|
|
|
|
|||||
|
||||||
From: George Huxtable (no email)
Date: Mon May 08 2006 - 13:39:19 EDT
GregR wrote-
"Fred: I hear you on being careful not to
"over-adjust" it for errors, but I did need
to get a good initial calibration (something that
I couldn't seem to do in the backyard),
and from what I understand plastic sextants are a
lot more susceptible to IE than
the metal ones. In fact, the Davis instruction
manual seems to recommend adjusting
it whenever it's used ("Adjusting your sextant is
easy and should be done each time
it is used."). Maybe that's why they put knurled
adjustment screws on it instead of
regular screws?"
There are a number of things wrong with what GregR
says here.
First, in no way does such a zero adjustment get
any sort of "calibration" of the sextant, which
would require a check, by some means, of the arc
divisions right around the scale, not just at the
zero point. In general, one normally has to accept
that the initial calibration, of the markings
around the scale, is approximately correct.
Certain sextants will have had that calibration
checked by some authority, in which case there
will be a table of known errors at different
angles tacked to the inside of the box. It is
possible, but rather difficult, for a user to make
his own calibration checks by star-to-star angles,
as Alex Eremenko, in particular, has written about
in the list's archives. All such calibrations are
made on the understanding that any index offset
has been allowed-for first. You will therefore
never find any offset noted for zero degrees, in a
calibration certificate.
Second, obtaining zero index offset, to high
accuracy, by adjusting the fiddly mirror screws is
a difficult task, nearly an impossible one. When
you have made your best attempt, what do you do
next? You check it out, by viewing a distant body,
to see what (exactly) the scale reads when the
images align. What if they don't exactly align at
zero on the scale, but at a small offset of the
order of a minute or so? Do you go back and do
another round of mirror adjustments? No, you just
accept what you get, and allow for that initial
offset by a simple bit of arithmetic. In fact, no
matter how big the zero-offset is, it does not
affect the result in any way, as long as it's
allowed for. What's wrong here is the mind-set, in
thinking of such an offset as a sextant "error",
which has to be minimised or zeroed-out if the
sextant is to do its job properly. Nothing of the
sort.
Third, it's true that plastics expand with heat
more than metals, and this, depending on the
design details may make the zero offset less
stable. I have not found that to affect the
performance of my own plastic sextant (Ebbco), in
practice. I always check the zero before a set of
observations, and after, and sometimes
interspersed between observations, but never find
a significant change. For other plastic sextants,
I have seen accounts of significant short-term
intability, and wonder if that may be related to
looseness somewhere in the mirror mountings. I am
careful, when I am using the sextant, not to put
it down where a shadow-edge falls across the
mirror mountings, to avoid thermal gradients in a
sensitive spot, but am no more careful than that.
Fourth, if a user keeps readjusting the zero
offset by tweaking the screws, he has lost the
opportunity to monitor the stability of his
sextant by noting such changes.
If he records it, and allows for it, never
adjusting it, he will get a feel for its long-term
reliability.
Fifth, what's been missed is how quick-and-easy
such an index check is. Simply the work of a
moment, to point to the horizon, or Sun or star,
align the images, note the reading. Much less
time-consuming than the business of tweak, then
look, then delicate tweak again, that is involved
in trying to zero it out. It's easy to check it
before and after, and between, altitude
observations.
A trick of the old-salt navigators, once they had
adjusted any tilt out of their mirrors, and got
the zero roughly right, was to introduce a spot of
salt-water, or urine, to initiate a bit of
corrosion of the adjustments, so that they would
never be touched again.
GregR is new to the sextant game, but he will
learn as he goes.
=================
He added-
"BTW, don't know if this has been done before, but
I came up with a slightly
off-the-wall method of "faking" a horizon since
the sun's dec is now too high
to get a LAN shot with an artificial horizon at my
latitude (34°14.9' N):
I took a length of surveyor's string and attached
one end to the side of the
house at my eye height, with the other end
attached to a tripod also set at
my eye height (and checked for horizontal with a
carpenter's line level).
From across the backyard (~20') I'm able to get
"reasonable" LOPs - the
intercepts on those are running anywhere from 3.7
to 8 miles (though I do
get the occasional one that's way out of the
ballpark, so this method isn't
perfect. In fact, I can induce a several-minute
error by slouching slightly
vs. standing up straight).
Even if it's not accurate enough for real
navigation, it did serve its purpose
in giving me something to practice bringing sights
down with (and it's long
enough to be able to rock the sextant to find true
vertical). Now that I've got
that part down, time to work on improving the
accuracy with a real horizon. :-)"
===========
Fair enough, as long as GregR is aware of the
inaccuracies involved. Bending his knees to change
his height by one inch will change the angle by
about a quarter-degree, which would be enough to
shift his position by about 15 miles
-- GregR
|