Linear slides
 

I've seen a couple of teams use Bosch extrusion or 80-20 to make accurate linear slides. I was wondering how the slider actually worked, or if you didn't use either of those, how you made yours work?

We've only made three real efforts at linear slides. The first was a cylinder-in-a-cylinder which worked well. The second was sandwiching a piece of aluminum extrusion between two pieces of HDPE, which didn't work well at all. The third was to cut a giant track in a piece of solid aluminum, stick in a smaller piece of aluminum, and guide the smaller piece with dowel pins in two slots cut in the exposed sides. That gets a bit iffy, because it's a lot of machining for relatively little gain.

Any thoughts?

Re: Linear slides
 

There are many companies which make linear slides.

One light duty slide is the "drawer slide".

Lowes, Home Depot, etc. sell these. General Devices and Accurride are industrial manufacturers.

There are many other industrial-type slide mechanisms. Igus has a low profile slide that looks really cool (I forget the part number). Rollon has some excellent slides that are relatively cheap. Thompson, IKO, NSK, THK, Hiwin (and others) all sell slides that range from medium to very heavy duty.

If you can choose one of these company's slides that are standard parts, you will be much better off with regard to price and lead time.

Andy B.

Re: Linear slides
 

For our arm in 2002 we used "linear bearings" found in 80/20's catalog. They are designed to bolt on to the extrusion and have HDPE for the slider to ride on. The rest of the assembly was just extrusion. It worked extremely well.

Re: Linear slides
 

I agree with Andy's comments. Buy em if you can.

In 2000 we made our own "roller elevator lift" using extruded sections and home-made rollers. It was 5 sections, and it extended to 3X length using a cable pull of only 1X. It was a mechanical multiplier. It was fast and low-friction (rolling not sliding). But... it was VERY high-maintenance. The biggest problem was keeping the roller axles securely aligned to their extruded pieces. The tiny little fasteners that you like to use with the extruded stock just dont have that much strength and stiffness.

It can be done, but be careful if you make your own. Maybe test it on a fixture - early - to debug.

Ken

Re: Linear slides
 

We generally make our own slides. Last year was the first time in several years that we did not have a linear slide on the robot. We take 1.25x1x0.125 aluminum C-Channel extrusion and build it into a frame. The ends of the frames have solid blocks of aluminum that are bolted into the channels. We make teflon or delrin rollers that are bolted into the blocks with shoulder bolts. The rollers are slightly undersized to move easily in the channel. Basically it is a big, heavy duty drawer slide and they work great. We originally developed it for the 2000 game and designed it to be able to pull BOTH robots onto the bar, assuming our partner could grab onto us. In 2001 our robot could place a ball on a goal from the floor when the goal was on the ramp, so it is fairly stiff as well.

We have also used HDPE or similar materials as guides with reasonable success, but the key is always power both ways. Otherwise your mechanism will get hung up. If you haven't already, check out my white paper on lifts and arms. I'll also dig around and see if I have our designs readily available as AutoCad files.

The the most important things are:

1) Each piece has to fit well, no binding allowed. Leave adequate gaps between all parts for smooth operation. We like 0.001-0.005" clearance between rolers and shafts and 0.010 between rollers and channel.

2) Minimize slop. While you need some clearance between parts, you don't want too much either. Much more than the above and things start flopping around all over.

3) Power both ways. Yes, gravity is a pretty reliable way to get a lift to come down, but we've found that you need to keep pulling on things to keep them lined up properly.

4) If you are using a winch set-up to power your motion, be careful how you rig it. We prefer the "cascade" type mentioned in my whitepaper because it forces the motions of multiple stages to be even. If the extensions get uneven then you can have jamming problems.