In a message dated 18/04/2009, rog@.......... writes:

Gary,
I am curious.
What are typical times for Lehman half-life  decay?
What are the main causes of  decay?
Rob



Hi Gary & Rob,
 
    With a Lehman, you first need to set both the cross  balance and the 
period of the pendulum, but with NO external damping and with  the sensor coil 
DISCONNECTED. Put a peel off label on the mass and another on a  match box 
or similar. Mark lines on both with a fibre pen. Also  mark dots on the box 
at +/-1mm, +/-5mm and +/-10 mm and place it on the  frame just clear of the 
mass by 1~2mm. You place the line on the box  at the centre of the frame, 
centre the mass by hand to align the marks, release  it and see which way it 
drifts. You adjust the cross balance screws so that it  stays central. 
 
    Then you deflect the arm 10 mm, release it and time  one FULL cycle of 
oscillation when the arm swings past the central alignment and  when it 
swings past again going in the same direction. Then adjust the tilt of  the 
boom, usually with a screw on the base near the mass end, till you  get the 
chosen period, which should be a minimum of 15 seconds. I use  20 seconds. If 
you have difficulty in getting a stable period of 15 seconds, you  will need 
to improve your suspension. You may have to trim the cross balance  while 
doing this. If you have lock nuts and spring washers on the adjusting  screws, 
tighten them and recheck the cross balance. 
 
    You can then repeat the deflection to 10  mm and time how long it takes 
until the swing is reduced to ~half  amplitude, 5mm. You should aim for a 
time of >4 minutes. If the time is  much less than this, you have a lossy / 
poor system which may need  correction. You can get problems if the top wire 
is not firmly clamped, if the  suspension can flex, if you are trying to use 
either a point or a knife edge  lower 'bearing', if the arm can rotate 
about it's longitudinal axis, if the  frame is not sufficiently rigid, or if 
base mountings are not firm. No material  is free of loss when the load on it 
is changed.  Air flow damping may  be significant if you are trying to use 
capacitative position sensing, but it is  usually very small for coil + magnet 
or LVDT sensors. 
 
    Then plug the coil into the amplifier, switch on  the power, deflect 
the mass, release it and check that it swings at least  one full cycle. The 
amplifier input resistor / input circuit may add  damping. Progressively add 
more damping until on release from the 10  mm deflected position, the arm 
swings 0.5 mm PAST the zero and then back to  zero.  It is quite easy to judge 
0.5 mm if you use a fixed x4 mounted  magnifying glass to observe the 
balance position. Trying to do this 'by eye' is  more difficult.
 
    I hope that this will help.
 
    Regards,
 
    Chris Chapman
    






In a message dated 18/04/2009, rog@.......... writes:
Gary,
I am curious.
What are typical times for Lehman hal=
f-life 
  decay?
What are the main causes of 
decay?
Rob


Hi Gary & Rob,
 
    With a Lehman, you first need to set both the=
 cross 
balance and the period of the pendulum, but with NO external damping and=
 with 
the sensor coil DISCONNECTED. Put a peel off label on the mass and another=
 on a 
match box or similar. Mark lines on both with a fibre pen. Also 
mark dots on the box at +/-1mm, +/-5mm and +/-10 mm and place it on=
 the 
frame just clear of the mass by 1~2mm. You place the line on the=
 box 
at the centre of the frame, centre the mass by hand to align the marks, re=
lease 
it and see which way it drifts. You adjust the cross balance screws so tha=
t it 
stays central. 
 
    Then you deflect the arm 10 mm, release it an=
d time 
one FULL cycle of oscillation when the arm swings past the central alignme=
nt and 
when it swings past again going in the same direction. Then adjust the til=
t of 
the boom, usually with a screw on the base near the mass end, till you 
get the chosen period, which should be a minimum of 15 seconds. =
I use 
20 seconds. If you have difficulty in getting a stable period of 15 second=
s, you 
will need to improve your suspension. You may have to trim the cross balan=
ce 
while doing this. If you have lock nuts and spring washers on the adjustin=
g 
screws, tighten them and recheck the cross balance. 
 
    You can then repeat the deflection to 10 
mm and time how long it takes until the swing is reduced to ~half 
amplitude, 5mm. You should aim for a time of >4 minutes. If the ti=
me is 
much less than this, you have a lossy / poor system which may need 
correction. You can get problems if the top wire is not firmly clamped, if=
 the 
suspension can flex, if you are trying to use either a point or a knife ed=
ge 
lower 'bearing', if the arm can rotate about it's longitudinal axis, if th=
e 
frame is not sufficiently rigid, or if base mountings are not firm. No mat=
erial 
is free of loss when the load on it is changed.  Air flow dampin=
g may 
be significant if you are trying to use capacitative position sensing, but=
 it is 
usually very small for coil + magnet or LVDT sensors. 
 
    Then plug the coil into the amplifier, switch=
 on 
the power, deflect the mass, release it and check that it swings at=
 least 
one full cycle. The amplifier input resistor / input circuit may add 
damping. Progressively add more damping until on release from th=
e 10 
mm deflected position, the arm swings 0.5 mm PAST the zero and then back=
 to 
zero.  It is quite easy to judge 0.5 mm if you use a fixed x4 mounted=
 
magnifying glass to observe the balance position. Trying to do this 'by ey=
e' is 
more difficult.
 
    I hope that this will help.
 
    Regards,
 
    Chris Chapman