Re: [ATM] Focal Length Accuracy

[Ken Hunter <atm_ken_hunter@yahoo.com>]

--- Dominic-Luc Webb <dlwebb@canit.se> wrote:

> 
> On a similar note, I recently bought a ring type spherometer
> from Mike Hunter and decided to calibrate it. I am not sure
> how other people do this. It was machined to specific
> tolerances, but calculating back from the "known" ROC (from
> optical tests) did not yield the expected ring diameter. It
> came down (I think) to whether the ring was landing on the outer
> or inner edge, etc. 

Dominic...

The spreadsheet I sent you has 2 pages. One is for spherometers where there are balls placed
aroound the edge of the spherometer and the formula for that spreadsheet takes into consideration
the diameter of the balls. The second page of the spreadsheet is for Ring type spherometers and
the default entries were made for your spherometer. This page does not take into consideration the
very small flat part of the ring into the formula. If you require more accuracy, we can do away
with the tiny flat area which subjects the "ring" to much more stringent handling care being
required to protect the "knife edge" of the ring or... You can input the flat size into the other
page as the "ball diameter (with a bit of tweeking since a flat is not a ball) to improve the
expected readout of the spreadsheet program.

> I ended up using an "effective" ring
> diameter, calculating back from a couple of mirrors with "known"
> ROCs. This brings me back to our earlier discussions about the
> accuracy of a spherometer, now that I actually have one in my
> hands with a decent guage.
> 

It is remotely possible that I have an error in the formula though I did run it through a few
friends that are actually good at math (I am not). How much error in spherometer diameter are you
talking about? I did the best I could - I measured the diameter with a micrometer, marked it with
chalk and cut it out with an axe...

;0)

> How does one calibrate a spherometer such that you would
> put more weight on the readings from this than a Foucault or
> other optical test where you measure the point where light
> is actually focusing?
> 
> Dominic
> 

Hopefully, the accurate measurement of the spherometer diameter and the sagitta and the input of
the data should result in a very close value of focal length. I don't trust most readouts that
give measurements to the N-th decimal place and have posted before on that subject (6.005 inch
F6.08345 mirror for instance) however the Excel spreadsheet formats the rows and columns into nice
even displays which have possible misleading accuracy results due to the added numbers. I did try
to limit the number of decimal places to one more than actually necessary for an accurate reading.

How can you really tell if the focal length is 48.0000 or 47.9995 inches? That's a tough one to
answer and the actual physical metrology is tougher yet. What I provided as a spherometer is
accurate enough for most ATM efforts during grinding, actually being able to certify that
measurement to the limits of the spreadsheet display is another thing.

You'll notice that I also over sampled the spreadsheet by making the measurement intervals very
small. This was done to be able to handle different units of measurement. For instance. If your
readout device is capable of only .001 inch resolution, having a spreadsheet that increments in
microns is kinda overkill BUT... if you DO have a readout than shows microns, the spreadsheet will
work fairly well. 

I compromised and made measurement intervals that should be able to accomodate everyone that wants
the utility of the spreadsheet even though printing it out might take a few more pages of paper
than necessary though you can just print out the pages you are interested in. It's hard to please
everybody.

Let me know if I can be of more help.

Ken Hunter

Ken Hunter




       
____________________________________________________________________________________
Moody friends. Drama queens. Your life? Nope! - their life, your story. Play Sims Stories at Yahoo! Games.
http://sims.yahoo.com/  
_______________________________________________
ATM mailing list http://www.atmlist.net/