NavList:

   

Reply

Fred wrote: "One would expect
that some smoothly varying curve passing through the points would
describe the errors, so that one could interpolate between points
to
estimate an error.  Unfortunately, this may not necessarily be so. "

The varying patterns of errors seen that Fred refers to is likely due
to complex interactions. Older sextants had the racks cut by a
dividing process with one tooth(more accurately, the space between two
teeth) being cut at a time, with the sextant rotated between
successive teeth on a dividing engine. The sextant's errors would be a
reflection of the dividing engine's errors, which in turn would
reflect errors in the lathe that cut its tangent screw(in turn
reflecting errors int the gears linking the lathe spindle to its lead
screw, in the leadscrew itself, etc.)eccentricity errors of the
engine's spindle and of its tangent screw, axial slip in the tangent
screw and so on. Differential heating of the engine or of the sextant
during cutting of the teeth would add to the confusion.

Members may be interested to know that the first dividing engine with
any pretence to high accuracy was made and described by Jesse Ramsden
in 1777. A number of exact or very close copies were made by Troughton
and other instrument makers and one of them was borrowed from the
Science Museum in London in June 1942 by D Shackman and Sons to
produce "upwards of 2000 sextants" for the US Navy.  Shackman's was
still active in February 1990, making cameras rather than sextants,
and the then Chairman wrote " Mrs Eastwell....recalls visiting the NPL
with her father(Mr Reuben Shackman) when he used to take the sextants
for inspection." After numerous(and increasingly anxious)requests, the
engine was returned to the museum in November 1951. It is not clear
whether Shackman's were making Mk II s. Certainly there is at least
one sextant they made(in the Smithsonian collection) that closely
resembles the Hughes and Son "Mate" sextant, made for Kelvin and
Wifred White in 1942.

I do not know how the rack of modern sextants is cut. A hobbing
process is inherently more accurate, but this is not the place to go
into details.

Frank wrote: "Since I have a Tamaya clone, can you elaborate on this?
I'm not clear which
direction you're talking about. "

In a C Plath/Tamaya and clones, a "thrust pre-load spring" bears on a
bronze insert in the end of the micrometer screw and keeps the right
side of a thrust collar in contact with a thrust block. The sight
should be made in such a way that this load on the thrust block is
always increased, not decreased. A moment's thought will show that
doing otherwise will tend to take the right side of the thrust collar
away from its bearing surface. Then, when turning of the micrometer
screw stops, the spring will return the thrust collar to its rightful
position, but with uncertain force, dependent on variable friction
from the parallel portions of the micrometer bearings and from the
index arm bearing. Turning one way gives a positive final setting.
Turning the other gives an uncertain final setting. On sextants like
the SNO-T and Frieberger, thrust can be taken at both ends of the
screw, so that potential backlash should be less than 12 seconds. On
the Mk II that I own, the backlash is variable and can amount to a
couple of minutes.

Members who would like a photo of my calibration setup and a labelled
photo of the micrometer screw thrust arrangements may e-mail me off-
group.

Bill
--~--~---------~--~----~------------~-------~--~----~
To post to this group, send email to NavList@fer3.com
To , send email to NavList-@fer3.com
-~----------~----~----~----~------~----~------~--~---

   

Reply