In a message dated 09/04/2009, rog@.......... writes:

Is a  scalpel blade against a steel ball bearing the current best consensus 
for a  Lehman pivot?
rob



Hi Rob,
 
    Crossed cylinders can be a bit better.  The surfaces need to highly 
polished. The SEP uses 1/8" OD polished Tungsten  Carbide needle rollers, but 
these may be rather expensive / difficult to obtain  depending on where you 
live. See _http://www.bgs.ac.uk/education/school_seismology/seismometer.html_ 
(http://www.bgs.ac.uk/education/school_seismology/seismometer.html)  You  
mount both the vertical rollers on the frame. 
 
    However, a polished SS plane rolling on a SS ball  bearing works 
perfectly well and it may cost only 50 c! See _http://www.smallparts.com/_ 
(http://www.smallparts.com/) 
 
    I use a 1/2" SS bearing and a large scalpel blade  'trimmed' oval with 
a carbide disk and glued to the arm with two part  acrylic adhesive. You can 
use bearings down to 1/4" OK, but mount them on the  vertical frame, NOT on 
the arm! My top suspension is an 8 thou music wire. I can  set the period 
to 60 seconds OK and it works for some days, but it has  a significant tilt 
drift. My 'normal' set period is 20 seconds. I  measured +/-2mm drift over a 
fortnight with this, in a max drift range of  +/-10mm.
 
    The two 'suspensions' which are  very likey to give trouble are the 
knife edge on a plane and a point in a  cup. The materials in contact are 
loaded close to or over the maximum allowed  stress. I advise against trying to 
use either of these types. 
 
    It may be beneficial to attach the top support ~3/4  along the arm as 
shown above. The balance between the mass and the weight of the  arm itself 
can be set to minimise the vertical load on the bottom bearing.  This 
improves the stability.
 
    Regards,
 
    Chris Chapman






In a message dated 09/04/2009, rog@.......... writes:
Is a 
  scalpel blade against a steel ball bearing the current best consensus fo=
r a 
  Lehman pivot?
rob


Hi Rob,
 
    Crossed cylinders can be a bit bett=
er. 
The surfaces need to highly polished. The SEP uses 1/8" OD polished Tungst=
en 
Carbide needle rollers, but these may be rather expensive / difficult to=
 obtain 
depending on where you live. See http://www.bgs.ac.uk/education/school_seismology/seismometer.html&nbs=
p;You 
mount both the vertical rollers on the frame. 
 
    However, a polished SS plane rolling on a SS=
 ball 
bearing works perfectly well and it may cost only 50 c! See http://www.smallparts.com/
 
    I use a 1/2" SS bearing and a large scalpel=
 blade 
'trimmed' oval with a carbide disk and glued to the arm with two part=
 
acrylic adhesive. You can use bearings down to 1/4" OK, but mount them on=
 the 
vertical frame, NOT on the arm! My top suspension is an 8 thou music wire.=
 I can 
set the period to 60 seconds OK and it works for some days, but it has 
a significant tilt drift. My 'normal' set period is 20 seconds. =
I 
measured +/-2mm drift over a fortnight with this, in a max drift rang=
e of 
+/-10mm.
 
    The two 'suspensions' which are 
very likey to give trouble are the knife edge on a plane and a point=
 in a 
cup. The materials in contact are loaded close to or over the maximum allo=
wed 
stress. I advise against trying to use either of these types. 
 
    It may be beneficial to attach the top suppor=
t ~3/4 
along the arm as shown above. The balance between the mass and the weight=
 of the 
arm itself can be set to minimise the vertical load on the bottom bea=
ring. 
This improves the stability.
 
    Regards,
 
    Chris Chapman