A patent,
back in motion.

A mechanism I grew up around.

My grandfather, Leo James “Jim” Hubbard, was a 1949 graduate of the University of Rhode Island whose engineering career included work at Hughes Aircraft, General Electric, and Xerox. He went on to found Precision Handling Devices, turning his patented designs and material innovations into paper-feed equipment sold around the world. URI documents that career.

I grew up around these tractors. My father’s factory did the injection molding for the plastic parts, and I was there all the time as a kid, including working after school. The mechanism fed perforated printer paper using a continuous belt whose pins fit into the holes along the paper’s edge.

This project began with a search for my grandfather’s original patent. Finding the drawings on Google Patents gave me a reference for recreating a mechanism I already knew well, as a model I could turn around, look inside, and watch in motion.

The starting point was his 1974 patent, Feed mechanism. We used Astra to read the description and work through the geometry, Meshy.ai to make an initial visual study, and Blender to build and refine the moving assembly.

Start with the source.

Figure 2 shows a single tractor with its guide plate open. The sectional drawings explain what that view hides: the belt supports, the sprocket recesses, and the shape of each molded drive element.

UNITED STATES PATENT3,825,162

Original drawing, cropped and rotated for reading. Part numbers and linework are preserved. The full patent contains 12 figures.

One belt. Two working surfaces.

Above the belt, a pin enters a hole in the paper. Below it, a half-cylinder fits into a rounded recess in the sprocket. Those two portions form one drive element. As the sprocket turns, it carries the belt along a straight feeding path and around each end of the frame.

The slotted guide plate holds the paper close to the belt while giving the pins room to pass. In use, two tractors support the opposite edges of a sheet. This reconstruction concentrates on one tractor, so the engagement is easier to see.

The material mattered.

My recollection is that my grandfather encountered the film through work on electrical components for rockets, before it was widely available, and recognized its potential for a paper-feed belt. Seeing a useful connection between those two applications is part of what makes this mechanism interesting to me.

The patent explicitly names Kapton, DuPont’s polyimide film, among the materials suitable for the belt. The design called for a lightweight strip that could bend repeatedly while maintaining its dimensions and pin spacing. Heat resistance also allowed the drive elements to be molded directly onto the belt. Those properties helped the belt keep the pins aligned with the paper as it moved.

Three tools, distinct roles.

I had never worked with Blender or Meshy before this project. With Astra helping me through the setup, I could get an initial model on screen and then refine it through specific feedback. I could describe what I wanted in ordinary language, then respond to what appeared on screen.

That first result gave us something concrete to work with. Refining the shapes, materials, and motion took further iterations. I could contribute by comparing the model with the drawing and drawing on my experience with the parts in the factory.

The Meshy study stayed alongside the reconstruction for comparison. The moving parts were built separately in Blender, where their spacing and motion could be checked.

Meshy · first interpretation

Blender · refined reconstruction

The details that mattered.

The first version had a silver finish. The parts I knew from the factory were black injection-molded plastic. The belt was amber, but it needed to let light through. And the top guide had rounded ends that did not match the patent’s more angular outline.

Those became concrete revisions: black plastic parts, a translucent amber belt, and a guide plate with straight ends, bent sections, and an integrated stepped tab. Comparing the model with the drawing made each correction easier to explain.

The most satisfying part was seeing the pins come around the sprocket and line up with the paper. A mechanism I had grown up around was now moving on screen, built from my grandfather’s patent drawings.