I had access to a large collection of old bicycles in various states of disrepair and was interested in finding a creative way to reuse them. Originally I was trying to figure out the best way to try to make a rickshaw type vehicle, then I started thinking of adding an electric motor. While surfing on the internet I came across some simple microprocessor based robotics boards. While looking at these, the notion of a heliostat somehow was born. Some more meditation on the pile of bicycles revealed the thought that they could easily be modified to provide what is known in astronomy circles as an alt-azimuth mount. By tilting the frame at the proper angle the front forks could also work as an equatorial mount, but I thought it would be easier, and more user friendly, to design an alt-azimuth system that would not only have simpler alignment, but could perhaps also be designed in a fault-tolerant way.
The bicycle frame gets turned upside down, and mounted on a pipe where the seat post would normally fit into the frame. On the H1 prototype, I cut the back half of the frame off, thinking that it would simply get in the way and wasn't needed. After some experience now, I'm going to try leaving the back wheel portion of the frame on next time so that it can be used to hold counter-weights for the mirror assembly, which will help reduce stress on the mount.
The use of off-the-shelf components and use of the bicycle frame greatly reduces the difficulty of fabrication. Besides the use of a welder in building the frame, only basic hand tools are needed. This should make the project well within the ability of any experienced handyperson.
Bicycle SelectionBoys BMX style bicycles seem to have the best frame geometry. If you look closely you'll notice that the seat post and the front axis of rotation for the forks are not usually parallel. This can be worked around (by compensating for the pitch angle), but it's nice to get a frame where these are as parallel as possible. Look for front forks that use round tubing. You won't care about any of the hardware on the bike except for the bearings in the handlebar. I've found these very easy to find for free. If you have options, one other thing to look for is the geometry of the handle bar. A nice straight bar is ideal.
PreparationStrip everything off of the old bike, leaving just the frame and front forks and handlebar. On the H1 prototype I used an angle grinder to cut the frame for the rear wheel off at the crank and seat mount locations. In retrospect, I wish I still had that to support a counterweight that would greatly reduce stress on the mounting pole.
Front forks are often slightly curved. Find a nice spot 4 to 6" from the end of the front forks where the tubing is relatively straight for another few inches. I like to put some tape around the tubing to mark where I want to cut. You can cut the forks by hand with a simple hacksaw, or use an angle grinder with a cut-off blade if you like to make some sparks. Be sure and where protective goggles/faceshield, use hearing protetion, and always wear tough gloves when using an angle grinder.
Once you've cut the forks you'll need to sand about 1/2" of the paint off from the end of each of the now four ends of tubing from the forks. You can use sandpaper or power it out with the angle grinder again.
Fork ExtensionMeasure the inside diameter of the tubing from the front forks. I simply went to the hardware store and purchased a piece of 1/2" black steel pipe. It was a little small, but close enough to do the job, and very cheap. Insert the pipe into the fork tubing to determine how far it will go on the upper and lower pieces. Add about 18" to that distance and cut two pieces of the tubing to size. You'll also want to prep the metal where the welds will be by sanding off any paint or grunge.
A third piece of the tubing needs to be cut to act as the altitude actuator support post. This gets mounted inbetween the base of the two front forks. The other end is supported by two braces welded to the front fork extensions approximately 9" from the end. You'll need to locate the actual desired actuator mount shaft location before the dimensions of this piece will become apparent. 3/4" angle iron works nicely for the braces (~14"). A bench grinder, or angle grinder can be used to grind the end of the angle iron so that it makes a nice joint with the round tubing. I used the threaded end of the pipe here to put a galvanized coupler on the end, which also helped give the width that was needed between the actuator mounting plates.
Two short pieces (4.5") of 1.5" steel bar (1/8" thick) are used to support the actuator mounting shaft. You'll need to drill a 1/2" hole near the end of each piece for the shaft. I used a short (3.5") piece of 1/2" all-thread for the shaft. This allowed me to mount the pieces of bar on the shaft by using a nut on each side and adjusting the nuts to achieve the desired spacing and alignment. The support braces are not parallel so the bars need to be twisted slightly. A vise and large crescent wrench work nicely to achieve this.
All of this directly effects the altituded control arm geometry. Refer to the appropriate documents to help you calculate the ideal dimensions for your actuator. The trigonometry used to calculate the final angle during operation depends on the dimensions of three sides of a triangle formed by the two actuator mounting points and the mirror assembly axis of rotation. What this means is that you don't have to be too precise during fabrication, but you still need to know what you are aiming for. The distance on the mirror assembly of the actuator mounting point from the axis of rotation is one side of the triangle. The longer the distance is, the greater your resolution will be, but if it's too long, your actuator won't go far enough for the desired range of motion. The minimum length of the actuator arm, as when the mirror assembly is in a parked horizontal position, defines the second side of the triangle - the actuator motor mount point must be low enough that this position (i.e. 90 deg elevation) can be achieved. The remaining side of the triangle (refered to as C) is the distance between the mirror assembly axis of rotation and the actuator motor mounting shaft.
Getting Ready to WeldSetup the bike frame, upside down, on your work/test stand. Insert the fork extension pieces, and put the fork ends on top. Make sure they are turned 180 deg from normal so that the width is maximized. After calculating the ideal value of C for your setup, you can create a jig to help weld everything together in the right position. Use a piece of 1/4" or 3/8" all-thread to stabilize the fork ends at the desired separation. Take a piece of wire or string attached to the center of that shaft and attach the other end of the wire/string to the actuator mounting shaft inbetween the two support brackets in so that the distance matches your calculated value.
I took a piece of cardboard and drew a right triangle to assist in locating the final mounting position of the actuator support brackets. The hypoteneuse is our value for C, the base is held horizontal, and the vertical side is yet another number that has to be calculated using trig. I punched the all-thread on the forks through the appropriate point on the cardboard, and clipped the lower part to the frame to hold the base line horizontal. Holding the support bracket assembly at the desired position measure the final length of the support post.
Azimuth Motor Mount