Belt grinder build, part 1

I wanted to experiment with building a belt grinder. But not having experience with one, I wasn't sure if it was something I had much use for, so I didn't want to build a typical belt grinder design such as Jeremy Schmid's, John Heisz's, Cosmas Bauer's, or this one.

And of course, out of wood, because it's what I know best, and quickest and cheapest.

I used some scrap MDF that used to be a headboard to cut some disks to make the main drive wheel out of.

My 3/8" Forstner bit made a hole that fit a 35 mm bearing just snug, but I needed a press fit. So I ground a slight bit off the outside of the Forstner bit to make it drill holes that fit tightly.

I marked a line on the drill to help guide how deep to drill (the depth stop on this drill press is missing).

The bearing fit quite tight. I used two C-clamps and a board to press it in the hole.

For bearings mounted in a wheel, a press fit is essential or the bearing will work its way loose eventually.

The wheel will consist of three layers. The problem when laminating layers together is that they float on the glue and get out of alignment while clamping together. So I drilled two holes to fit screws (just large enough that I could push the screw through the holes), then aligned the layers and tapped the screws into the second layer with a hammer.

The screws then acted as alignment pins while I glued two layers together.

Aligning the third layer (which has the other bearing in it) is more critical. Any misalignment results in wobble of the wheel.

I clamped the third layer on, then carefully aligned it so the wheel did not wobble, then screwed the screws further in so they got into the middle layer.

Then gluing the third layer on, same way as the second layer, with the screws as alignment pins.

I also made a smaller wheel for the top using the same method. But this was small enough that I could press the ball bearings (roller skate bearings) in with a vise.

I previously ground that 7/8" drill down so it made a hole to tightly fit the 22 mm bearing.

I tried to trim the wheel to perfectly round on the bandsaw, but the blade wasn't stiff enough. The workpiece just pushed the blade to the side as it cut.

So then I rigged it up on the table saw, which is much stiffer.

I also arranged it so I could push the wheel forward and back along the shaft so I could cover the cylinder and make sure its surface was perfectly parallel to the shaft.

Then adding a chamfer to the edges on the belt sander.

I trimmed the top wheel the same way, but added a slight "crown" to it. Basically, it's slightly barrel shaped. This helps the belt track on the wheels.

The big wheel was slightly out of balance, but the bearings did not turn freely enough to balance it just by gravity. Previously I worked around this by letting it roll on smaller bearings inside the main bearing, but with these bearings just 16 mm inside diameter, the only bearings I had that were small enough were ones for router bits. I was going to put two of these bearings on a shaft, but then I had the idea of just mounting the router bits, with bearings in two pieces of wood.

It's probably better to balance the wheel by drilling holes in it, but I didn't want to have a big hole in the surface, so instead I used two wood screws to add weight to the light side.

My original idea had been to have the back column of the sander slanted so the top wheel would be forward enough for the front of the belt to be vertical. But if I use it with different lengths of belt and moved the top wheel along the post, the belt wouldn't stay vertical. So I made the column vertical and made the top wheel holder long enough to bring the front of the top wheel in line with the bottom wheel.

Gluing one side of the top wheel holder together. It's easier to glue in two stages so there are fewer pieces of wood that need aligning at a time.

After gluing up the first half, I realized I wanted one hole to be a slot to allow for adjusting tracking, so I cut that out with a coping saw while it was clamped up.

Then gluing on the second layer, relying on my workbench surface and a square to help keep it aligned as I clamped it.

Then jigging the rest of the sander together just with clamps and testing how the tracking worked, just by inserting a nail in the slot to force the shaft up the slot.

Tracking wasn't working very well. I wasn't sure why, but with everything held together precariously with clamps, I figured I should make it more solid before working on that some more.

I drilled some screw holes in the plate that has the stub axle for the lower wheel, used a short nail tapped into every hole to transfer the hole locations, then used those divots to locate pilot holes, then screwed it together, with just two screws at first.

Figuring out where the top mount goes for a 48" belt and marking it.

Then cutting a shallow slot where I marked it.

Two washers screwed into the top mount, overhanging the edge slightly, will engage the slot in the top.

I cut a slot in the bottom back of the mount and installed a threaded insert in the column.

This allows me to screw a thin piece of wood against the back of it, which rotates the top mount to push the top wheel up for tensioning the belt.

But belt tracking adjustment still didn't work very well. I checked that everything was sufficiently square (it was). I found my bottom roller had a slightly negative crown to it, so I fixed that. That made it slightly better but still didn't fix the problem.

I thought maybe the problem was with the shape of the top roller, so I switched to a roller skate wheel, but my tracking adjustment still wasn't very effective.

So I decided to investigate changing the tilt of the wheel, and the rotation about the vertical axis (yaw angle) independently. Both of those were able to change the tracking. But lifting the right side caused the belt to track to the left, while moving the right side forward also tracked it left. The way I had the slot, I lifted the right side while also moving it back. Those two caused opposite adjustments to the tracking, so they cancelled out. Oops! So I really needed a slot that slanted in the opposite direction.

So I sawed the slot out a bit more for an opposite slant. It was a bit of a hacked-up looking hole!

And with the slope I had, I needed something to push the axle forward for tracking adjustment.

So I threaded a hole into the back of the bracket using a piece of threaded rod with some slots cut into it with an angle grinder to act as a wood thread tap.

And with this change, tracking worked perfectly. I'm testing this by manually spinning the lower wheel with a nail as a crank.

Before I go further with this build I need to figure out how to mount the motor, and that depends on which motor I'm going to use. I need to decide which of these three 1750 RPM motors I should use. I will have more on that next week.

Continue on to part 2

See also: