NavList:

   

Reply

Pierre:

As a quick answer, I have always been taught that easterly and westerly
deviations should sum to zero (or, more correctly, if the axis of a
plotted deviation curve should be at zero degrees).  In fact, if they
don't then this is a signal the compass's lubber line is not aligned
with the keel of the boat.

The Handbook of Compass Adjusting (one of those wonderful, albeit hoary
publications from the US DHMA (or NIMA, as they call themselves these
days), http://pollux.nss.nima.mil/index/uncon/products/HoMCA.pdf,
paragraph 318, gives a list of the types of errors one might see in a
deviation curve.  If it were possible to reduce the deviation curve (eg,
as plotted on a Napier Diagram) to a straight line and the straight line
were off-axis (ie, giving a constant deviation regardless of the boat's
heading) this would be called an "A" coefficient.  The possible causes
for "A" coefficients are given as "misaligned compass," "incorrect
observations" and "asymmetrical alignments of horizontal soft iron (rare)"

Compass adjusting on steel vessels is an art unto itself, with Flinders
bars, quadrantal compensators (the steel balls one sees in ships
binnacles), "soft irons," etc.  Fiberglass or wooden boats pleasure
boats have modern small-boat compasses that don't have "soft irons," so
assuming you're talking about the world of pleasure boats, yes, easterly
and westerly deviations should cancel.

Lu Abel



Pierre Boucher wrote:
> Is it correct to say that if a compass is properly installed (align,
> compensated, etc..) that the sum on the residual west deviation equals
> the sum of the residual east deviation?
>
> Pierre Boucher
>
> ************************************************************************
> Pierre Boucher N
> formation en navigation de plaisance
> Ste-Therese (Quebec) Canada
>
> la "VOILE" ... le reste n'est que du vent...
> "Sail, Sail !"... the rest is only wind...
> EMAIL: pboucher@lavoile.com
> http://www.lavoile.com
> ************************************************************************
>
>

   

Reply