Re: [ATM] Understanding Foucault

[mdholm@telerama.com]

Dave Harbour's description of the Foucault test is very good.  The only quibble
I would have is that he shows what is known as the Millies - LaCroix method (or
the "tornado" graph) for analyzing the data.  Careful work by more than one
mathematically inclined atm has shown that the M-L method is weak.  Now that we
have computer programs, they are the way to go.  I am a big fan of Sixtests.  It
uses about the most sophisticated math that can be applied to Foucault results,
and since Jim Burrows knows his math pretty well, we can be reasonably sure it
is correct.

If you have to do it by hand, the best is probably to steel yourself to using
Texerau's analysis.  It's not all that hard with a calculator, and, so far as I
know, gives correct results.

About moving source versus fixed source.  Moving source works well and has some
advantages.  The drawback is that knife edge movements are approximately halved.
 This means that the requirement on precision of measurement (how many
thousanths of an inch toward and away from the mirror) is twice as tight.  Since
you can buy a fair dial indicator for less than $20 from Harbour Freight, this
isn't as big a problem as it once was.  Most ATM's find that their readings are
uncertain to more than the 0.001 to 0.002 errors that may remain in a cheap dial
indicator.  The source of uncertainty is difficulty reading zone nulling and is
independent of the measurement tool.

With moving source testers, it is common to cover only half of the light source,
and to use the same knife edge at the source as for the probe.  Many atm's find
that they can get more sensitive shadow readings by blocking most of the source
with a second knife edge, leaving a slit, as commonly shown in moving source
Foucault testers.  The "main" knife edge still serves both for the source and probe.

Mark Holm
mdholm@telerama.com
_______________________________________________
ATM mailing list http://www.atmlist.net/