Building a Simple Barn Door Tracker
 |
| The tracker, with a Nikon D3100 mounted to it |
If you've ever tried taking pictures of the stars with a camera, you'll quickly notice two issues. One, it's really dark and it's almost impossible to see anything without a long exposure. And two, once you do a long exposure, the stars start smearing into little lines. So why do you get these trails? Well, the Earth is actually spinning on its axis, and the trails you see is actually the Earth's rotation. The stars haven't moved, but the Earth, and by extension the camera, has, and gives you that motion blur.
 |
| Those stars should be round! |
So how do you fix this? You can simply spin the camera at the same speed as the Earth's rotation, cancelling out the spin. Now you can buy a tracker for well over $300, or you can build your own with some simple parts for $50, or less if you have some of the hardware components, which are very commonplace. There are several designs out there, and I chose to make the curved rod version, as it guarantees an equal arc movement for the duration of travel, unlike a straight rod.
For this build, I followed some pretty good build tutorials from people way smarter than me. I will just detail my personal build process here, but definitely check them out for good instructions as well as how to use it.
Parts and Tools
The crux of this build are the hinge, a curved threaded rod, and a motor/gear system to spin the rod. The hinge is a cheap 2.5" hinge from Home Depot, and the wooden base is some MDF board I had lying around. You can use plywood or really anything flat for it. The wood I had was .7" thick, which seemed to be a decent thickness, and was just enough for the hinge screws not to poke out the other side.
The curved threaded rod dictates your gear ratio and hardware, so it is important to stay consistent. For me, I had some 1/4"-20 steel threaded rod lying around and some assorted 1/4"-20 nuts, so that's what I used. You don't really want anything thicker than that, since you will need to bend the curve in by hand. The TPI is important too, as that will dictate how quickly you need to spin the nut to raise it. In addition, I had two lock nuts to secure the rod to the top base, a washer for that, and a regular nut for the rotation mechanism.
The motor and gearing works together, and dictates the radius of your curved rod. I bought a small 6V 1RPM motor off of Amazon used for cooking spits that I planned to power with 4 AA batteries, and 3D printed my own gears. Since it is impossible to know for certain if a motor will actually spin at exactly 1RPM (accuracy really matters here), I also bought a potentiometer that let me regulate the voltage to speed up or slow down the motor, as well as double as a power switch. If you have random gears lying around, you can use them as well with some modifications. Since the Earth spins one degree every four minutes, the formula is as follows:
Rod Radius = RPM / (0.004375 x Rod TPI) * Gear Ratio
Since I had a 1RPM motor and 20TPI rod, I would need a rod radius of 11.427" if I maintain a 1:1 gear ratio. That seemed like an acceptable size to me, since it wouldn't be too tight that I would have difficulty bending the rod and warp the threads too much.
The length of the rod also dictates how long the tracker will be able to run without intervention. In my case, 1 foot of rod was equal to 4 hours, although conservatively it's probably closer to 3 due to some of the radius being unusable to attach it onto the board. That's still plenty for my purposes, and it can easily be reset in less than a minute by disengaging the gear and spinning it back, although you will need to realign the camera after that. The formula for calculating maximum theoretical runtime based off of length is below:
Hours Runtime = (12 x Rod Length) / (𝜋 x Rod Radius from above)
 |
| Gear mid-print |
The tools used were pretty standard, and while I used power tools, they can also be substituted with hand tools. For cutting the MDF, I used a miter saw for very quick and accurate cuts. For the inset cutout described later, I used a drill with a spade bit, as it was uncontrollable on a drill press. For all the holes, I used a drill press with standard bits.
Build Log
 |
| The template I made in CAD, 1:1 scale on printer paper |
The second step was to build the hinge base. I first cut the wood into a 12" section, and then attached them using the hinge. Then, measuring from the hinge axis, I measured out the radius of the rod, which was roughly 11.427". I then drilled thru-holes that fit the free fit diameter of the rod. Depending on how accurately the rod is curved, you may want to make it a little wider so it doesn't hit the edge as the rod travels through. The rod is then secured to the top plate using a lock nut and washer, with a wing nut to top it off because that's what I had.
 |
| Bottom of the top board |
I then drilled a hole for the motor shaft to go through. To line up the gears, I first attached the rod gear, and using the motor gear, I marked out four opposing corners. I then drew lines to cross them to find the midpoint. Since my gear shaft was 12mm, I used a 1/2" drill bit, which left me a little bit of wiggle room for fine adjustments. Once the motor was in place, I drilled the mounting holes for the motor mount. I mounted mine at an angle, so it leaves a bit of space for the tripod mount and switch.
 |
| Motor gear slot. I made a mistake with my first gear, so it's not perfectly circular |
For the bottom tripod mount, I used a 1/4"-20 tee nut that fits perfectly with the standard tripod mounting thread 4" from the hinge. However, the tee nut was too short for the tripod bolt, so I had to use a spade bit to dig into the wood to shorten the distance. Obviously that didn't make a very clean hole, and ideally a Forstner bit should be used, but the spade bit worked in a pinch, and is much cheaper.
 |
| Inset with tee nut at the bottom |
The top ball head mount was mounted at 6" from the hinge, so as to better center the actual camera's weight above the tripod. This was a simple 1/4"-20 carriage bolt that goes through the wood, and is held in place by the ball head. Normally, tripod heads are mounted with 3/8" bolts, but the cheap one I got had a threaded insert included to drop it down to 1/4"-20. These inserts can be cheaply obtained, or alternatively a 3/8" carriage bolt is just as cheap and stronger.
 |
| Ball head with the bolt head visible |
Lastly, the electronics needed to be wired together. This was a simple solder job using some 16AWG silicone wire I had lying around, which was a bit overkill for the low voltage and current of the motor and battery source. One thing to be aware of is the direction the motor should be spinning. You want the nut to travel down the rod, and the motor needs to spin the opposite direction. Also, when soldering, make sure no solder gets onto the motor can, as that can lead to a short. The potentiometer had screw down clamps, so no solder was needed on that end. I didn't have a battery tray available, so I salvaged an old remote controller and hot wired the terminals to the potentiometer.
 |
| Don't forget the heat shrink! |
Lastly, attach the motor onto the board with the motor mount, slide the gear onto the shaft, and you're done with the construction! The next step is to calibrate the motor such that it makes exactly one revolution per minute. I did that with a simple timer and a mark on the gear. I was lucky, and my motor made exactly 1 rotation per minute at full power, so the potentiometer was actually superfluous, but it is still useful as a power switch.
 |
| The bottom of the tracker, with a Manfrotto quick release clone mount for the base tripod |
Future Optimizations
There's a few things I want to do to update it and make it better. First, I don't like how the electronics are exposed and dangling, and I would much rather build an enclosure for it. I will probably use an old plastic tupperware container and mount it onto the base. For weather resistance, I would feel better covering the potentiometer and electronics in some conformal coating, especially when it gets cold and dewy at night. Similarly, I would seal the wood with some polyurethane or something. However, since the camera I'm using isn't even weather sealed, it feels like a moot point. Lastly, the ball head mount I bought was a tad too small, and not quick release which made setting up the camera difficult. I am thinking about replacing it with a quick release head in the future. Also, the alignment process is somewhat difficult, so I'm thinking about getting a laser or sight attached to it to streamline the process.
Conclusion
Overall this was a pretty quick build, as the barn door tracker is quite simple mechanically, and the hardest part was designing the gears. I learned quite a bit about how MDF behaves when worked on, as well as how aggressive spade bits can be. I will definitely be grabbing some Forstner bits in the future to use with my drill press.
As for the actual performance, the tracker actually did very well! I put it through its paces, and it works as advertised. The crucial part is the polar alignment, as if it's not aligned exactly it's still going to trail a little bit. But even in the below image, I was able to take images over 20x longer than the maximum recommended untracked time based on the rule of 500 without any trailing issues, which is great. I would definitely recommend building one if you're interested in taking pictures of deep space objects or really anything in space with a camera, and don't have the budget for a prebuilt tracker. It won't hold a telescope, but at that point you'd be looking at more precise systems anyways.
 |
| 45 second exposure at 200mm. No more star trails! |
Comments
Post a Comment