I want to design and create a gearbox that allows an input shaft to be continuously rotating in one direction, while the output shaft rotates one full turn in the same direction but then half a turn in the opposite direction and repeat that cycle.

The application for it would be attaching to a drill and automatically tap holes to cut threads.

The only way I can see it working at the moment is by creating a gear which has teeth missing so that it rotates to a certain point (1 full turn of output) and then slips, being pulled back by a return spring) and hits a stop to restrict back rotation to half a turn.

However I have concerns about the reliability of this design and the fact that it would be difficult to allow it to work in both directions.

Having a mechanism that can operate in both directions by changing the rotational direction of the input shaft is ideal because it would allow it to be used to cut both right and left hand threads.

Any ideas about how to achieve such a gearbox would be greatly appreciated.

And sorry if this sounds really confusing, it’s quite difficult for me to explain.

  • $\begingroup$ It's difficult to imagine why anyone would go to so much trouble when a servo motor, or even a stepper, would do the job at a fraction of the cost and with programmable control. We'll have a think though but I doubt that it's going to be called a "gearbox". $\endgroup$ – Transistor Apr 30 at 19:41
  • $\begingroup$ Yes Transistor, I understand your point but instead of going though the process of designing a dedicated tool for driving the taps, I wanted to create a device that could be driven by a standard drill through a hex shaft. $\endgroup$ – Billy Apr 30 at 20:02
  • $\begingroup$ If you can dig up a description of an automatic tapping machine (or maybe some lathes or screw machines) from the 1950s, roughly, it might have what you want. $\endgroup$ – Pete W Apr 30 at 20:20
  • $\begingroup$ Note that if you want to "ping" someone in the comments or posts you need to use the form @username with no spaces (even if there are in the username!). Then your comment will appear in their inbox. $\endgroup$ – Transistor Apr 30 at 22:05
  • 2
    $\begingroup$ Why? Taps are designed for use on lathes which have the flutes shaped to remove waste and tap to the full length of the tap including the shaft. Driven in to depth then wound out - none of this turn and back stuff. $\endgroup$ – Solar Mike May 1 at 5:02

Use a planetary gearset. Drive the center continuously at the speed you want to tap. Use a mutilated gear to drive the ring in the opposite direction intermittently and at a speed that will back the chuck up. The chuck runs with the pinion gears. There needs to be a dog clutch rigged to the mutilated gear to hold the ring gear fixed when not backing it up.

Alternatively, you can built an internal star wheel directly into a planetary gearset. Drive the center, fix the ring, and attach an internal star to a pinion gear. See figure 15.6 here

enter image description here


The best I can think of is a barrel-cam indexer mechanism.

enter image description here

Figure 1. Image source: nolte-nc-kurventechnik.

See the Motion index drive on YouTube for a better view with fewer cam followers.

The animation in Figure 1 shows a helical barrel cam but there are many other options. For industrial indexers there is often a period of "dwell" put into the motion so that with continuous rotation of the barrel there is intermittent indexing of the table. In your case you would put a portion of reverse motion rather than dwell.

I know this would be relatively easy with a four-way indexer which, in your application, would give 90° forward, 45° back. This might be even better for the tap than full turn forward, half turn back.

Note that input and output for this configuration are at right angle to each other.

YouTube video. The video shows back and forth motion and is worth a look.

  • $\begingroup$ that's a nice solution! $\endgroup$ – Pete W Apr 30 at 21:52

The simplest approach, I think, will be using two stacked gears on a single shaft - one with 2/3 of its teeth missing, one with 1/3 of them missing. The gear with 1/3 gears missing engages to output gear (2/3 of the time) through two gears, the one with 1/3 missing - through one gear, spinning freely while the other one is engaged and vice versa. This way the output will rotate 2/3 of the time forward, 1/3 back. Adjusting gear sizes (teeth ratios) will easily convert it to 1 rotation forward, half back.

It's going to take some precise measurements and good precision to assure the gear spinning freely meshes perfectly with the "arriving" toothed part - might require first and last tooth of the propulsion gears to be shorter than the rest.

Please forgive the unprofessional software, messy look and inaccurate operation of the visualisation - it's what I had at hand and it's to show the idea.

enter image description here

  • $\begingroup$ Sorry for late response, I haven’t checked this post in a while. I’m not exactly sure how the method you are describing is supposed to work given that if the gears are stacked they would be fixed to the same shaft so even if they do have 2/3 and 1/3 of the teeth missing respectively, I don’t see how the direction of the output would change. Maybe I’m missing something. Would you be able to draw out a rough diagram? $\endgroup$ – Billy Jun 18 at 4:20
  • $\begingroup$ @Billy One of the two stacked gears drives the output indirectly through an even number of gears, the other through an odd number. Added a gif of something approximating the idea. $\endgroup$ – SF. Jun 18 at 10:15
  • $\begingroup$ That’s really clever. Makes perfect sense now. $\endgroup$ – Billy Jun 18 at 23:48

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.