re_ gage blocks & interferometer precision
Posted by
Elliot Burke
on 2001-11-08 04:09:06 UTC
Les Watts writes:
There are two convenient long conherence length lasers:
HeNe of various wavelengths and diode lasers.
The HeNe, if the tube is short enough, will lase at a single frequency.
This frequency is known to be within .5 ppm of the nominal line center,
because the line is very narrow. So over a 1 m range a µm accuracy is
possible. The properties of the laser vary slightly with isotope.
Coherence length is many meters, I never had a bench long enough to measure
it, even with multiple bounces.
The HP interferometer makes measurements to 10's of meters, if I remember
correctly.
By stabilizing the laser is is possible to get 0.1 ppb frequency accuracy,
this is done for some metrology applications. There is a convenient iodine
transition that is used with a vapor cell for this.
There are lots of other interesting lasers, but they put you out of the $100
range.
A much bigger source of error for long paths than wavelength is the
variability of the index of refraction of air. Humidity and pressure,
sometimes temperature are monitored in the beam path. For a long path
several points are measured. Some interferometers operate in a vacuum or in
a helium atmosphere to eliminate this problem.
Diode lasers are convenient, but the frequency is not inherently stable or
known. They can be stabilized to atomic transitions quite well. They have
an additional nice feature in that by modulating the current the frequency
is modulated. This makes heterodyne detection quite easy.
It is possible to make interferometers for $100. But not a precision length
measuring interferometer, I'm afraid. The stabilization hardware is too
complex.
If you just want to look at fringes, get a laser pointer, a piece of
beamsplitter glass from Edmund Scientific, and a couple of first surface
mirrors. Some sort of two axis tip mounts are needed for the mirrors. A
short focal length (10-20 mm) lens is used to expand the beam after it exits
the interferometer and to focus onto the test surface.
I use the laser pointer interferometers in my work quite a bit, they are
mounted on plywood sheets and have attached microscope lenses for examining
flatness in small areas.
The little batteries that come with the pointers wear out fast, I used 3 AA
batteries in series for longer life, though this extra current doesn't do
the laser much good. Measure the current and put a series resistance in as
needed.
On a metal surface with good shiny finish you should be able to see fringes
indicating the surface flatness.
Be sure to keep movements less than .2 µm over the time period of a
measurement.
Elliot Burke
>I guess what we would all like to have is a nice laserA subject near and dear to me.
>interferometer. I had a physics professor I used as a consultant a couple
>years ago who claimed he could throw together one from junk parts for about
>$100. I don't know
>how you would get a long coherence length laser and precise
>temperature control for that price though. Ought to check with
>him again.
There are two convenient long conherence length lasers:
HeNe of various wavelengths and diode lasers.
The HeNe, if the tube is short enough, will lase at a single frequency.
This frequency is known to be within .5 ppm of the nominal line center,
because the line is very narrow. So over a 1 m range a µm accuracy is
possible. The properties of the laser vary slightly with isotope.
Coherence length is many meters, I never had a bench long enough to measure
it, even with multiple bounces.
The HP interferometer makes measurements to 10's of meters, if I remember
correctly.
By stabilizing the laser is is possible to get 0.1 ppb frequency accuracy,
this is done for some metrology applications. There is a convenient iodine
transition that is used with a vapor cell for this.
There are lots of other interesting lasers, but they put you out of the $100
range.
A much bigger source of error for long paths than wavelength is the
variability of the index of refraction of air. Humidity and pressure,
sometimes temperature are monitored in the beam path. For a long path
several points are measured. Some interferometers operate in a vacuum or in
a helium atmosphere to eliminate this problem.
Diode lasers are convenient, but the frequency is not inherently stable or
known. They can be stabilized to atomic transitions quite well. They have
an additional nice feature in that by modulating the current the frequency
is modulated. This makes heterodyne detection quite easy.
It is possible to make interferometers for $100. But not a precision length
measuring interferometer, I'm afraid. The stabilization hardware is too
complex.
If you just want to look at fringes, get a laser pointer, a piece of
beamsplitter glass from Edmund Scientific, and a couple of first surface
mirrors. Some sort of two axis tip mounts are needed for the mirrors. A
short focal length (10-20 mm) lens is used to expand the beam after it exits
the interferometer and to focus onto the test surface.
I use the laser pointer interferometers in my work quite a bit, they are
mounted on plywood sheets and have attached microscope lenses for examining
flatness in small areas.
The little batteries that come with the pointers wear out fast, I used 3 AA
batteries in series for longer life, though this extra current doesn't do
the laser much good. Measure the current and put a series resistance in as
needed.
On a metal surface with good shiny finish you should be able to see fringes
indicating the surface flatness.
Be sure to keep movements less than .2 µm over the time period of a
measurement.
Elliot Burke