STAR TRACKER Mk2

Home

1. Gearbox and Drive

The original test table drive arrangement is shown opposite. It consists of a large 1.8° stepper motor directly coupled to the worm shaft. The worm drives the large bronze main worm gear which has 180 teeth. This gives 36 arc seconds per step of the motor, which unfortunately is not a fine enough drive resolution for star tracking. However, it did provide an excellent starting point for a new tracker. The test table was probably made in late 1980's, the rigidity and quality of the machined parts is clear to see and would put many of the present day telescope mounts to shame.

 

The test table components were stripped, cleaned, repainted and new bearings were fitted. The original stepper motor was replaced by a smaller 7.5° stepper motor driving through a 250:1 gearbox. The new arrangement gives a resolution of 0.6 arc seconds per step of the motor. When I was cleaning the worm I noticed that there was some fretting damage to the teeth on one side of the worm. This was probably caused by the test table spending most of its life in one set position. I replaced the worm with the one from the other axis and lapped the 'new' worm to the main gear using 3 micron diamond lapping paste followed by 1 micron. (This type of process is not recommened for those without experience in lapping).

 

Most high quality telescope mounts have Periodic Error Correction (PEC). This is a method for correcting for the fixed errors in worm gearboxes used for telescope drives. (As a worm rotates, the spiral worm has a certain amount of eccentricity with respect to it's rotational axis. This eccentricity causes the main worm gear to speed up and slow down for every revolution the worm. The form of this error is typically sinusoidal and repeats for each rev of the worm. Low quality telescope mounts typically have periodic errors of ±20 arc seconds or more, high quality mounts have periodic errors of ±3 arc seconds or less.) I wanted to incorporate PEC into the new drive, eventhough this added complexity to the drive train and electronics.

 

If the rotational position of the worm is known, the stepper motor can be slowed down or speeded up to exactly compensate for the worm periodic error. To achieve this, I coupled a shaft encoder to the worm shaft. This was achieved by attaching a gear to the front face of the flexible coupling which is fitted between the gearbox and worm shaft. The encoder is an 8-bit absolute gray scale encoder, giving 128 discrete pulses per rev.

 

2. Electronics

I wanted to keep the size of the tracker as small as possible. I was aware that the electronics would probably cover a couple of Eurocard boards and so a case to house them would be larger than I would have liked. The metal cover over the worm gear had some clearance to the rotating gears, so I decided to set myself the challenge of fitting all the electronics and displays under this cover. This proved to be tougher than I had originally thought, but I did eventually manage to plan a layout.

 

 

The drive electronics are based on a Philips SAA1027 stepper motor drive chip. This drive frequency is obtained from a crystal oscillator which is subdivided via a CMOS 4059, to provide the desired motor step rate. This subdivision can be easily modified using four Binary Coded Decimal (BCD) switches, accessible via a perspex window on the side of the tracker cover. The subdivision for sidereal tracking is a value of 3989. The four decades of adjustment give very fine resolution for setting the drive frequency.

 

The main tracker casing had a conveniently located bevel on which I fitted two seven-segment led displays. The one on the left shows the frequency subdivsion value for the motor drive. If the PEC function is in use, this shows the subdivision value varying throughout each rev of the worm. The right hand display shows the output from the shaft encoder on the worm. This displays 1 to 128 as the worm shaft rotates. The LED intensity is adjustable by a switched potentiometer.

I attached a 0-360° scale to the main worm gear. This is read through a magnifer made from an objective lens from a very old, but ideal, microscope eyepiece. This scale also has adjustable LED illumination.

The encoder (Bourns ACE, Farnell part no. 935-8234) has a Gray scale output. In a Gray scale, only one bit of the 8-bit output word changes for each step of the encoder. The output is therfore non-sequential in normal binary counting terms. The conversion of Gray scale to binary or BCD is not straightforward, so I decided to use a look up table stored in an EEPROM for the conversion. I couldn't justify buying an EEPROM programmer for this task and borrowing one would never be as verstile as owning one, so I decided to build one instead. I used the same EEPROM (Atmel AT2828C64B) that I used in the PEC circuit, so this gives me complete flexibility to reprogram the encoder and PEC values stored in memory. The EEPROM programmer is the right hand card in the image to the right. The other card is the lower drive card which carries the EEPROMs. The drive electronics use 16-bit data for the frequency subdivision, so there are two 8-bit EEPROMs for the encoder circuit and two for the PEC circuit.

 

Lower Drive Card

Upper Drive Card

Drive and Gears

3. Polar Alignment Scope

For a polar alignment scope I used the same type of collimator as used on the Mk1 tracker. I added a red LED to illuminate the reticule. The LED intensity is controlled by an adjacent potentiometer. I added a fold mirror on the scope output. The scope gives a correct, non-inverted image. It also allows the scope to lie along the existing plate on the tracker, helping to keep the scope out of harms way. The yellow flip up dust cover on the end is from an old shampoo bottle!

 

4. Tripod

This is based on an old wooden surveyors tripod. I've added a turntable with fine adjustment and a lock for use during polar alignment. A small red led lamp is fitted underneath for general illumination during setting up.

I modified a bubble level from Maplin Electronics to allow it to be backlit with a red led, this helps to carry out the initial levelling of the tripod, prior to polar alignment.

 

Back To General Tracker Design

Top