Wooden gear cutting jig

There's nothing that symbolizes machinery like gears. Seeing that I like to make machines out of wood, it follows then that I should want some of these machines to have gears on them. Wooden gears, of course.

I have cut numerous wooden gears on my table saw for various machines, such as my Marble machine 2. One of the biggest problems with this is precisely measuring the angle between the teeth, and rotating the gear by exactly that much between cuts on the table saw. With the gear jigged up on a shaft to hold it securely, it isn't possible to attach a protractor to measure accurately. And even when using a protractor, there is always the problem that the protractor has to be moved (seeing that it doesn't span 360 degrees), and the possibility of getting some slightly odd teeth intervals as a result.

gearbox with counter To help rotate the gear by a precise amount for the purpose of cutting teeth, I decided to build some kind of mechanism for rotating the jigged-up gear. The idea is to have something to roughly take the role of a machinist's dividing plate, but much smaller, lighter, and less precise.

Rummaging through my junk, I was fortunate enough to find a remote servo (for controlling valves and such), of roughly 1940's technology. The box consisted of a small split phase motor, 5 stages of reduction gears to a big gear on a half inch shaft, a rheostat, and some sort of mechanical balance contraption to function a bit like a bridge circuit to compare the rheostat in the box to some master rheostat somewhere else. When unbalanced, the balance would connect either phase of the split phase motor to AC power, causing it to rotate in either direction.

YouTuber AvE made a video about one of these servos

I removed the motor and the gears closest to it, leaving me with two gears of reduction. I mounted the last remaining shaft and gear backwards so that the long shaft protruded out through a hole I drilled in the aluminium plate, and attached a knurled knob to it for fine adjustment. I then needed some means of measuring rotations of the knob. I originally thought of using a counter from a tape recorder, but I found an old set of mechanical car odometer wheels in my parts collection. The gear attached to the last wheel was exactly the right pitch to match with those in my gearbox. Unfortunately, the first (0.1 km increments) wheel of the odometer was designed to jam up on running backwards. The rest of the wheels didn't have this property. So I moved the 0.1 km increment wheel to the other end, and made the gear drive the 1 km increment wheel directly. With the various gears, I got a 1:39.12 gear ratio.

gear cutting jig in action   cutting gears with tablesaw
The gear cutting jig in action on the table saw.

I then made a sturdy wooden box to attach to the aluminum frame of the gearbox, and keep my gears free from dust. The way I use the jig on my table saw sled causes the sawdust to get thrown directly towards the gearbox, so keeping the dust out is important. I even made a little window for the counter so no dust can get in that way either.

The whole mechanism runs really smoothly. A light twist to the output shaft will set the odometer counter spinning. This turned out to be a bit of a problem when cutting gears, because I always apply a slight amount of torque to the output shaft while cutting to avoid random errors from gear backlash. But with the mechanism turning so easily, I need to add a brake to the input knob so that it doesn't start spinning from this slight torque.

A gear ratio of 1:36 would have made the counter increments 1 degree. Unfortunately, that was not the gear ratio that I had. I was already lucky enough to have a gearbox as suitable for the job as it was. Besides, most gears I make don't have a number of teeth that would divide nicely into 360 anyway. As it is, I have 391.2 counter increments per revolution. When I cut the gears, I calculate how many increments I need to advance per tooth, and then use the constant calculation feature of a calculator to generate the numbers one at a time, and dial them on my counter. 391.2 increments is unfortunately too few to just round to the nearest, so I have to visually interpolate fractional increments to get decently accurate gear tooth spacing. For most gears, I have to make two cuts per tooth, so I have to dial in two angles per tooth.

The two gears above were the first ones I cut with my jig. To round the tops of the teeth, I used a square file, held at a 45 degree angle, and filed between two adjoining teeth. This kind of approximates the ideal involute spur gear tooth shape. These two gears mesh amazingly smoothly for being cut out of plywood. Naturally, cutting the gears straight out of plywood is not ideal because the wood grain is not aligned with the teeth. Better gears can be made by inserting individual teeth into a plywood wheel. Unfortunately, this means the teeth have to be fairly narrow to leave enough plywood to hold them. At left are two gears made this way. The small gear only has 14 teeth, and it runs fairly noisily. This can be desirable for drawing attention, as in "look, it has wooden gears!"

Since building the gear cutting jig, I have come up with a simpler solution for getting all the gear angles. It's much easier to just make a protractor template out of paper and stick it to the piece of wood. This should help align the angles, and with care should be more than accurate enough. You can make very accurate protractor templates with an inkjet printer using my free online gear template generator or you could buy a copy of my fancy gear template generator.

To my Woodworking index, or go to

Woodwork tools & techniques index