Some current automotive applications use a missing tooth wheel for this. A
toothed wheel is either bolted to the front (or rear) or the crankshaft, or the
teeth are part of the crankshaft itself. A magnetic reluctance coil is used to
create pulses as the toothed wheel goes by.

The missing tooth is used to determine absolute position.

The automotive computers very carefully monitor the changes to engine RPM as it
turns. Changes in rotational speed can indicate a misfire (this is reported via
the OBDII diagnostics system). They are also sometimes compared with ABS brake
sensors at the wheels as a sudden decrease in rotational speed combined with a
sudden drop in tire speed can indicate the vehicle just hit a bump and did not
misfire.

Now that I think of it, the ABS wheel speed sensors are very much like what you
are looking for.

-----Original Message-----
From: piclist-bounces@mit.edu [mailto:piclist-bounces@mit.edu] On Behalf Of
David Cary
Sent: Thursday, May 17, 2007 11:38 PM
To: piclist@mit.edu
Subject: [EE] measuring rotation rate

Dear piclisters,

What are good methods for measuring rotation rate?

I have a rotating device.
Currently I measure how fast it rotates using a
one-pulse-per-revolution optical sensor,
and then assume that it rotates at a fairly constant rate between pulses.

Lately I wonder -- is it really rotating at a constant rate?
Or are there significant angular accelerations and decelerations between pulses?

I think an angular rate gyro should be able to directly measure such
angular acceleration.
The fastest one I've found so far advertises "300 degrees per second!".
I suppose nearly 1 revolution per second is nice.
But my device spins around 50 revolutions per second.

I suppose some sort of rotary encoder (along with accurately timing
the width of each pulse) could possibly work.
But all the off-the-shelf rotary encoders seem to assume they are at
the center of rotation.
It would be more convenient if I could bolt something to the outside,
with a pattern ring about a foot in diameter glued to the stationary
part and a sensor bolted to the rotating part. (Or vice versa).
(Perhaps replacing, the one-pulse-per-revolution sensor I have now).

Suggestions?

-- 
David Cary
http://massmind.org/techref/member/DAV-MP-E62a
-- 
http://www.piclist.com PIC/SX FAQ & list archive
View/change your membership options at
http://mailman.mit.edu/mailman/listinfo/piclist

-- 
http://www.piclist.com PIC/SX FAQ & list archive
View/change your membership options at
http://mailman.mit.edu/mailman/listinfo/piclist