On Sep 10, 2009, at 9:46 AM, Gibson Elliot wrote:
I would think that running the device with the axis tilted 45 deg.
off center to the left, and then to the 45 deg. off center to the
right, would allow for normalizing out the effects of mechanical
shifting or warping.
Not true. Any change in the angle with respect to vertical changes
the stress on the beam splitter support and the beam splitter itself.
This is a simple problem to resolve.
Appearances can be deceiving.
I am planning on building a small, very small version of the device
with a laser diode, etc... in a much smaller form factor on a
single substrate. This should minimize any mechanical issues.
If you'll note, the mirrors and splitter were mounted on a single
metal "substrate", or base. The base is not the problem - it is the
deflection of the optical components under stress which they were not
designed to handle. Both the components and their mounts require a
major structural beefing up. Optical test benches and components are
typically designed for stable horizontal use.
Using a CCD pickup and computer to analyze the motion in much finer
detail should also help.
Analyze what motion? There is no need to precisely record the
rotation at this point, and the interferometry does a much finer job
of measuring relative motion of the parts than a CCD pickup.
Anyone spare some parts? Small high grade mirrors etc.. to set this
up?
Gibson
Best regards,
Horace Heffner
http://www.mtaonline.net/~hheffner/
