NavList:

Don

There are plenty of encoders available, with very high accuracy.  One only need mount it concentric to the index arm axis of rotation.  Directly encode the motion of the index arm and eliminate all gearing, similar to a vernier sextant, only with a direct readout of D.M.S.

Brad

On Jul 23, 2017 2:25 PM, "Don Seltzer" <NoReply_Seltzer@fer3.com> wrote:

On Sat, Jul 22, 2017 at 9:33 PM, Pete Solon Palmer <NoReply_Palmer@fer3.com> wrote:

Just to throw up ideas for everyone to knock down:  The rotary encoder seems to be rated by resolution per turn.  In the sextant we only need 1/4 turn so maybe we could gear it up 1:4 to increase the rez?   Cons would include the complication of gears and gear lash.  Linear encoder is lookin' good.  Maybe have a low rez solution for degrees and a higher rez for the minutes?

I suspect that gearing would create both accuracy and backlash problems.

There is a type of rotary sensor that can be electronically 'geared'.  Resolvers are a magnetic sensor, sort of a rotary transformer.  Very rugged and no backlash.  A simple resolver generates a sine and cosine signal over a single revolution.  The ratio of the two can be used to determine the angle, typically to one part in 4000. There are also multispeed resolvers with multiple pole windings that generate a few sine wave cycles in one revolution.  With a 6X resolver, one could resolve a full revolution to one part in 24000, good enough for one arcminute.

It is has been decades since I worked with them, and I have no idea how available or expensive this technology is today.

Don Seltzer