Re: Pride Goeth Before a Fall

[Tom Droege <droege@wwa.com>]

Alain,

Thank you for all the good suggestions.  

As you can see from other correspondence I was thinking about that solution
from the start, just got tired of implementing all the details.  My brother
had told me of using leather in precision micrometers.  Yes, I do have a
dial comparator and can use it for test.  My usual approach is to try the
simple solutions first, then fix the things that are not good enough.   I
did leave room for an insert.  I can drill out the brass block and put in
something soft.  First I will try other screws.  

Tom Droege

At 12:31 AM 3/19/99 +0100, you wrote:
>david garnett wrote:
>> 
>> If the nut is made of metal, it will only contact in one place at a time,
>> however long
>> you make it ! In fact, the contact point will probably move around as
>> everything turns
>> leading to all sorts of interesting anomalies !
>> 
>> When they made the original diffraction grating ruling engines they used a
>> soft
>> nut - made of something like cork if I remember correctly - to average out
>> the
>> errors. Mind you, they were starting with a pretty good screw thread
anyway.
>> 
>> Still, the possibility exists that a PTFE 'nut' might improve things with a
>> bit of
>> averaging.
>> 
>> Are you sure that the shaft is straight enough ?
>> 
>> regards
>> Dave
>> 
>
>Exactly what I was going to suggest. To remove the cork screw stars, use
>a cork nut. Indeed this is how the lead screws for diffraction grating
>engines are made. The advantage is that this is done in a constant
>humidity room, whereas the telescope will not be used in such an
>environment. The basic idea is to pinch two pieces of cork around the
>screw. For example 2 pieces of metal sandwiching the 2 pieces of cork.
>The cork (make it long enough, like at least 20 threads long) will
>average out the periodic errors. If your mount is correctly balanced,
>you can use such a fragile nut. Also make sure your screw is straight
>(you can check that either by rotating it on a flat surface, or just
>looking at it  with one tip near you eye). Make sure its extremities are
>held correctly in ball bearings which do not move laterally. I would
>also remove the motor and turn the screw by hand, and see (in fact touch
>more than see :-) ) if it turns smoothly with no hard points. Ideally
>you should be able to put a comparator (I don't know the exact english
>term, a thing which shows the motion down to 1/100mm or whatever
>fraction of inches you can see) in order to see how things move at the
>tip of the optical tube while you rotate the screw.
>You can use other materials instead of cork. Here we have used a plastic
>style thing called turcite, if I remember the name correctly, which has
>about the same properties as cork, except it will not soak up water.
>I don't like the idea of fixing by software things which could and
>should have been fixed by mechanics first. Do it right, then it'll work.
>This type of drive system is very popular around here and works well.
>You can get a very high precision drive for almost nothing compared to
>the price of a high quality worm gear.
>Alain
>
>