Chris,

I have changed from the ball bearing, with wire suspension to crossed
cylinders for both the top and bottom bearing surfaces.
The wire suspension seems to cause noise when ever the the frame was
touched.
I wanted to get rid of the resonance of the wire so both the horizontal
boom, and the suspension are now tubes.

No more rotational problem around the horizontal mass axis either.

The last step was building the interface between the tubular suspension and
the vertical support, so I could use the crossed cylinders there as well.


  -----Original Message-----
  From: psn-l-request@.............. [mailto:psn-l-request@.................
Behalf Of ChrisAtUpw@.......
  Sent: Friday, June 15, 2007 2:10 PM
  To: psn-l@..............
  Subject: Re: How Pivotal is the pivotal relationship?


  In a message dated 15/06/2007, Paulc@........ writes:
    When constructing my first experimental Lehman units I used a piece of
steel channel iron mounted in a vice at the vertical member.

    The first bearing surface was the head of # 5 bolt polished, and mounted
into a holed drilled into the vertical member. The arm had a similar bolt
mounted in the end of an aluminum tube. This bolt had a hole in the center
of the head and a hard stainless ball glued into it.
  Hi Paul,

      You have two viable choices for a counterface for your SS ball
bearing - another piece of very hard SS - like a bit of a blade off a SS
craft knife - you can get break off blades - or a triangular tungsten
carbide lathe tool tip - the plain ones are not expensive. I polish mine
with diamond paste. Use Devweld 531, Devcon Plastic Welder, Loctite
Multibond, Permabond Quickbond or any other two component acrylic adhesive,
NOT an Epoxy. The acrylic glues have exceptional strength and are 'tough'.

    The suspension was a .020 piano wire and the mass was a 5 Lbs roll of
copper wire slide over the end of the aluminum pipe.
    The top of the suspension wire was simply drooped over the top of the
vertical member and  "C" clamped to it in order to made the pendulum as
close as possible to horizontal.
      You need to control any tendency of the arm to rotate about it's long
axis. The easy way to do this is to put a 6" crossbar just behind the mass
and have a V suspension to the top bearing. 7 strand SS fishing trace is
great for this and it comes with crimp tube fittings. 30 to 50 lbs line is
OK.
      You are better with a solid chunk of metal for the mass, like a slice
off a 3" to 4" OD brass rod.

  I have since gone with a lead mass

      You may find a mass of ~ 1 kg is easier, for both the suspensions.
      You need to provide a fixed / clamped top hinge point.

  This is exactly what I was using before the crossed cylinders.
    By moving the pivots laterally I could see the effect, and cause the
mass to stop where ever I wanted within reason.

    OK comes the problem.  After putting guitar tensioner on the top of the
vertical member, so I could adjust the pendulum level, I could no longer
cause the mass to stop anywhere.
    It seemed that no combination of level the pendulum, changing the mass
position moving the bottom pivot would allow the pendulum to stop.

    Finally I realized that the adjustment apparatus had moved the top wire
suspension point back slightly so that it was a bit behind the vertical
member.
    No longer directly vertical of the bottom pivot.

    I find that I need to have the top pivot just slightly forward of the
bottom pivot to get any pendulum I build to balance.

    Is there some rule of thumb...  It seems the closer to 90 degrees the
angle between pendulum and vertical becomes the easier the pendulum will
shift off balance.

    Currently my bottom pivot is crossed hard cylinders, and I want to do
the top this way, but I would like to know more about the "Pivotal
Relationship" before I construct the bearing.
      Tungsten Carbide drills are not expensive and the shafts make very
good bearings. You can also buy tungsten carbide needle rollers. You can
also use shoulder bolts type 435 from McMaster Carr.

      If you use a guitar tensioner, lead the wire through a hole in a
horizontal bolt and two washers and clamp them to define the top flex point
of the wire accurately - you need precisely defined clamp edges. I 'dish'
the washers slightly with a conical punch on a wood block.

      If the length between the bearing and the balance point of the arm
close to the mass is L, the period T = 2 x Pi x Sqrt( L / (g x sinA)), where
A is the angle between the true vertical and the line joining the centres of
rotation of the top and bottom bearings. It is typically ~1/3 degree, maybe
less - a very small angle. If you have a ball or a vertical roller on the
vertical support column, the centre of rotation is the centre of the ball or
the roller. If your ball or vertical roller is on the end of the arm --->
change it!  g = 9.81 if L is in metres, g = 32.2 if L is in feet.

   All of these considerations have been taken care of in my new sensor.

      Make the seismometer base a bit longer than the arm and make the
crossbar at the support end about 1/2 the base length ---> make a single
unit supporting everything on the three screws. If you try to use separate
units, as described on psn, you will get never ending problems. Set the
height / position of the mass, taking account of the sensor and damping
systems. You set up the suspension angles with three levelling screws on the
base of the seismometer. Set up the cross balance first, then set up the
period and re-trim the cross balance as necessary. Then set the damping to
~0.7 critical. If you push the arm 10mm to one side and release it, it
should swing ~1/2 mm past the balance point.

      Hope that this helps!

  Yes it does immensely, and validates some of the ideas I thought of
myself.

      Regards,

      Chris Chapman

  Thank you again Chris.

  PauLC
  W1VLF





Chris,
 
I have changed from the ball =
bearing, with=20
wire suspension to crossed cylinders for both the top and bottom bearing =

surfaces.
The wire suspension seems to cause =
noise=20
when ever the the frame was touched.
I wanted to get rid of the =
resonance of the=20
wire so both the horizontal boom, and the suspension are now =
tubes.
 
No more rotational problem around =
the=20
horizontal mass axis either.
 
The last step was building the =
interface=20
between the tubular suspension and the vertical support, so I could use =
the=20
crossed cylinders there as well.
 
 

  
-----Original Message-----
From:=20
  psn-l-request@.............. =
[mailto:psn-l-request@...............On Behalf=20
  Of ChrisAtUpw@.......
Sent: Friday, June 15, 2007 2:10=20
  PM
To: psn-l@..............
Subject: Re: How =
Pivotal is=20
  the pivotal relationship?


  
In a message dated 15/06/2007, Paulc@........ writes:
  
    
When constructing my first =
experimental=20
    Lehman units I used a piece of steel channel iron mounted in a vice =
at the=20
    vertical member.
    
 
    
The first bearing surface was =
the head=20
    of # 5 bolt polished, and mounted into a holed drilled into the=20
    vertical member. The arm had a similar bolt mounted =
in the=20
    end of an aluminum tube. This bolt had a hole in the center of the =
head and=20
    a hard stainless ball glued into it.
  
Hi Paul,
  
 
  
    You have two viable choices for a =
counterface for=20
  your SS ball bearing - another piece of very hard SS - like a bit =
of a=20
  blade off a SS craft knife - you can get break off blades - or a =
triangular=20
  tungsten carbide lathe tool tip - the plain ones are not expensive. I =
polish=20
  mine with diamond paste. Use Devweld 531, Devcon Plastic =
Welder, Loctite=20
  Multibond, Permabond Quickbond or any other two component acrylic =
adhesive,=20
  NOT an Epoxy. The acrylic glues have exceptional strength and are=20
  'tough'. 
  
 
  
    
The suspension was a .020 =
piano wire and=20
    the mass was a 5 Lbs roll of copper wire slide over the end of the =
aluminum=20
    pipe.
    
The top of the suspension wire =
was=20
    simply drooped over the top of the vertical member and  "C" =
clamped=20
    to it in order to made the pendulum as close as possible to=20
    horizontal. 
  
    You need to control any tendency of the =
arm to=20
  rotate about it's long axis. The easy way to do this is to put a 6" =
crossbar=20
  just behind the mass and have a V suspension to the top bearing. 7 =
strand SS=20
  fishing trace is great for this and it comes with crimp tube fittings. =
30 to=20
  50 lbs line is OK.
  
    You are better with a solid chunk =
of metal=20
  for the mass, like a slice off a 3" to 4" OD brass rod. 
  
 
  
I have =
since gone with=20
  a lead mass 
  
 
  
    You may find a mass of ~ 1 kg is easier, =
for both=20
  the suspensions. 
  
    You need to provide a fixed / clamped top =
hinge=20
  point. 
  
 
  
This is =
exactly what I=20
  was using before the crossed cylinders. 
  
    
By moving the pivots laterally =
I could=20
    see the effect, and cause the mass to stop where ever I wanted =
within=20
    reason.
    
 
    
OK comes the problem.  =
After=20
    putting guitar tensioner on the top of the vertical member, so I =
could=20
    adjust the pendulum level, I could no longer cause the mass to stop=20
    anywhere.
    
It seemed that no combination =
of level=20
    the pendulum, changing the mass position moving the bottom pivot =
would allow=20
    the pendulum to stop.
    
 
    
Finally I realized that the =
adjustment=20
    apparatus had moved the top wire suspension point back slightly so =
that it=20
    was a bit behind the vertical member. 
    
No longer directly vertical of =
the=20
    bottom pivot.
    
 
    
I find that I need to have the =
top pivot=20
    just slightly forward of the bottom pivot to get any pendulum I =
build to=20
    balance.
    
 
    
Is there some rule of =
thumb...  It=20
    seems the closer to 90 degrees the angle between pendulum and =
vertical=20
    becomes the easier the pendulum will shift off balance.
    
 
    
Currently my bottom pivot is =
crossed=20
    hard cylinders, and I want to do the top this way, but I would like =
to know=20
    more about the "Pivotal Relationship" before I construct the=20
    bearing.
  
    Tungsten Carbide drills are not =
expensive=20
  and the shafts make very good bearings. You can also buy tungsten =
carbide=20
  needle rollers. You can also use shoulder bolts type 435 from McMaster =

  Carr.
  
 
  
    If you use a guitar tensioner, lead the =
wire=20
  through a hole in a horizontal bolt and two washers and clamp =
them to=20
  define the top flex point of the wire accurately - you need precisely =
defined=20
  clamp edges. I 'dish' the washers slightly with a conical punch =
on a wood=20
  block.
  
 
  
    If the length between the bearing and the =
balance=20
  point of the arm close to the mass is L, the period T =3D 2 x Pi x =
Sqrt( L / (g=20
  x sinA)), where A is the angle between the true vertical and the line =
joining=20
  the centres of rotation of the top and bottom bearings. It is =
typically ~1/3=20
  degree, maybe less - a very small angle. If you have a ball or a =
vertical=20
  roller on the vertical support column, the centre of rotation is the =
centre of=20
  the ball or the roller. If your ball or vertical roller is on the end =
of the=20
  arm ---> change it!  g =3D 9.81 if L is in metres, g =3D =
32.2 if L is=20
  in feet. 
  
 
  
 All =
of these=20
  considerations have been taken care of in my new =
sensor.
  
 
  
    Make the seismometer base a bit longer =
than the=20
  arm and make the crossbar at the support end about 1/2 the base length =
--->=20
  make a single unit supporting everything on the three screws. =
If you=20
  try to use separate units, as described on psn, you will get =
never ending=20
  problems. Set the height / position of the mass, taking account =
of the=20
  sensor and damping systems. You set up the suspension angles with=20
  three levelling screws on the base of the seismometer. Set up the =
cross=20
  balance first, then set up the period and re-trim the cross balance as =

  necessary. Then set the damping to ~0.7 critical. If you push the arm =
10mm to=20
  one side and release it, it should swing ~1/2 mm past the balance =
point.
  
 
  
    Hope that this helps!  
  
 
  
Yes it =
does immensely,=20
  and validates some of the ideas I thought of =
myself. 
  
 
  
    Regards,
  
 
  
    Chris Chapman 
  
 
  
Thank you =
again=20
  Chris.
  
 
  
PauLC
  
W1VLF