shaft encoders

[Lex]

has anybody ever used or tried to use an optical thatis laser mouse as a shaft encoder.

[bojan]

Hi Lex,
As far as I know, there are two types of encoders widely used for precise measurement of movement… optical (wheel or tape with slits) and magnetic.
Those two methods operate inherently free of slippage (electronic or mechanical) and they are repeatable, in both incremental and absolute versions.
The problem with optical mouse is the method used for “coupling” with the moving part which can be anything. This is OK for a mouse… However, this feature makes the outcome somewhat dependent of surface texture, colour and contrast and algorithm used in mouse… And I am not sure the repeatability can be controlled properly (obviously, if you move in one direction x metres, and then back the same x metres, your encoded readout should be the same. One count less or more and you have slippage.
In my view this can be achieved only by using the repeatable pattern on the moving surface and simple detector circuitry and this is exactly what was used in optical and /or magnetic encoders J
So, I would recommend sticking to reliable methods.
Some time ago I have seen on the web one such application for Bartel’s scope program, the author built an interface box to read two optical mice (one for each axis), but I can not find the page any more….
Have a look at this document, it may be interesting...
http://www.oru.se/oru-upload/Institu.../rapport21.pdf

[DavidH]

Lex,

I had thoughts along these lines, but a simple test of an optical mouse shows that the "slippage" inherent in returning the cursor to a known point on a computer screen means that the mouse (in its unaltered state) will not be up to the task. I also tried printing a sheet with alternating light and dark lines to try to improve the repeatability, but this did not work either. At the time I was thinking of using the mouse as a linear encoder for focus control.

Maybe somebody with more knowhow or persistence could manage to make it work, but from my point of view buying a linear encoder will be simpler and more effective.

Regards,
David.

[Lex]

Hi bojan thanks for the link its worth reading all pp 12 to20 give a good idea
of what is possible .... simple table top experiments of our own found that if you moved the mouse on a thin disk with hole ie az axis and keep the subsurface stationary you get a good result perfect circle now to get that infoto be read by guide
davidH go and look at the sight bojahas suggested
what i presumed was a laser is an LED

[bojan]

I found this on yahoo group.. the conclusion is the same: do not use optical mice :-)

http://tech.groups.yahoo.com/group/s.../message/16829

[bojan]

Of course, to fully test mouse encoder repeatability (regardless of type) acceleration must be switched off (some mouse drivers allow for that in configuration file, some of them want parameter in the command line, and some are hopeless).
Also, highest sensitivity must be specified, this will make mouse to use maximum resolution.

[Lex]

seems to me they had a mount problem or a mounting problem or a percieved mountain of problems. perception is everything.