Jan 21 2002
Wesley A. Traub 

SD1/2 Progress Report
---------------------

All:

Good news!  Both SD1 and SD2 are working now.  Brief details follow.

Here is the bottom line: I think we had 2 interacting problems, and their
interaction is what caused the nightmares we have been having over this
for a long time now. 

First is that the large air pad on sd2 has no damping stubs on its middle
air orifice (ie, number 2 out of 3 in a row).  This is because the
screw plugs are Loctited in so strongly that I twisted 2 Allen keys 360
degrees and broke off a third trying to remove the plugs.  This also
distorted the flatness beyond the original 1-wave maximum, to about 2
waves, still OK.  The lack of a damping stub means that the pad is prone
to self-induced vibrations.

Second is that the newest dihedral (d2) tends to vibrate itself.  This
dihedral is mounted slightly differently than its predecessor d1, with a
triangular spreader plate in back, tensioned with a single large spring,
instead of 6 small springs as in d1.  The net result is that when d2 is
placed on sd2, the two bad elements gang up and invariably the coupled
system vibrates. I can stop the vibration cold by simply lifting up one of
the three feet supporting the dihedral, with the tip of a screwdriver, to
prevent any suspicion of damping by my finger.  The effect is crystal
clear, when you know where to look.

Knowing this, and not wanting to spend 2-3 days remounting d2 (the RTV
takes that long to cure), I exchanged the dihedrals, to demonstrate that
d1 would not vibrate on sd2.  Luckily this turned out to be true.
Moreover, putting the bad d2 on the good sd1 did not vibrate either.  
Thus separating the bad couple and giving each a good partner worked well.
(Maybe I should take up marriage counseling.)

Both sd1 and sd2 now have their full magnet complement, and the magnet
gaps are set to be about 0.035-0.050 inch.  There is good clearance on the
side air pads, and goodclearance on the motor coil between the motor
magnets.  I now have detailed measurements of all these tolerance ranges,
and how to achieve them.  Three types of shims are needed on sd2 to get to
this state: a shim on the motor carrier plate, another on the top
pull-down magnets, and a third on the plate that attaches to the motor
coil itself.  Only by decoupling all these elements can the mechanical
clearances be achieved, at least on sd2.  The last shim is not used on
sd1, I believe.   

I tuned sd2/d1 to get rid of bad scraping spots along the track, and
completely ignored the detailed stub pressures (contrary to advice from
Anorad and Specialty Products).  I ran both carts in vacuum at three
places, -1 cm, -100, and -190.  I did a complete range of pressures at
each station, and found the region where drag was low (ie motor current
was low), and vibration was small (here under 50 mvpp, average 20 mvpp or
0.08 micron).  A good solution for all pressures in vacuum is 60 psi, on
all 4 regulators.  In air, 70 psi all around is fine, and 80 a bit better.

I will look into getting a spare large air pad, so we can improve the
range of operation of sd2.  I do not plan to remount d2 unless something
new comes up.  At present both lines are operational.  I would very much
like to see a display on our monitor of the 4 crucial parameters:
peak-to-peak position error and motor current for sd1/d2 and sd2/d1; right
now some of these are measure with little clips going to a voltmeter,
adequate for testing but not for operations.