Need Solution for Peculiar Ring/Pointer Problem

[Maddy]

I have a peculiar problem. I have a circuler ring(1 metre dia, marked
from 0 to 360 in degrees) which is free to rotate. There is a pointer,
not connected to the ring which can be rotated seperately around the
ring, touching the ring. I want to digitally check which part of the
ring the pointer is touching(with an accuracy of about 0.1 degrees).
The ring cannot have any power source on it since it rotates. The
pointer can be powered from its rotation assembly. Can anybody give
any suggestions? (one suggestion was to put a wire on the ring and use
the wheatstone principle, but I am not sure how)

 

[Joseph Hansen]

Well, you could put a bunch of magnets on the ring, evenly spaced, and a
pickup coil on the pointer, and then use a counter to count how many magnets
the pointer had moved past... or you could make an optical encoder to do the
same thing. The electronics for both are very well established, it would
simply require designing the physical aspects of it. Say, an optical
encoder ring just below your ring, attached so it rotates with the ring, and
a photodiode / LED pair hanging from the pointer to read it. You can get
the encoder ring custom made for you (sorry, can't provide a source as I
have never needed the service, but I have heard that there are those who do
such for a fee....) or you could take aluminum or steel sheet and make holes
at the appropriate intervals to make it yourself. The reader can be ripped
out of almost any printer, many disk drives, etc. As for the magnet
solution, the evenly spaced magnets could be done with a length of magnetic
cassette tape, and use the read/write head to write the evenly spaced spots
onto the tape, and then to read them during operation.

If you don't want something hanging from the tip of your pointer, you could
just use the encoder ring to measure how far the ring had turned and in
which direction, and another to measure the movement of the pointer, and add
or subtract the two values appropriately. There're lots of tutorials on
encoders on the web that will teach you how to get direction and distance
information out of an encoder.
news:[email protected]...

 

[Robert Baer]

Well, you could put a bunch of magnets on the ring, evenly spaced, and a
pickup coil on the pointer, and then use a counter to count how many magnets
the pointer had moved past... or you could make an optical encoder to do the
same thing. The electronics for both are very well established, it would
simply require designing the physical aspects of it. Say, an optical
encoder ring just below your ring, attached so it rotates with the ring, and
a photodiode / LED pair hanging from the pointer to read it. You can get
the encoder ring custom made for you (sorry, can't provide a source as I
have never needed the service, but I have heard that there are those who do
such for a fee....) or you could take aluminum or steel sheet and make holes
at the appropriate intervals to make it yourself. The reader can be ripped
out of almost any printer, many disk drives, etc. As for the magnet
solution, the evenly spaced magnets could be done with a length of magnetic
cassette tape, and use the read/write head to write the evenly spaced spots
onto the tape, and then to read them during operation.

If you don't want something hanging from the tip of your pointer, you could
just use the encoder ring to measure how far the ring had turned and in
which direction, and another to measure the movement of the pointer, and add
or subtract the two values appropriately. There're lots of tutorials on
encoders on the web that will teach you how to get direction and distance
information out of an encoder.

If i remember correctly, ther are afew rotary optical encoders
available intheDigiKey catalot (switch section).
Also try Mouser.

 

[JeffM]

rotary optical encoders

Robert Baer

Yup.

Maddy,
The 1st thing that occurred to me was an absolute optical encoder.
http://images.google.com/images?q=optical-encoder+slots
(The entire Chapter1 page at polyu.edu.hk is good stuff.)