Re: ATM 5 hour error-free straight bolt tracking

[The Silent Observer <silent1@ix.netcom.com>]

doug kniffen wrote:
> 

<snip>

> It seemed deceptively simple, a straight bolt fixed
> at 90 degrees to one board and a second board that "rode" the bolt while
> sliding against the first.
> 
> The are no "magic" ratios involved and only one critical dimension. Motor
> rotation isn't important. Even the drive rate doesn't matter.  If the
> distance on the second board , between the bolt pivot and the slide contact
> point, equals 6 hours of thread it should track to the mechanical limits of
> accuracy. After sketching this thing out it appears that one could easily
> get five hours of tracking in the basic configuration. A second
> configuration sketched might allow up to eleven hours, though it could be
> tough to eliminate mechanical slop as the meridian is crossed.

<snip>

> 
> Maybe I'm all wet on this, but I'm going to build one since it looks good
> on paper. Anyone have comments??

I think I've correctly visualized this.


                          /
                         /
                        /
                       /| <- bolt at fixed point on this board
                      / |
                     /  |
                    /   |
                   /    |
slide here -----V /     |
_________________/______|___
                      |___| <-- clock motor

The sliding joint is aimed at the celestial pole, and whatever is on the
upper board rotates at sidereal rate (assuming your clock rate and screw
drive match up properly with the dimention from the bolt pivot to the
sliding edge).

I'm curious what can be used to keep the bolt precisely localized on the
second board -- a pointed bolt riding in a drilled dimple would seem to
have a lot of slop, a ball joint would have too much friction and
overload the clock motor.  You might be able to mount the motor on the
upper board with a gimbal hinge (hinge line through the intersection of
the bolt axis and the board surface) and you could probably put a knife
edge on the sliding board, which would ride on a metal plate, in order
to minimize inaccuracies due to wear.

Is the requirement for the distance from bolt pivot to sliding edge to
be 6 hours worth of threads the critical dimension?

And how do you visualize flopping over the meridian for a longer term
drive?  Some sort of very weak spring plus reversing the motor just as
the board crosses vertical (could probably be done with microswitches,
if you have a reversible clock motor), or something more elegant? 
Perhaps make the sliding board a right angle, slide on the bend for the
first six hours, and put a collar on the bolt that will pick up the
sliding edge as it comes even with the bolt, then lift by that collar
for the next six hours (don't see how to make the transition smooth, but
I can't say it's impossible)?

-- 
Some of their knowledge is corrupt, and inaccurate, being gleaned from
all manner of sources.  But like us, they are wizards too.
                                                    -- Jaldis the Blind

Donald Qualls, aka The Silent Observer           NAR # 70141-SR Insured
Rocket Pages                http://silent1.home.netcom.com/launches.htm
Telescope Pages            http://silent1.home.netcom.com/astronomy.htm

Opinions expressed are my own -- take them for what they're worth
and don't expect them to be perfect.