Re: my $0.02 and laser interferometers

Elliot Burke wrote:>

> >
> > For those interested in using interferometers to measure distance, a
> > suitable laser source costs much less than $100. The optics required
> aren't
> > too complex. Basic precision can be 0.5 ppm, if the temperature and
> > humidity of the air path are compensated for.
>
> I tried to breadboard something like this many years ago. I couldn't get
> any fringes. One of the problems is I couldn't afford the 1/4 wave
> retardation plate, to convert plane to circular polarization. Diode lasers
> should do fine, at $9 for a 5 mW output at 660 nM.
>
> Jon
>
> Something must have been very wrong, it's hard NOT to get fringes with a
> coherent source.
> There are many interferometers that don't need quarter wave plates (QWP), in
> fact, most don't. Look in any university library, Steel's book on
> interferometry is a good one, also Malacara's Optical Shop Testing.

Well, getting fringes is one thing, but getting fringes that MOVE linearly in a
direction related to the movement of the far mirror requires a specific
form of interferometer. The one I found in a book used circularly polarized
waves throughout the interferometer. When the fixed and movable arms
were recombined in a beam splitter, you got varying amounts of linearly
polarized light, and the two cross polarized outputs were in quadrature!
That looked like the best way to send a signal right into a standard DRO
counter circuit.

> The diode lasers are sometimes useful, in a limited way, for interferometry.
> There are two limitations:
> 1. the wavelength changes by about 0.25 nm/�C, or 0.03%/�C. Not very good,
> unless you are very well stabilized. Also there is no easy way of knowing
> precisely what the wavelength is, since they are all different by 1% or so.
> Wavelength is not tightly controlled.

Right, but I'd rather try to temperature stabilize a tiny diode laser than run
a big gas laser for thousands of hours. As for calibration, you would likely
want to calibrate any interferometer. Once you had the temp set to
some repeatable setting, you should be able to calibrate it once and
forget it.

> 2. the coherence length most most LD's is short. A path difference of more
> than a few mm gives very low fringe contrast. There are some LD's with
> longer coherence length, but it varys with drive current, age, and other
> factors.
> Unless you want to be a laser engineer, stay away from metrology with LD's.

Yup, that is, indeed, a problem, if you have a 2-3 M round trip path on the
measuring arm. I'm pretty sure you can get a lot more than a few mm,
definitely a few cm, but that still falls short. There are some tricks with
external cavities, etc, but that gets out of the range of the $9 laser diodes.

> The HeNe laser, on the other hand, is an inspired choice for metrology.
> NIST has certified its wavelength to be know to 0.5 ppm, so if the index of
> refraction of air is corrected for by temp and humidity measurement, a 1 ppm
> accuracy is easily obtained.
> The coherence length of the HeNe is also very long, many meters, if a single
> or double longitudinal mode laser is chosen.
> I have obtained microinch stability over 60 inch paths over periods of
> hours, on a big surface plate in a temperature controlled room.

Yes, that may be an argument for HeNe. Of course, for machine tools
in even high-end hobby use, 1 ppm is certainly not needed!

Jon