Barn Door Tracker

On the home page I mention a barn door tracker I had built but had not been too successful with it. This is the latest version based on Gary Seronik's design. I used a curved piece of threaded rod to drive the top platform with gears and a stepper motor taken from an old Epson printer. I found a simple stepper motor control circuit by Rob Paisley and put this into a box to control the stepper motor. I will write a more detailed description in time but for now this is the image of Casseopeia taken on the first attempt with this device. I was pleased to see no trailing in the stars.

Olympus E500 3 x 1 minutes subs 1 dark stacked in DSS.

A plane flew over in the first frame.

The top and bottom boards are from 18mm plywood because that's what I had to hand. Earlier versions have used 12mm softwood boards approx. 100mm wide. I maintained the 100mm width as I felt it had worked well enough. The board size is 100 x 300mm. I used cheap and cheerful hinges on earlier versions but found there was too much play in them so invested in a high quality hinge to do the job.

The drive screw is from M5 threaded rod which was bent to suit the radius needed to drive the board through 15 degrees. The movement is simple you have to lift the top board through 15 degrees in an hour to match the sidereal rate of the stars. I won't discuss the maths here as Gary Seronik does a good job on his website.

The drive screw and control box

The drive mechanism has been recycled from an old Epson printer for the stepper motor and the gears. This is where you have to do some calculation of your own to ensure the gears you have will turn at the right rate to lift the top board. In my case I needed the stepper motor to be turning at 6 RPM to turn the main gear.

In order to fit the main gear to the drive I epoxied an M5 threaded nut into the middle.

Close up of the drive mechanism

The control circuit from Rob Paisley's site has been modified slightly as I only needed to drive the motor in one direction. I swapped the 1M pot for a 150K pot as this gave me the range I needed to finely control the rate of rotation. However I was never truly satisfied with the performance and the control knob was easy to move and upset the timing. So I settled on a 200K multi-turn preset which was easier to set and adjust. As it sat on the board it was out of harms way. So the hole in the box was used for a power on LED with an on off switch next to it. The addition of the 200k preset has enabled me to set the tracker up well enough for a five minute exposure to be possible so I shall have to test it on the next clear night.

Another modification was to remove the control circuit for the reverse function. It is easier and faster to lift the top board disengage the gears and spin the main gear in reverse to reset. I retained the reset switch as it proved helpful as a means of turning off the output temporarily.

I'm a dabbler in electronics and although I believe the circuit is correct you would be well advised to check out Rob Paisley's site for a full explanation of how this circuit works and for an accurate layout.

The above modifications proved necessary due to interference in the circuit which led to the stepper motor randomly going into reverse for a few steps. Removing the unwanted components overcame this problem and this image of Orion was taken as a result.

EOS 550D 60s f5.6 ISO 800 single shot.

Processed in Canon Digital Photo Professional to reduce the skyglow which was prominent in the lower half of the picture

M42 EOS 550D on Barn door mount 6 x 60s subs no darks f5.6 ISO 800 55-250 EFS Stacked in Deep Sky Stacker. Processed in Light Room

M42 EOS 550D on Barn door mount 10 x 60s subs no darks f5.6 400mm ISO 800 Stacked in DSS processed in Light Room and further processed in Photoshop Elements (Levels adjustment). Finally Imagenomic Noiseware applied to reduce Sensor noise.

In its present form the tracker has served to get photos of the night sky but use of longer lenses means errors are magnified. One source of errors has been the vibrations of the stepper as it has plodded around and around with its 7.5 degree step angle.

I decided to experiment with an Arduino micro-controller and an easydriver board to introduce micro-stepping to the tracker which produces a smoother result. I'd used an Arduino for another project a year or so ago and so set about designing the programme.

It was a simple concept one button to activate the platform and start raising the top section which can be reset to its start position by pressing a second button. In the event the platform reaches the top of the travel when my back is turned, I added a photo interrupter salvaged from an Epson printer. This stops the motor and I can then press the reverse button to reset the platform.

This shows the prototype setup with the Arduino, breadboard, and Easy driver, the photo interrupter is on the right. The original board is still in place in the control box.

Having created a prototype I then proceeded to create a second prototype on a prototyping shield on the Arduino. This enabled me to see how it all worked in a more compact form and refine the programming and wiring of the circuit.

Then came the plunge into committing the whole design to a permanent Strip-board layout and installing it in a project box. First though an order to Bitsbox for the required parts. I decided to use two keyboard switches through the top of the box which would require some careful drilling to achieve the accuracy I required to avoid the switches binding in the sides of the hole. In the event because two of the Arduino header strips are placed further apart than the 2,5mm pitch of the strip-board the strip-board was slightly skewed and so some fine adjustment was needed to avoid the one switch binding.

The switches were soldered to the Strip-board along with headers for connection to the Arduino. The layout of the strip-board changed during the course of the build and connections were relocated to made installation in the box easier.

The layout of the components within the project box was planned with the aid of Autosketch which is a drawing package from the same stable as the Autocad software used by Architects and Engineers. It is easy to use but no longer available having been discontinued by Autodesk in 2016 shortly after version 10 was released. I'm using version 5 which is around 15 or more years old!!

The final version fitted to the tracker on its first night of use. Tweaking the programming further resulted in an accuracy of +0.04 secs after 28 minutes runtime. The result; stars which are round enough after 5 minutes exposure.