Linear motion

[bool101]

Hi, I'm writing to get some advice on how to take a the circular motion from a DC motor and convert it to linear motion. I have nil background in mechanical engineering so I'm looking for either an intro guide or an off the shelf kit. I'm understand the basic concepts but haven't looked at the math behind it. Any advice would be appreciated. Thanks.

 

[kchriste]

A worm gear type setup is the simplest: Basically turn a long machine screw with the motor and the "nut", if prevented from rotating, will travel up or down the length of the screw.

 

[Tomble]

There's also the "rack and pinion" type arrangement, where a gear (pinion) meshes with a long flat toothed strip (rack). The relative motion of one vs the other is at the same speed as the teeth of the gear, which is proportional to the angular velocity and to the width of the gear.

Unfortunately has the disadvantage compared to the worm gear arrangement, that it has much lower "mechanical advantage", IE it can give less force. In some cases that might not matter. The flipside to having lower mechanical advantage is a correspondingly higher "velocity ratio" so the rack moves faster. Regardless, I'd generally opt for the worm gear unless there was a good reason not to, especially as most cheap motors are so very fast.

 

[Hank Fletcher]

The worm gear solution described above is possibly the solution with the most positioning accuracy - you can see it at work in homemade CNC machines.

The simplest and fastest solution is to use DC motors to power wheels on a car. The wheels go around, the car moves forward - circular to linear motion conversion.

The next step up from that is a wheel with a pulley, which can be just as fast as a car but not as accurate as a worm when in comes to positioning*. Still, it's the solution I'm opting for in my trombone playing robot to move the slide back and forth, and I'm seriously considering using some small motors with a single pulley in the design of some small actuators.

*Sorry that's not exactly true. A single pulley will never be as accurate or have the mechanical advantage a worm can provide, but many pulleys and a lot of line can. Many pulleys and line can also be much quieter and result in less wear in the long run, whereas all worms inevitably wear and will get noisier as a result. Noise isn't a big deal for most applications, but is in mine. The reason for favouritism of the worm gear is that it provides a lot of mechanical advantage in a very compact form, whereas a lot of pulleys (or, as another example, a lot of spur gears) take a lot of space, manufacturing, maintenance, etc.

 

[Leftyretro]

And of course there is the classic crankshaft that converts a rotor to linear motion.

Lefty

 

[Sceadwian]

They do make linear stepper motors. MPJA'a last catalog had one in it.

 

[stevez]

You could also google on ballscrew. It is very similar to the worm gear - though it's little more than a threaded rod or shaft that drives a matching nut. Best example I can think of is the carriage drive on a lathe.

 

[shirazmacuff]

Hi, I’m working on a similar project. I have built the slide with mild steel for rigidity. Normally it should have brass bushing because you don’t want the steel slide to wear quickly. It’s a lot of work and you need material, drill press, tap and die set and some others tools. I know the post is very old but if you need details reply and I’ll see what I can post.

Now I need to build a system which would track the position of the slide. I’m thinking to use LVDT. The linear slide is pretty small (6 inch movement) and I need it to have approx 0.01 inch tolerance. The commercial ones can work but funds are pretty low right now so I’m hoping it can be built. I also have a lot of parts and scrap, if any can be used it would be of great help. I would welcome any ideas.

 

[Leftyretro]

Hi, I’m working on a similar project. I have built the slide with mild steel for rigidity. Normally it should have brass bushing because you don’t want the steel slide to wear quickly. It’s a lot of work and you need material, drill press, tap and die set and some others tools. I know the post is very old but if you need details reply and I’ll see what I can post.

Now I need to build a system which would track the position of the slide. I’m thinking to use LVDT. The linear slide is pretty small (6 inch movement) and I need it to have approx 0.01 inch tolerance. The commercial ones can work but funds are pretty low right now so I’m hoping it can be built. I also have a lot of parts and scrap, if any can be used it would be of great help. I would welcome any ideas.

True LVDTs are a bit hard for home-brewing as linearization and temperature compensation can get pretty hairy. There are many applications that use optical sensing of a strip with holes or slits at the step resolution needed, that's a lot simpler for home-brewing.

Lefty

 

[KMoffett]

Also, the pulley and belt conversion of rotary to linear...think scanner!

Ken