Searching \ for 'angle sensor from mouse?' in subject line. ()
Make payments with PayPal - it's fast, free and secure! Help us get a faster server
FAQ page: www.piclist.com/techref/io/sensors.htm?key=sensor
Search entire site for: 'angle sensor from mouse?'.

Truncated match.
PICList Thread
'angle sensor from mouse?'
1998\01\19@120427 by Cser Laszlo

flavicon
face
Hi!

I would like to build an angle sensor, and I thought the sensor in a
mouse would do as well. Could anyone tell me the features of the mouse
sensor? Sensitivity, ... ? I'd like to measure angles with a resolution
of 0.57 degrees. What kind of output comes from a mouse? What signals are
needed to be applied?

Bye: cotan

1998\01\19@140343 by mdmiller

flavicon
face
Cotan,

The sensors in most mice are basically "cheap" optical encoders.  If an
optical encoder is used, you would need to have one that counts at least
360 / 0.57 = 632 counts per revolution in order to measure 0.57 degrees.
Although a mouse does not have this sort of resolution, I'm sure encoders
that do are available.  Additionally, communicating with a mouse would be a
little more involved than an encoder.  Many encoders have two phase
outputs.  Each output has a 50% duty cycle, and is 90 degrees out of phase
with the other.  This allows you to record the pulses produced by each
output and measure distance and direction with two pins of your controller.
Communicating with a mouse is a little like "bit banging" RS-232 _and_
knowing the protocol a mouse uses to communicate with it's host (not to
mention the additional circuitry required - ie: a mouse draws it's power
from it's host).  I believe Microchip (http://www.microchip.com) has an
AppNote on building a mouse controller with one of their PIC devices ...
quite insightful into how a mouse works.

Hope this helps.

-- Mitch


------------------------------
Mitch Miller
spam_OUTmdmillerTakeThisOuTspamnque.com
------------------------------


On Monday, January 19, 1998 12:47 PM, Cser Laszlo [SMTP:.....cotanKILLspamspam@spam@EIK.BME.HU]
wrote:
> Hi!
>
> I would like to build an angle sensor, and I thought the sensor in a
> mouse would do as well. Could anyone tell me the features of the mouse
> sensor? Sensitivity, ... ? I'd like to measure angles with a resolution
> of 0.57 degrees. What kind of output comes from a mouse? What signals are
> needed to be applied?
>
> Bye: cotan

1998\01\19@143611 by Andy Kunz

flavicon
face
>sensor? Sensitivity, ... ? I'd like to measure angles with a resolution
>of 0.57 degrees. What kind of output comes from a mouse? What signals are

Contact skipspamKILLspamrcboats.com and ask him about the sensor for his APC-3.  It
has VERY good resolution, and is quite simple to interface with.

Vibration could a problem - you want a relatively steady surface.

Andy


==================================================================
Andy Kunz - Statistical Research, Inc. - Westfield, New Jersey USA
==================================================================

1998\01\19@170221 by Mauro, Chuck

flavicon
face
Cotan,

As a very recent and former mouse EE at Microsoft, I can tell you that
Mitch Miller is basically correct about the operation of today's typical
opto-mechanical mice.  Most mice have a 400 DPI resolution, so a single
mouse-ball increment is 0.9 degrees.  Mitch is also correct that there
are really no commercially mice that can resolve 0.57 degrees...

Mice really use 2 orthogonally opposed shafts and opto-sensors to
resolve X/Y motion.  The sensors employed today are PTR's (high gain
phototransistors) that use dual transistor elements in a single package,
and provide a quadrature based output.  The mouse packets that are sent
to a host, whether they are PS/2 or serial, only contain changes in
mouse state data that has been derived from this orthogonal relationship
(or button activity).  No other angular information is present or
implied.  Mice have been designed (at least experimentally) to encode
body rotation as well as translation, but I do not know of the model or
manufacture that tried this (basically, this requires 2 sets of
orthogonal encoders)...  If you are just interested in angular
displacement of the ball when the mouse is moved (again: 0.9 degrees),
then simple small displacements will provide that type of information
(delta X or delta Y of 1).  Unfortunately, today's uC algorithms in the
mice need to take mechanical slop and other process variations into
account when monitoring the mouse sensors, and this typically involves
some type of hysteresis on the encoding mechanism that often results in
a logical loss of small increment data (this is done to avoid mouse
cursor jitter or drift when the mouse is stationary, and occurs when the
uC is trying to encode the signal from a PTR that is just at the input
pin switching threshold, [resulting in loss of electrical noise
immunity])...

PS/2 mice operate from +5V, and the AC and DC information concerning the
interface is (was) available as an IBM technical document (PS/2 Mouse).
The Serial mouse protocol is available from Microsoft - although as a
former insider, I had access to all of the internal doc's, and I've
never had to deal with the external developer support group, so I don't
know what kind of experience you are going to have...  See the Microchip
App notes that Mitch referred to - it contains very good information on
this topic...

Good luck in your attempt at adapting a mouse for your needs...

Chuck Mauro

> {Original Message removed}

1998\01\19@175650 by Bill (WL) Boulton

flavicon
face
At 10:47 AM 19/01/98 -0800, you wrote:
>Hi!
>
>I would like to build an angle sensor, and I thought the sensor in a
>mouse would do as well. Could anyone tell me the features of the mouse
>sensor? Sensitivity, ... ? I'd like to measure angles with a resolution
>of 0.57 degrees. What kind of output comes from a mouse? What signals are
>needed to be applied?
>
>Bye: cotan
>
cotan

Don't know what you mean about "angle sensor" in a mouse. Mechanically, the
ball transfers motion to 2 shafts, each with a simple rotary encoder
generating a 2 bit Gray code. Sensitivity depends on the size of the ball,
the size of the collar on the shaft and the number of slots in the disk at
the end of the shaft. All parameters are at manufacturers discretion an the
ball to shaft contact is by friction and not positive. You may be able to
apply the principal to your problem, but I don't think the actual hardware.

Bill

1998\01\20@091250 by Charles Hoss

flavicon
face
Cser Laszlo wrote:
>
> Hi!
>
> I would like to build an angle sensor, and I thought the sensor in a
> mouse would do as well. Could anyone tell me the features of the mouse
> sensor? Sensitivity, ... ? I'd like to measure angles with a resolution
> of 0.57 degrees. What kind of output comes from a mouse? What signals are
> needed to be applied?

cool idea !
I think you'll need some gears to
achieve the resolution you need ,
and I'd suggest a zero set point too .

bye
charley

1998\01\20@140659 by Lee Jones

flavicon
face
> I would like to build an angle sensor, and I thought the sensor in a
> mouse would do as well. Could anyone tell me the features of the mouse
> sensor? Sensitivity, ... ? I'd like to measure angles with a resolution
> of 0.57 degrees. What kind of output comes from a mouse? What signals are
> needed to be applied?

As other people have mentioned, a mouse's encoder
probably doesn't have sufficient resolution.

It might if you gear it up sufficiently using one or
more stages.  I'd use a cog belt to avoid slipage.
And sprockets plus cog belt is usually easier to make
in small quantities than gear trains.

If you're willing to buy an encoder, U.S. Digital makes
an extensive line.  They have shaft encoders with over
2000 counts per revolution.  Output signal is quadrature
encoded.  A friend of mine uses them for high resolution
telescope pointing (a commercial product).

They also have disk and linear patterns if you want
to "roll your own" encoder.  They have a web site at
http://www.usdigital.com.
                                               Lee Jones

1998\01\20@170214 by johnb

picon face
Lee Jones wrote:
> As other people have mentioned, a mouse's encoder
> probably doesn't have sufficient resolution.
> It might if you gear it up sufficiently using one or
> more stages.

Have a look inside a low-cost Hand Scanner. You find a high gear ratio
from the roller to a disc/photo-cell arrangement that is very similar to
the one inside a mouse. What you don't get, of course, is a
bi-directional system, but it almost certainly gives the resolution
needed (from memory - it's a long time since I took one apart).

John Blackburn,
South London UK.

More... (looser matching)
- Last day of these posts
- In 1998 , 1999 only
- Today
- New search...