In a message dated 06/04/03, dyouden@......... writes:

Hi Dave,

       Thanks for the reply.

> I guess complexity depends upon personal viewpoint to some degree. What you 
> have described seems complex to me. I have built systems similar to what 
> you describe and the complexity springs from having to prevent rotation.

       ? It is two flat sheets of glass with 4 pairs of small magnets and two 
strips of metal mounted horizontally on top of the top sheet --- it's not too 
complicated! I would expect it to be rather easier to prevent rotation than 
to do the linear positioning. 

> My stack of tubes originates from the challenge of making a relatively 
> accurate prismatic shape without resorting to machine tools, which most of 
> us don't have. Commercial tubing is pretty repeatable stuff, although not 
> especially accurate.

       I have very considerable difficulty in visualising the apparatus that 
you are trying to describe. Is there any chance of you putting a hand drawn 
sketch on the Web, please? In my experience, commercial tubing is not usually 
very accurate and very often it is not quite circular. Hard drawn seamless 
tube is best. If you just want a lot of mass, why not use solid bar? I am 
trying to figure out how you maintain the alignment with four pads...

> I have built a commercial system, with a moving mass of about 10 Kg. (It 
> cost about $800,000) with 0.3 nanometer laser position feedback. I have 
> never thought of looking for seismic effects on this device, but I'll bet 
> that they are there. That particular device is used for metrology, and has 
> an overall position accuracy of less than 10 nanometers.

       I am very curious to know why you used an optical measurement system? 
Getting down below about 10 nm is hard work optically. What frequency range 
was this over?
       Resolving 0.1 nano metre is not difficult or expensive with a linear 
capacitative detector. However, when you get down to these dimensions, 
thermal expansion effects are dominant and you can only stabilise the centre 
of position of the detector. Even fused silica has an expansion coefficient 
of 0.2 ppm / C Deg     

> The noise from air bearings is low enough so that I have been able to 
> machine 10 Angstrom RMS surface finishes in aluminum on these bearings.

     Great. My experience with air bearings has not been quite so good. What 
precautions do you take with the air supply? What pressure, filters, flow 
restrictors and pressure regulators do you use / propose for this? 

       Regards,

       Chris Chapman

In a message dated 06/04/=
03, dyouden@......... writes:



Hi Dave,



       Thanks for the reply.



I guess complexity depends=20=
upon personal viewpoint to some degree. What you have described seems comple=
x to me. I have built systems similar to what you describe and the complexit=
y springs from having to prevent rotation.




       ? It is two flat sheets=
 of glass with 4 pairs of small magnets and two strips of metal mounted hori=
zontally on top of the top sheet --- it's not too complicated! I would expec=
t it to be rather easier to prevent rotation than to do the linear positioni=
ng.=20



My stack of tub=
es originates from the challenge of making a relatively accurate prismatic s=
hape without resorting to machine tools, which most of us don't have. Commer=
cial tubing is pretty repeatable stuff, although not especially accurate.




       I have very considerabl=
e difficulty in visualising the apparatus that you are trying to describe. I=
s there any chance of you putting a hand drawn sketch on the Web, please? In=
 my experience, commercial tubing is not usually very accurate and very ofte=
n it is not quite circular. Hard drawn seamless tube is best. If you just wa=
nt a lot of mass, why not use solid bar? I am trying to figure out how you m=
aintain the alignment with four pads...



I have built a=20=
commercial system, with a moving mass of about 10 Kg. (It cost about $800,00=
0) with 0.3 nanometer laser position feedback. I have never thought of looki=
ng for seismic effects on this device, but I'll bet that they are there. Tha=
t particular device is used for metrology, and has an overall position accur=
acy of less than 10 nanometers.




       I am very curious to kn=
ow why you used an optical measurement system? Getting down below about 10 n=
m is hard work optically. What frequency range was this over?

       Resolving 0.1 nano metre is not dif=
ficult or expensive with a linear capacitative detector. However, when you g=
et down to these dimensions, thermal expansion effects are dominant and you=20=
can only stabilise the centre of position of the detector. Even fused silica=
 has an expansion coefficient of 0.2 ppm / C Deg     



The noise from=20=
air bearings is low enough so that I have been able to machine 10 Angstrom R=
MS surface finishes in aluminum on these bearings.




     Great. My experience with air bearings has not=20=
been quite so good. What precautions do you take with the air supply? What p=
ressure, filters, flow restrictors and pressure regulators do you use / prop=
ose for this?=20



       Regards,



       Chris Chapman