Making Mack

Mack is a collection of servos, fiberglass, epoxy, fake fur, latex, and aluminum roughly arranged in the form of a mounted moose head. He has six axes of motion: eyebrows, mouth, eyes left & right, eyes up & down, neck left & right, eyelids, and mouth. There's also a speaker in his mouth for all the witty things he has to say. Each moving part is controlled by an RC servo, which in turn is driven by an RC transmitter. When selecting RC supplies, I chose the Futaba SkySport 6 for several reasons:
Why did I make Mack? We put an addition on the house a couple of years ago that has a cathedral ceiling. The area in the peak looked empty, and I wondered what we could do to spice it up. I made a mental list of things I've seen done in rooms with high ceilings, and one of the items was of a mounted trophy head. One thing led to another, and now there's a moose on our wall. Mack's not mounted in the open peak area, though. He's too big to fit there because there's a ceiling fan that would bump his nose. Perhaps if I do another, smaller one I can put it under the peak.

Moose construction took place in bits in pieces, so I have no idea how long it actually took me to finish. The job I had when I started the project required a lot of travel. About halfway through I changed jobs, and although I don't travel at all right now, I need to put in extra hours to meet various delivery deadlines. So, there was almost no continuity, and occasionally I would work out a design to some problem only to forget it when I got around to the implementation.


I sculpted Mack in Roma #2, and I molded the sculpture in 5 pieces using Ultracal 30. Once the mold was done it was time to make the underskull. Mack wasn't going to have a foam skin, so I didn't need a core mold. Underskulls are generally made from rubber molds made from the core pieces so the foam fits properly. I didn't feel like bothering with a core and then making a rubber negative of that just for the underskull, so I cheated. (This was the first of many cheats, actually. There are more coming, so if you're at all queasy about that sort of thing, it's time to hit the back button!)
Anyway, I still needed an underskull. After I lined the mold pieces with 1/4" of clay, I put a layer of cellophane over the clay. At that point I had a poor man's negative of the core, and I built up the epoxy resin & fiberglass skull pieces from there.  Once they were done, I reassembled the mold and put in fiberglass strips that joined the skull pieces. When that set up, I pulled the complete underskull from the mold.


Key to making all this work was a strong neck that could support a cantalevered head but could be moved by a hobby servo. I had Hosford's (a local machine shop) make six plates from 1/2" aluminum and then mill recesses for thrust bearings. I spaced the plates about 10" apart and stacked the "vertebrae" in a stair-stepping fashion.

There are two pivot points. One would have been easier, but also would have looked very stiff, kind of like what you see at Major Magic's or Chuck E. Cheese's. The extra joint gives it a nice, broad sway, but also introduced the serpentine swaying effect you get with this sort of design (which is amusing in it's own way, but inappropriate for a moose). To counter that, I attached elastic to the neck ribs that support the fur and give the neck it's shape. The elastic provides just enough resistance to stifle the serpentine motion without overloading the servo.

The servo wasn't your garden variety mini that you find in conventional kits. There'a a company named FMA Direct that makes 1/4 scale servos, and they make one called the S500. This unit is still pretty small - 2.5"×2.25"×1.75" - and at 4.8v provides 250 oz. torque. I could have handled something twice that big (and would have been even mightier), but servos like that are made by specialty outfits and run several hundred dollars. That's fine for ILM, but not the likes of me. The S500 is a bargain at $45.

The eye mechanism is pretty klunky, but I was able to make it in my basement with a consumer-grade drill press. The pivot point between the eyes allows them to roll up and down. The support is mounted to the bridge of the nose, and the servo is mounted on a plate attached inside the nose. Left and right motion comes from a servo mounted to the main eye support. The eyes are cedar balls I pulled from decorative caps which are sold as trim for stair railing.
The eyelids are also crude, but they're durable. The lids were cast from the stone mold of the moose. First, I painted in a few layers of latex. Then I took some fabric that was somewhat fuzzy on one side and embedded it fuzzy side down in the last layer of latex. Last, I painted epoxy on the exposed fabric. The real trick was to mount the lids so that they didn't touch the eyes. Not having an CAD program to test various mechanisms or a machine shop to build them, I didn't bother to do any up-front design. I just made the mounting arms on the lids long enough that I could make it work by trial and error.


These pictures were taken once the underskull was complete and half the mechanics were installed. I had to resculpt the nose because the mouth wasn't open enough in the original sculpture. It would have been workable with a foam nose skin, but the slush cast latex skin resisted too much and the servos couldn't open up the mouth.


Fitting a speaker inside the moose was a challenge. It needed to be small so it would stay hidden and not interfere with the mechanics, but it also needed to sound good. The effect would have been ruined if Mack sounded like he was talking through a 1960's Heathkit AM radio. After a disappointing experience with a midrange speaker purchased from Radio Shack, I ended up buying a set of computer speakers. Not the little ones that are the size of your fist, but the ones that are about 8" tall. That style relies on the chamber that surrounds the rather small speaker inside to produce a full sound. The chamber beefs up the bass, which is what I needed. I only needed one, but they aren't sold individually.

Unmodified, the speaker was still too big to fit in Mack's nose. I hacked apart the primary unit and removed the amplifier board. Then I chopped off the now empty bottom portion of the case and built a new bottom for it. This shrank the unit enough that it could fit inside the moose. I installed the amplifier in a hobby case and ran wires from it to the speaker. Then I placed the amplifier unit in the base of the neck.



For a while I thought I would carve the antlers out of some material like dense styrofoam, and then paint on a latex or epoxy skin. I wanted to avoid sculpting, molding, and casting and just do something simple. As with everything else on the moose, the antlers needed to be lightweight, so I couldn't do something like buy a huge box of Sculpy and bake the antlers into submission. The shape I wanted seemed too complex for an approach like casting a latex skin and then then filling it with expanding foam. There would be lots of nooks, crannies, and points sticking in wierd directions acting as air bubble magnets for the latex. Been there, done that. I also wasn't excited about making silicone molds for hard foam casts. Silicone is expen$ive and is relatively slow to cure, plus you need a mother mold. The whole process for a pair of antlers would drag out for days, especially since I was doing this in the evenings and couldn't guarantee being able to work on it every evening.

But after all that worrying and fussing, I wasn't able to find a carvable material. If I'd had the time to really look into it, I probably could have found something workable. So the antlers are rigid urethane foam cast from silicone molds.


The device that makes Mack work as an automated performer is a little gizmo called the M5, from Effective Engineering. For those who don't want to or can't spend the bucks on a Gilderfluke or equivalent system, this is definitely the way to go. Go to their website (one of the liks below) for detailed info, but I can sum it up here. The M5 allows you to record and play back movements produced by an RC transmitter. The recordings can be made using a cassette player, which makes it easy choreograph the movements to a mono audio track.

Having mentioned the cassette deck, I must also say that if you want to use it for a more professional puppet, use a CD. But, for testing purposes and livingroom displays, a cassette works fine. I was fortunate in that a friend of ours, Paul Hahn, has a recording studio in his basement. We recorded the audio track first, then played it back while puppeteering the moose. The audio went to a speaker and into one channel of the recording deck, and the other channel was hooked up to the M5. Paul is a solid baritone, so we used his voice. We overcranked the the recording deck by about 25 percent, then undercranked the playback by another 25 percent. (These probably aren't the terms used by audio people, but my background is effects photography, so that's how I think of it.) This dropped the pitch to a range that we liked, and masked it so that people would accept it as Mack's own voice rather than Paul's.


Supplies
RC control devices for puppets and other applications.
Among all the other interesting supplies, they carry XR1592, which is the rigid urethane foam I used for the antlers.  I also used their latex for the nose.
Ann Arbor hobby shop
Burman Industries Lots and lots and lots of things. I ordered VRM-65, which is a brushable grade of silicone.
Stadium Hardware A great source in Ann Arbor for small machine screws, aluminum stock, epoxy, and other things needed for moose construction.
Hosford & Company Ann Arbor machine/blacksmith shop. Good guys, good prices.


Back to the main Mack page
Back to the Home Page
Hosted by www.Geocities.ws

1