RE: [CAD_CAM_EDM_DRO] Re: Optical mouse for DRO anyone?

I had also floated this idea a while ago.



The specs I found (agilent) provided a quadrature output from the
optical chip. The mice I inspected usually have

An interface chip (usb/ps2) that receives the quadrature output from the
sensor is responsible for counting and serializing and usually has
an overflow status indicator for the counters.



I assume the tracking capability of the sensor chip is limited by the
sample rate against shapes moving out of view.

I'll assume we only use the sensor chip providing quadrature output.



I think you want a high contrast pattern with sharp edge.

I was thinking maybe on an angle with the possibility of using both X
and Y outputs in something of a 4 phase output vs. 2 phase to
generate more precision.



I think repackaging the sensor/led combo into a 2"x1" pcb and using the
optical fence/strip from a dot matrix ought to work.

I haven't done this yet.



I'm also thinking of using one as an encoder on DC motor for servo
application.

I haven't done the repeatability study, and I don't run a shop either,
If I were earning a living at it I'd buy good scales.





________________________________

From: CAD_CAM_EDM_DRO@yahoogroups.com
[mailto:CAD_CAM_EDM_DRO@yahoogroups.com] On Behalf Of dgoadby
Sent: Tuesday, April 17, 2007 2:15 AM
To: CAD_CAM_EDM_DRO@yahoogroups.com
Subject: [CAD_CAM_EDM_DRO] Re: Optical mouse for DRO anyone?



--- In CAD_CAM_EDM_DRO@yahoogroups.com
<mailto:CAD_CAM_EDM_DRO%40yahoogroups.com> , "mrclicit" <chris@...>
wrote:

>
> Has anyone had success using an optical mouse or it's components to
> make a DRO or an accurately reading scale? If so, can you share any
> tips, links, or other info?
>
> TIA, Chris
>

Hi Chris,

I've been playing around with this idea (no guns involved here) for a
while. The original chip (designed by Agilent) had a 16X16 array and
is really a camera plus some smart software in the chip. A good start
to interfacing one of these can be found here:
http://www.contrib.andrew.cmu.edu/~ttrutna/16-264/Vision_Project/
<http://www.contrib.andrew.cmu.edu/~ttrutna/16-264/Vision_Project/>

A Thomson version of this chip has a resolution of 1600cpi with a
20X20 sensor and it tracks (acc to data sheet) at 40 inches per second.

Initially I am looking at making datums for my CNC set-ups using
these. If I can get 0.001 reliably then I will make a non contact
datum. Using a longer lens I will then look (no pun intended) at
workpiece detection and even broken tool detectors might be possible.
Hopefully a compressed air bleed across the lens will keep it clean.

One promising idea is to track using moire fringes for the scale.

Regards

David





[Non-text portions of this message have been removed]