head (or the base of the Wedge) must be totally flat to the ground. If the base is not flat, not only will the latitude angle be 'off' but your 'tracking' will go slightly 'up hill' or 'down hill'. So every time you move your telescope, your must adjust the tripod legs until the base of the head, as measured with a spirit level, is as flat as possible.

a3. The polar axis of the head (i.e. the bit that's now inclined at an angle of 52 degrees or whatever) must be 'pointed' at the North Celestial Pole (NCP). This is 'a tiny bit off' Polaris (the Pole Star). For most purposes, pointing at Polaris is 'good enough' :-)

For long-exposure astro-photography you will have to do better than simple 'point at Polaris' alignment - this is addressed later.

Having achieved all 3 of the above, you can load your telescope onto the head (if not already done) and start to think about selecting an object to view ... and turning on the motor drive.

What is required from the drive ?

The drive has to 'oppose' the rotation of the Earth. i.e. it has to turn the mount head 'backwards', so that the telescope is 'stationary' when pointed at stars (rather than rotating with the earth). The speed of rotation required is 1 complete revolution (360 degrees) per day (actually, it's 360 degrees per 'Sidereal day', however 'once per day' is near enough :-) )

There are many ways to achieve this, however 2 fundamental requirements must be met :-

a1. The gearing must be such that during normal tracking the motor will be running at it's 'normal' speed - constantly running 'too fast' or 'too slow' will drain your batteries faster (and may even burn out the motor).

a2. The actual movement of the telescope / head must be measured and fed into a 'closed loop' control system. This needs to compare the actual movement against some accurate 'reference' (e.g. a clock). This allows any mechanical errors in the drive (e.g. imperfect gearing, motor speed variation induced by changes in drive resistance etc) to be 'fed back' into the motor drive speed control (so it can 'catch up' or 'slow down' when the actual movement is found to be too small (undershoot) or too large (overshoot)).


Autoguiding overcomes poor alignment by adjusting the head (in both axis) via a 'closed loop' control system using a camera and software to 'track' a 'guide star'. The camera is often fitted to a short tube refractor co-mounted with the main telescope (sometimes the camera is fitted to the 'spotter scope'). Modern 'planetary' (movie style) camera's can 'double up' as autoguiding cameras

It is possible to use an 'off axis guide adaptor' at the main scope eyepiece to 'divert' part of the view to the camera, however such adapters typically cost rather more than purchasing a decent short-tube refractor ! So a co-mount is usually the way to go

Do I need autoguiding ?

Basic Autoguiding allows you to keep objects in the center of the field of view (for visual observing) without having to perform an accurate (or even any) Polar Alignment

More fancy Autoguiding is designed to support long-exposure astrophotography - and this is where we have to ask 'why bother ?'.

The fact is, you are always 'better off' taking 60 images of 10 seconds each, rather than 1 image of 10 minutes = quite apart from the fact that you can 'throw away' any of the 60 effected by vibration / atmospheric distortion etc., the magnitude of the noise is reduced by the square root of the number of stacked images (so 60 stacked reduces the noise 7.7 times). Further, todays stacking software is quite capable of 'realigning' images where the stars have moved a bit from shot to shot (so Alt-Az field rotation effects can also be ignored).

This means that, instead of autoguide 'tracking' accuracy of 1 pixel in 10 minutes (or more), we can 'get away with' 1 pixel in 10 seconds !

If there is sufficient light (imaging Moon, Mercury to Saturn) you can 'make a movie' (multiple frames a second) rather than 'take a picture' (multiple seconds a frame). This means you only need an accuracy of '1 pixel in 1/50th second' !

This is within the reach of 'manual guidance' (video frames that have 'smeared' due to vibration etc will be discarded by the stacking software) - and (at lower magnifications) you might even 'get away' with no tracking at all !

What guide camera ?

Commercial 'solutions' (such as the Starlight Xpress Lodestar (about £300) can cost you more than your imaging camera, however they do have the advantage of built in 'mount control'

The fact that control can be built into a small camera housing suggests that you can 'build your own' using a PIC or similar !

If you use Open Source software (such as 'PHD') on a computer, all you need is a basic 'web cam' to feed images to it. Unfortunately, the average web-cam is just not very light sensitive ! Whilst you can modify many web-cams to reduce the frame rate (and thus increase the sensitivity) a better source of suitable camera's is the CCTV industry

You need a CCTV camera designed for use in low light, with a sensitivity of 0.01 LUX or better. I found a basic '700 TVL .001 LUX' mini-camera with 1/3" sensor on eBay (from China) for less than £10 (don't waste money on a useless 'zoom lens', audio support or 'black light' (iR) LED's, all of which you will have to remove anyway).

CCTV camera's usually output analogue 'composite video' (PAL or NTSC). Whilst this means you need a 'digital video converter' (USB 'dongle') for use with computers, the raw data stream can be used more or less directly by a PIC controller (see later).

The astronomically expensive Lodestar (£300 or so) delivers a miserable 752 x 580 pixels. Since all we want to do is 'track' the star near the center of the field of view, things like 'spectral response / colour cast', 'pin cushion distortion' etc. are all totally irrelevant = indeed a basic 700 TVL (TV lines) camera (generating PAL 720x576) should be quite capable of producing good results

Low end cameras suffer from 'hot spots' (which can be 'worked around' on-the-fly by semi-clever software i.e. if the 'star' fails to move when the drive motors are 'off', plainly it's not a star :-) ) and 'drop outs '(dark pixels), all of which may have to be 'mapped out' (see 'dark frames' and 'light frames' in my Astrophotography pages)

Second-hand digital camera ?

There is no point in generating 50 frames a second from your autoguide camera if your autoguiding software is going to take a second (or even 2) to process each frame ! This begs the question, 'can we use an old DSLR / digital camera instead' ?

Unlike most CCTV cameras, DSLR camera's can be set to higher 'sensitivity' (ISO) and longer exposure times

Even old digital cameras, at 3-5Mpixels, about 10x the resolution of the 'average' auto-guiding camera (the Lodestar is 0.44 Mpixels). Unfortunately, getting the photo out of the camera is the problem - few old DSLR's had a USB connection, although you might find one with an AV 'preview' output - plus, course, you need a way to take a continuous 'stream' of photos.

Old cameras often had a 'cable release' socket that can be used with a simple push-button switch (or PIC i/o pin driven)

Using your main imaging camera

Mdern 'planetary' imaging (video) camera's can 'double up' as the autoguide camera, especially as 'absolte accuracy' is not very important (the stacking software will re-align frames)

One problem with planetary imaging is that extreme magnifications are necessary to 'bring out' the features on the planets surface - and the higher the mag. the worse the effect of any vibration. Note that the problem with Jupiter is it's rotation (around once every 10 hours), so the 'red spot' is going to 'smear' very quickly

Next page :- Basic motor control