NavList:

Hi Bill

I enjoyed your article at your Sextant Handbook website.  Each and every inspection certificate I have ever seen details the error at whole degrees, with rather large separations

between them.  As such your treatment of backlash and errors to be found within a turn of the micrometer is well founded.  I do have extensive background in precision robotic machinery,

and as such, fully expect backlash errors to be present.  Backlash happens.  It is the permissible degree that separates the good from the bad.  Worm gears aren’t perfect.

It is the degree of precision in creating the worm that separates the good from the bad.  Your website does explore these relevant features.  While we chase these last bits

of sextant accuracy, the errors in the sextant become a signal, rising from the noise.

It was for this reason that I have explicitly avoided micrometer sextants.  The vernier type sextant may be a touch harder to read (at least at first), but readily becomes second

nature.  All of the backlash inherent in the micrometer is eliminated, since there is no micrometer.  The position of the worm is irrelevant, once the observation is complete,

the vernier is set to the arc, without further reference to the gear whatsoever.

There should be some play remaining in the alignment of the vernier to the arc in vernier-type sextants, as a function of thread engagement.  For those vernier-type sextants

which have a spring aligned along the arc, the vernier-arc play will be limited by the opposing forces of the compression spring and the stiction and friction inherent in the instrument.

A freely moving index arm should easily be moved by the spring, resulting in fairly tight position of the vernier to the arc.   With this type of instrument, a remaining error

will be the centering error in the placement of the index arm, something that may permit you to separate it out.

I wonder if your investigations will cause you to wander down this path in the future?  I would be most interested in a head to head comparison of the spring loaded vernier

mechanism to the spring loaded micrometer mechanism.   Is my avoidance of the micrometer justified?

In the early days of micrometer sextants, was there ever a sextant created which had both a vernier AND a micrometer?  I would expect that, at a minimum, prototypes were created

to convince the remainder of the engineering staff that the micrometer actually functioned as designed, permitting the vernier to be eliminated in production.

Best Regards

Brad

Lately there has been much written about accurate determination of index error, some of it relating to the instruments used and some relating to the observers’ physiology. Recently, Richard Pisko mentioned the subject of backlash, in the context of surveying instruments and I see there was quite extensive (though not always well-informed) discussion of this on the list in 2005. Not for the time being having any new old sextants to pull apart, I have devoted a little time to considering aspects of accuracy of a few micrometer sextants of various ages.

All are second hand but in good condition and have been completely overhauled. I first made a start on the subject in my blog post of 6 July 09 at www.sextantbook.com. under the head of "Interesting Overhaul Problems" and I have followed it up with another account, posted today, under the head of "Chasing tenths of an arcminute".

The account is rather lengthy and is illustrated, so I have posted it on my web site rather than here,for those interested in the details, but briefly my findings are as follows.

In sextants that follow C Plath’s original micrometer design, backlash is negligible, under 3 seconds. This design, which uses a leaf spring to oppose axial movment of the worm shaft by pressing a collar on the shaft against a thrust surface on the bearing, was followed by Tamaya, Hughes and Son after 1939, USNavy MkII, Astra, SNO-M, Cassens and Plath and, of course, C Plath.

In sextants that depended on close manufacturing or hand fitting of the worm shaft bearing to reduce backlash, such as the Freiberger Trommelsextant, SNO-T and pre-1939 Hughes and Son, backlash was a respectable 6 seconds, but by further attention to detail could be reduced to close to that of the first group. If known to be in good condition, it probably doesn't matter from which direction you approach a reading with this type and the Plath type.

The third group, that mounted the worm shaft and swing arm between adjustable coned centres as bearings, includes Heath and Co and later Kelvin Hughes sextants. In my example, I could not by any means adjust out the backlash of about 75 seconds, but it was at least fairly consistently of this amount. I remain puzzled as to its source. My findings might caution you always to approach a reading from the same direction with this type of micrometer design.

The bearing between the swing arm and index arm, and the index arm bearing itself are in principle other sources of backlash. If well made in the first place and properly adjusted, wear here should be negligible, and backlash from this source is unlikely to be important.

As I have mentioned in a previous post, calibration certificates tell us only about instrumental accuracy at various whole degrees and nothing about how accurately the micrometer divides these degrees. In my blog post for 6 July 2009 (“A Worm Turns”) I gave some results and in the latest post I give some more. Many people, I am sure, will be surprised at the size of some of the errors. Bear in mind that the worst are those that deviate most from a horizontal straight line as, over a full rotation, there are always exactly 360 degrees in a circle and the errors have to sum to zero (in practice there is nearly always a deficit of plus or minus a handful of seconds).

On this basis, the SNO-T, a 1938 Husun with a new worm, a 1938 Husun with the original worm, an Ajax Engineering US Navy MkII and a Freiberger all perform well, but before deciding that the Tamaya, Heath and SNO-M don’t cut the mustard, consider that they were second-hand instruments and that 10 seconds at the periphery of a radius of 150 mm represents a movement of about 4 thousandths of a millimetre. Invisible particles of dust and fibres and minute nicks can thus easily lead to variations of this amount.

Bill Morris
Pukenui
New Zealand
----------------------------------------------------------------
NavList message boards and member settings: www.fer3.com/NavList
Members may optionally receive posts by email.
To cancel email delivery, send a message to NoMail[at]fer3.com
----------------------------------------------------------------