EL Wire Mechanical Donkey (Parade of Lights Float)

NOTE: This is an ongoing project that will not be completed until it is ready for use in early December 2016. Thus, not all sections of this "instructable" have been written yet.

09/01/2016: Our church decided to create a float for the "Parade of Lights" in 2016, in order to inform local residents about our community services and church school. The chosen theme of the float is "Animals of the Nativity". (The theme of the parade is "The Miracle of Christmas".)

The float base is an 8-foot-by-5-foot utility trailer. To the side rails of the trailer, we plan to attach images of various animal faces, made out of electro-luminescent (EL) wire. The centerpiece of the float will be a mechanical donkey, also lit with EL wire. Around the lower perimeter, we plan to hang standard house Christmas lights. On the rear of the trailer, we will attach an informational sign.

We will also have a lighted "star" attached to the pickup truck that pulls the float, along with signs on the truck.

Attached is an image of a "flyer" that we used to promote the float plan, in order to invite people to help work on it. Note that graphics used in the flyer came from the web, and may be copyrighted by their respective owners. No infringement intended.

Also attached is a drawing based on the idea to use a 20-foot long trailer. This idea was discarded, and we decided to go with the shorter trailer.

Before mentioning the float idea to anyone outside of my own family, I decided to create a cardboard mock-up of the mechanical donkey, just to see whether it even appeared to be feasible to create a more permanent version of it. I used an image of a donkey that I found on the web, and enlarged it on paper, then used the paper image as a template for drawing and cutting out the cardboard pieces.

The cardboard pieces were attached together using 4-40 screws, nylon lock nuts, and nylon washers.

The mock-up allowed me to move the legs and head, in order to see how the motions could work.

I also viewed several web videos in order to get a better handle on how a donkey walks. I did not intend to mimic that behavior in all of its aspects (it is quite complex), but wanted to mimic some of the more basic characteristics of the walk (i.e., to move the legs back and forth, and to move the head up and down).

Here are the videos:

09/05/2016: Early on in the process of constructing the mechanical donkey, I thought about constructing it primarily out of plywood. Before cutting out any parts, much thought was put into various ways of mechanizing it. Attached is one of the early drawings (the yellow paper). It shows wheels located behind the donkey, with rotating wheels that move push rods, which in turn move the legs.

A second drawing (made with MS Paint) shows a central shaft driven by a motor, connected to front and rear shafts that would turn the lever arms attached to the push rods. Note that this drawing includes no mechanism for moving the head of the donkey.

A third drawing adds another shaft used to move the head of the donkey.

A fourth drawing shows the top view of a mechanism based on gears rather than on pulleys, using a single full-width shaft to move all of the legs. Implied are much shorter shafts on the gears that move the push rods. This version would require fairly large gears.

My thought with these drawings was that I could attach the push rods fairly close to the hip pivot points, thus needing only about 4 to 5 inches of horizontal motion at each hip push rod connection point. Once I assembled a complete leg, and tested the amount of pull required to move it (based on its weight), I realized that the connection point needed to be much lower on the leg (for leverage reasons), meaning that more distance of motion was required.

As a result, I chose not to assume that any of these drawings is final, and I plan to arrange the mechanics once the cogged pulleys and belts arrive.

One of the critical parts of the thinking is in how to attached the lever arms, since they will likely be homemade. The fifth drawing shows one possibility, using threaded rod as a shaft.

09/05/2016: Using the parts of the cardboard mock-up as templates, I drew them onto plywood. Originally, we had planned to double the dimensions of the donkey, meaning that it would cover four times the surface area, but after researching the costs of building and mechanizing it, we tossed out that idea.

I had planned to used ordinary construction plywood, for cost reasons, but the wood available at the local store was too warped, and had too many knot holes in it. I opted to upgrade to a better grade of plywood, because it was smoother and cleaner.

Although I am still in the process of cutting out the parts, the attached photos provide a good overview of that process. I have been using a rotary cutter for this purpose.

After drilling holes in the wooden parts (the ones cut out thus far), I used some machine screws to attached the parts of the legs and hooves together, in order to determine how well they could move, and to check the amount of force needed to move them.

Although the cardboard mock-up included four joints (rotation points) from the rear hip to the rear hoof, we decided to remove one of the joints (to simplify motion). Based on that decision, I screwed together the hip part with the upper leg part.

Also,note that the lower leg is in two similar sections, sandwiched together. That was simply a fallout of copying the mock-up, which needed extra cardboard in order to be strong enough to manipulate. The same thought was given to the upper leg and the hip area, at least for a while.

See attached photos.

09/05/2016: Once I realized that the rear legs were too heavy, I tried connecting the "push/pull" point lower on the leg. This decision halved the amount of force (to about one pound), but of course, implied that the push rod must move around twice as far, which would change how the mechanism would need to be constructed.

In order to reduce some of the weight, I decided to cut holes in the leg parts. So far, I have only done this on the rear legs. I will wait until later to decide whether to do so on the front legs.

As a quick side task, I decided to "test drive" one of the animal faces. Each face must be lit with EL wire, bent into the shape of the face. We intend to use black foam board as a backing for each face. As an experiment, I printed out a sheep face, and taped some EL wire to the printout, then lit it (via AA batteries), to see how it might look.

After that experiment, the next one involved creating the first EL wire animal on the proper black background. The animal face is made of white foam board, with the EL wire taped to it, and all of that taped onto black foam board. When lit, the face outline shows quite well, and even the small white foam eye inserts show up, even though they have no lighting of their own (the eye outlines are close enough to bleed some light onto them).

The main body is made of 1/2 inch plywood. I decided to use 2x4 braces to screw it together, and steel sleeves (from the hardware store) as bearings for pulley shafts (which I lubricated after installing the shafts). In order to make sure that the holes aligned properly, I screwed the 2 sides together before drilling the holes. While installing the wooden braces, I inserted metal shafts into several holes, to help keep them aligned. I constantly tested the ability to rotate those shafts, while tightening the braces.

The 4 larger pulleys are single-flanged, simply because the company I bought them from did not have any dual-flanged ones in stock. I had to add the 2nd flange to each one. I found some wooden discs at the hobby store, and used JB Weld to attach those disks. The 4 smaller pulleys are dual-flanged from the factory.

I tested the mechanics a bit by rotating one of the flat steel lever arms (which shows in the photo of the motor). The attached video shows this test. The lever arms are nothing more than deck braces purchased from the hardware store.

The pulleys have 10 mm shaft holes. The local hardware store has no 10 mm rods, but it does have 3/8 inch rods, which are close to the proper size. I chose to use 3/8 inch threaded rods for pulley shafts, and to secure the pulleys with nuts and lock washers. I may add some JB Weld later, to make sure nothing comes loose (not sure yet).

Attached are some photos of the motor control box that will drive the motor. Its primary components are an Arduino UNO board, and a Pololu motor controller board. There is also a terminal block that helps to organize the connections. On top of the box are 3 switches. One toggle switch turns on power from a 120V-to-12V convertor. The other toggle switch turns on power from a 12V truck cigarette lighter plug. The pushbutton switch tells the Arduino to run the motor until the button is released.

Another important component is an Arduino Lilypad accelerometer. This is used to help determine when the truck is moving, so that the motor can be turned on when the truck is moving, and turned off when the truck stops moving.

The fully wired control box worked fine with the in-house convertor, and worked with the cigaretter lighter plug, as long as the truck engine was not running. After starting the truck, the control box stopped working. After research, I reckoned that the alternator was providing over 14V of power, and according to the Arduino UNO web page, it is not recommended to supply more than 12V, even though the maximum rating is 20V. Thus, I will likely need to insert a 12V regulator of some kind, between the power supplies and the Arduino.

Once the pulleys and belts were installed, it was time to install the motor, and give the internal mechanism a test. I found a couple of steel angle brackets at the hardware store that, when screwed to each other, were just about the correct height to use as a simple motor mount (there are 3 mount holes on the motor's gearbox). I used steel straps to secure the body portion of the motor to the inside of the donkey. For the most part, I used nuts and lock washers to secure the pulleys in place on 3/8 inch threaded rod. Later, I may use some JB Weld to help keep things from coming loose.

See the video for a test of the pulleys, driven by the motor.

The donkey must be raised up above the other animal pictures on the trailer, so I had to make a simple stand for it. I used some leftover and scrap 2x4 boards for that purpose. If I keep the stand, I will need to add some angled braces to it (triangles are stronger than parallel or perpendicular board joints).

One thing that I do not like about this stand is that it partially blocks the view of the legs on the opposite side of the donkey. The only way to correct that is to suspend the donkey from the top (i.e., with ropes or cables), which would imply placing high posts at the 4 corners of the trailer. Not sure I want to do that, because the cost of aluminum or steel channel is significant. I may just need to live with the wooden stand.

The attached photos and videos show the legs and head. At this point, the rear legs and head are attached to their driving pulleys (the legs via rotating lever arms and the head directly to the pulley). I used tie rod ends to enable the rotation and action of the connected parts. Note that 3/8 inch diameter rod ends are way overkill for this job (i.e., too strong and too heavy), but this is a learning process. The lever arms came from the hardware store, of course.

I made an important mistake (big duh) relative to the connecting rods. I purchased 3/8-16 threaded rod for connecting the tie rod ends together; however, the ends themselves have 3/8-24 threads. Thus, the rods won't screw into the ends. To work around that, I bought a few 3/8-24 bolts (1 inch long). I will JB Weld the middle rods (3/8-16) to the bolt heads. For the tests in the videos, I just used duct tape.

The motor controller box will likely be mounted to the side of the donkey stand, but during these tests, I just set it on top of the donkey, on one of the wooden braces between the body sides.

Need to connect the front legs sometime soon.

After connecting the front leg control arms to the front legs via their push rods, all 5 body parts were moving. However, I had to remove all of the control arms in order to glue their sections together. All tests were done just using duct tape as the fastener. Even after gluing, I put duct tape back on, in case any glued joint failed (broke apart) later, which did happen. The attached video shows all body parts moving.

I used a scrap piece of plywood as a base for holding the push rods while gluing them, to make sure that their lengths did not change from when they were first tested..

With the push rods disconnected, it was a good opportunity to being lighting the donkey, i.e., to start taping on some EL Wire. During this process, I used 4 different kinds of tape, including clear cellophane tape, white duct tape, clear duct tape, and black electrical tape.

To cover up the holes that I cut out of the rear legs, head, and neck, I used thin craft foam (the flexible kind).

Before reconnecting the push rods, we tried to finish up various details, including the face, mane, tail, and ears. The eyes and ears are made out of white foam board. The face is flexible foam material. The tail is white parachute cord with EL wire woven into it. EL wire runs all around the face, including up through the mane.

Still need to draw in some nostrils and possibly some lips, but whether they are seen at all would depend on the ambient light during the parade.

Attached are photos of the fully lighted donkey. It uses 5 separate EL Wire controller boxes, each one using 2 AA batteries.

Here are some pictures and a short video from parade day, with the completed float. They were taken before it got really dark. Things looked even better after dark. Enjoy!