Wild thoughts about encoders

[Konstantin]

I got this crazy idea about encoders for servomotors, well basically instead of using 2 separate encoders just use one mouse optical sensor, novadays optical mouses exel in positioning so why not using it to encode the XY table movement. One optical sensor under the table to tell the movement change to the computer and with some software feedback the driver or send the correct positioning directly into the driver.

Now I must add that I now nothing about electronics nor programming, and it above it is just a crazy idea I needed to flush out of my mind. Of course nothing is going to be easy in this theoretical setup.

One advantage of wich I can think of is that the actual XY table shift/position will be encoded (if we suppose that the optical sensor reading is error free) and not the lead screw revolutions thus no backslash compensation is needed.
Maybe the same CNC program could host the encoder mouse driver and send corrections to the servo.

Maybe I am talking loko here, I should go to sleep already.

Konstantin.

PS:
What do you think about it?

[Stevie]

my intelly mouse explorer is sitting here in pieces; it packed up a week ago; I'm not sure quite whats wrong with it; but i think the optical is still working; it was the center button that packed up

[ToyMaker]

The highest resolution opti-mouse I've seen puts out 400 counts/inch (~15.75 counts/mm) in x and y axes. If 0.0025 inch (~0.06mm) is sufficient resolution for your application, go for it. In a larger wood carver this would be excellent, in a metal working machine it is not very good.
Not all the mice (mouses, meese, ?) out there have this high resolution - caveat emptor.

robotic regards,

Tom
= = = = =
Q. "What's the difference between a fisherman and a lazy student?"
A. "One baits hooks; the other hates books."

[Konstantin]

I think thre new mouse fro M$ will do better, the Intellimouse Explorer 3.0

Benefits of the new IntelliEye technology

Improved tracking speed: The most common complaint optical mouse users have is that the cursor gets “lost” during periods of quick hand movement. The problem is, most optical mouse products can only move up to 14 inches per second. But usability research indicates that computer users can move the mouse up to 30 inches per second - far faster than the tracking capability of today’s mouse products. Microsoft’s IntelliEye optical technology is better. It enables the mouse to keep up with the hand, tracking up to 37 inches per second.

Capturing 6,000 frames per second, it delivers up to four times the performance of other optical mice. This superior tracking speed translates into smooth, accurate cursor performance

[Bloy2004]

I'm not sure, but the main problem would seem to be keeping the optical pickup surface free from impurities that could/would quickly degrade the performance.

[Ken_Shea]

The most common DPI resolution is 200-400 DPI, but now you can get 1000 DPI optical

http://www.razerzone.com/productoverview.html

I had posted a thread on using the optical mouse for A DRO, but the 400 DPI would limit the acceptable accuracy for machining in most cases. With the 1000DPI it is a whole different ball game.

I may look into it on the DRO part, as writing a small VB program to interface with the mouse is very easy.

[dherbman]

I like the idea. It could work, but I fear that there may be an accumulative loss of signal which would translate into large inaccuracies. It's no problem for a mouse if it loses a few beats here and there, but an encoder can't lose count.

[ynneb]

To add to your idea, I was thinking about the pits in a CD/ DVD.
I was thinking about if you had a linear arrangement made out of the same stuff as a CD.
You would then have a laser shining on this linear CD surface. Can you imagine the accuracy?

[ynneb]

With this idea, the laser would read the CD surface track and know exactly where it is. There would be no need for zeroing, because the reader would know exactly where it is, even if the machine was turned off and on.

The machine would have no problems with backlash because it reports exactly where it is, and not relative to belts, screws, etc.

This would simplify the electronics for servo drives, eliminate servo encoders, etc.

It would be impossible for the machine to lose count with this method because each step has its own code/number that it feeds back.
The whole reading unit would be encased to protect it from dust.

Just In case this idea is not patented, I ask that everyone note the idea was published on this forum and the date of its publishing. Just because this idea is in public display does not mean it is for public use. If there is a developer out there who wants to use this idea then plase contact me. ( Sorry to be a goose, but if it means a easy $10, 000 for me then so be it)

I have been thinking about this idea and similar for a while now. Including a laser shining on the tool itself and reporting positiong from that.

EDIT : What I mean is, anyone can use the idea for private use, but not for comercial use.

[Stevie]

lets take it a step more; use a CD burner to encode the strip; then the marks would be as accurate as the reader head; plus alinement would be automatic; no need to aline the strip

[ynneb]

Stevie, you are not getting any of my 10 thou........get your grubby hands off it

An excellent idea.

[Stevie]

just name me in the accounting for a small payment every month