Re: [ATM] Spherometer calibration, precision, etc.

[Richard <cnc@cncservo.co.uk>]

Hi Dominic-Luc,

Saturday, February 24, 2007, 5:30:41 PM, you wrote:

DLW> Diameter: 150 mm
DLW> ROC_design: 1000 mm  (sagitta: 2.816466 mm)
DLW> ROC_actual: 1010 mm  (sagitta: 2.788503 mm)

DLW> above (design vs what is actually obtained) differ by only 0.027963
DLW> mm (0.00110"). I am pretty sure I could measure this with my pinpoint
DLW> light and paper test, but I am curious how well you are actually
DLW> measuring this with a spherometer,

With a 1 micron gauge in my 80mm diameter disk, reading to ... lets say
+/- 2 microns, gives an error of +/- 2.5mm on the 1000mm ROC. Going to
a bigger disk if available, say 120mm, and improving reading accuracy
to 1.5um gives ROC accuracy of +/-0.825mm. Absolute best case of using
the whole 150mm (whilst still spherical :)) and approaching +/- 1um
accuracy by null methods would give an ROC error of +/-0.37mm

DLW> Actually, along similar lines, I never got round to purchasing one
DLW> of these high precision guages for a spherometer (best I have now is
DLW> 1.968 turn/mm screw). Maybe someone can remind me of the source,
DLW> and I'll just go ahead and order it.

Dominic, using a screw thread driven spherometer is a big source of
inaccuracy as you are having to constantly judge yourself the contact
force. If you hollow out a thick disk and chamfer the outside such
that the edge is a knife point, and in the same operation on the lathe
bore, or drill and ream, an accurate central hole to mount a DTI, then
you have a good accurate spherometer. The plunging DTI has the
advatage of being spring loaded onto the surafce with almost constant
force so can be very accurate. For the highest accuracy, use one of
Mitutoyo's 0.001mm resolution digital indicators (Ebay...) - there is
a picture of one on Mike Lockwoods webpage. You can set the gauge in
position and then zero it wherever you like. You don't have to zero it
on a flat, you can pile up a stack of Jo Blocks (Gauge blocks) to the
required sag and then zero on that so you now have a null testing
spherometer.

The circular knife edge and the "three balls" methods have their
relative advantages and disadvantages. You have to be careful with the
knife edge, both not to damge it, you, or the mirror under test, but
one distinct advantage is the ease with which you can very accurately
measure one imporatnat parameter of the instrument - in this case, the
inside diameter for which you use a three legged internal bore
micrometer. It's much harder to measure the three balls type,
largely because if they are home made they will often not sit at the
apexes of an equilateral triangle.

-- 

Best regards,
  Richard in the UK

_______________________________________________
ATM mailing list http://www.atmlist.net/