Alternative to Ball Screw or Rack-n-Pinion

[handlewanker]

Armchair engineering......I like that.....the horizon is boundless and allows you to explore without reservation of inhibition anything imaginable.

The 1/4" cable drive does have some limitation...as it's a twisted bundle of wires the application of tension is enough to make it stretch like a rubber band, even if not quite as much, but in a drive situation you would have the cable attached to ends of the machine frame and provided the frame did not flex the only other lost motion is the flexing in the cable under load....the longer the cable the more "backlash" or stretching appears.

If'n you get .02mm (.001") per 300mm (12 inches) elongation from the stretch in the cable....then you have backlash each time you reverse direction and it is accumulative with the cable length.

You don't get stretch so much in a solid wire, but you have the problem that as a solid high tensile wire increases in thickness, to offset breaking under tension, it becomes too rigid to wrap round a smaller shaft, drum or pulley, and the larger the driving shaft diam the less resolution you have.

You could have a number of solid wires of smaller diam laid side by side to enable them to wrap round a smaller pulley, but as the length increases the sag if not supported would cause problems.....and as it is desirable to wrap the wire round a shaft to get travel length, this then makes multi wires laying side by side a problem.

I toyed with the notion of a chain as it is capable of flexing round small diams, but that is only an armchair solution.

The neater rack and pinion method has already been noted to be successfull, so all roads lead to Rome eventually.

In any form, a rack and pinion drive would have to be hardened to make it proof against pinion/rack tooth impact wear...not apparent at first but as cratering develops so backlash starts to appear.

I have never seen a geared tooth set up using soft gears that withstood the pounding that the teeth are subject to in impacting on each other in a drive situation.

Alternatives are fine as long as they go one step further in the path of progress as opposed to doing it the same way but differently.

When all things are equal the simplest is always the best.

Equal means achieving the SAME results with different methods.

On the scale of complexity and cost, we have at the top of the ladder..the ball screw, next comes Acme screw drive followed closely by rack and pinion then cable and pulley.

It goes without saying, one man's meat is another man's poison...and we all would like the sheer accuracy, freedom from backlash and simple fixing of the ballscrew, but not the cost.

I wonder if'n a hydraulic cylinder to shift the table and a linear encoder to know where you were would be the ultimate in positioning and accuracy......there could be no backlash and the encoder would tell you to 100th of a mm where you were, and speed of drive is in the hydro pump capacity.

Many years ago when I worked a Kearns horizontal borer, the table movements were all indicated with light verniers, and you can't beat a vernier for knowing your exact precise position, even with high speed positiong drives going back and forth.....you only have to read them.....and backlash is non existent....and resolution is in the vernier engraving increments.

Supposing we had a hydraulic drive with a digital read out etc etc....someone with knowledge of encoders could come in on this and explain the workings of a linear encoder to position the table and the problems encountered in the reading at speed.

Would you lose position if'n the encoder missed reading the linear encoded table scale, or does it actually read the position on a scale as in a calibrated vernier...IE, the position is read as opposed to counting the number of lines that flash past the reading head....miss a line and you are forward or back of position, whereas with a calibrated vernier you are at one and only one position at any time.
Ian.

[JoeCosman]

by soft gearing I'm assuming you mean the delrin gear rack I posted from SDP-SI?

I figured that the shear loads on those gears would be the same amount as 5 start acme with acetal nut. I'd have to do the math to be sure, but from a visual standpoint, I think it would be in the envelope. my design right now is just plain chain/timing belt drive, more because I would be able to mill aluminium if I wanted to. most of my materials would be either foam, cast urethane, cardboard or pvc with the occasional mdf board thrown in for good measure.

so.. what about using a drill bit at an angle to 'worm' along a series of rods, like kylestrong87's idea for rack? heck, what about finding a gear size that fits perforated DIN rail?



or using a 36" drill as a leadscrew or ballscrew?

all roads lead back to a rack and pinion or chain when it comes to simplicity. But to me, it seems that the market is cornered simply because very few people are thinking farther outside the box. everyone wants cheap and accurate and scaleable. there just needs to be that one spark that can push everyone to an inexpensive alternative that alway's existed, but nobody thought to look for.

btw, I think BillTodd should cut his sprocket design and test it out.

[handlewanker]

The inescapable fact is the more exotic you go the more complicated the solution.

With a resolution need of +- 1mm you can just use good old Allthread and use a common hex nut to back it up to cut down the backlash......it will work for a while till the middle of the thread wears and you attempt to adjust the back up nut and find it gets tighter when you move to the ends, and still has backlash in the middle.....so you go to a spring loaded back up nut and this is OK provided you don't overcome the spring pressure on reverse drive.

Then the need escalates to 1/10mm resolution and the Allthread gets changed to Acme thread.....with back up nut etc etc.....same time you want to do a bit of Alluminium cutting and the resolution, while good for wood at .1mm is now required to go to .o2mm....the expectation is further satisfied by going to a ball screw, and so the machine, whatever has come of age and does sophisticated work with tolerances held to .02mm in every plane.

The problem is some of us want .01mm resolution and no backlash using Althread and Delrin nuts......the pigs, having been fed, will soon be ready for flight.

First lesson to be learned...you can't use soft metal on soft metal and have no wear.
Second lesson to be learned......you only get what you pay for.
Third lesson to be learned.......the law of physics is exact and absolute.
The fourth and final lesson to be learned is...business is business and pleasure is pleasure.....the two can never meet.

I quite support any method that can be thought of to drive a load along a plane path, but at the same time I would like to know what resolution the proposed design could provide.....eg maybe class 1------ .002mm (.0001"), Class 2-------.02mm (.001"), Class 3-------.25mm (.010") and class 4-------1mm (.040") etc etc.

This will seperate the ball screw from the wrapped string design and make the proposer think as to what it will best be suited for.
Ian.