accurate angle measurement

Hi everybody,

Can anyone suggest me some idea for measuring angles up to seconds? or
Where can I find the principle used by theodolites in measuring angles?

Thanks a lot

 

That is a bit little information. There can
be optical masks with micrometer resolution.
Is your problem a scientific problem to be
solved once, a technical problem to be
solved repeatedly or a business problem to
be solved cheapest in numbers.

Rene

 

Wow.

Theodolites implies angle - more of less.

A rate gyro, on a chip?

 

Joerg can make it cheaper


martin

 

martin griffith a écrit :

But only Philips and Infineon make that dreadfully required component.

 

Rene, I want to measure the angle position of a 4~6 cm radius disc.
Does it help?
Thanks

Rene Tschaggelar ha escrito:

 

Avago (was Agilent (was HP)) make some quite nice absolute and incremental
encoders (14 bit). ( If you have real money there are some laser based
encoders that work down to a few arc seconds, and also some rotary
transformer based solutions. )

You need to specify:-

Absolute or Incremental ?

Accuracy?

Angular speed and acceleration?

Do you require 360 degree revolution ?

How much do expect to spend ?

Dave

 

Quartz crystal cut angles are measured to seconds of arc to get optimum
temperature coefficient. Xray defraction and rugged mechanical
construction are key elements.
For a cheap system two a radial gratings and count the fringes as the
transparent upper grating is rotated over a lower grating could do it.

 

A little calculation shows that a pointer/scale one-arcsecond system on
the disk requires alignment
precision (and also axis-of-rotation position) of 0.14 microns. That's
not possible
with simple light-readout systems.

For reference, a CD squeezes about 6170 bits into one degree of disk
rotation,
with good diffraction-limited optics and lotsa tricks (like
servo-controlled tracking
mechanisms). That's 1.7 bits per arcsecond of rotation.

One can just about use lightbeam interferometry to measure a triangle's
sides to this
precision, but it's not gonna be a simple measurement.

 

Hi Dave,

My needings are: incremental, accuracy half second or second at most,
speed 1.5 revs/sec at most, yes 360 degree, $ 500 to 700 approx
Thanks

Dave ha escrito:

 

Add a 'k' to those values and you just might do it.

Do you really mean seconds ? To put what you've said in context, that's
less than the width of a motorway measured from the centre of the earth.

- Steve

 

Another way: you're asking for 1 part in 1296000 _accuracy_. Even that
kind of _resolution_ is not easy.

John Perry

 

Apparently you have never worked with any high resolution angle measurement
systems. I can do that using 1950's technology, all gears and some tubes
as needed. Back in the 1970 i saw 30 bit per revolution systems, i expect
that better can be readily achieved today.

 

A rotary transformer based solution is the Inductosyn (Farrand Controls
http://www.ruhle.com/). It's typically 360 speed (cycles per rev).

To encode the Inductosyn position, you can get an Inductosyn to digital
converter or design your own.

In a former life we did our own encoder. Using a 16 bit ADC, we encoded
position to 18 bits per degree, or about 0.01 arc-seconds. Accuracy
however was not better than the Inductosyn, which was about 0.5 - 1
arc-second for IIRC a 12 inch diameter model. Achieving this level of
accuracy required the use of patented calibration techniques.

Marc