In a message dated 10/02/2009, gmvoeth@........... writes:

build the mechanics to any type  of seismometer
then use several different  sensors to see which
produces the best  results?

Hi Geoff,
 
    That is not necessary.

does anyone know what the best  would be?

    Yes, but they are also sample time and design /  range dependant.
 
    In sequence of best first
    Capacitative. LCDT 10^-12 m
    Inductive. LVDT 10^-10 m
    Optical. 10^-8 m
    Hall Magnetic. 3x10^-8 m
 
    Wavelength optical is sometimes used, but it gets  excessively expensive 
below 5x10^-7 m
 
    The absolute stability may only be better than  ~10^-8 m.
 
    Velocity sensors tend to be grouped around  Inductive, but are frequency 
dependant. They are not so effective at long  periods.
 
    I am excluding the operation of superconducting  sensors. They can have 
extremely high stability.

are there mechanical ways to damp  (remove energy) that might work as well as 
 magnetic?

    Viscosity methods are difficult to use and strongly  temperature 
dependant.
    You can use phased electrostatic feedback as well  as electromagnetic. 

It might be nice to have a chord  for a pendulum that will damp as it flexes?

    Not usually necessary. Springs sometimes give  problems, particuarly coil 
springs. It is easier to apply magnetic damping to  leaf springs.

is it possible to levitate two  magnets against each other then measure 
changes in the field as the two magnets  bounce south against  south or vice-versa?

    Yes, but it is non linear and temperature  dependant. Solenoid + magnet 
can be used, but there is still a temperature  dependance and a highly 
stabilised current requirement.

it seems to me that whatever  sensor you use you need two identical for a 
proper input into a differential  amplifier?
that alone would increase your  output by 6Dbv or X2?

    That may not be necessary. A single sensor can  often drive a 
differential amplifier OK. Single sensors drive non differential  amplifiers just fine.
 
    Regards,
 
    Chris Chapman





In a message dated 10/02/2009, gmvoeth@........... writes:
<=
FONT=20
  style=3D"BACKGROUND-COLOR: transparent" face=3DArial color=3D#000000 size=
=3D2>
  
build the mechanics to an=
y type=20
  of seismometer
  
then use several differen=
t=20
  sensors to see which
  
produces the best=20
  results?
Hi Geoff,
 
    That is not necessary.
<=
FONT=20
  style=3D"BACKGROUND-COLOR: transparent" face=3DArial color=3D#000000 size=
=3D2>
  
does anyone know what the=
 best=20
  would be?
    Yes, but they are also sample time and design /=
=20
range dependant.
 
    In sequence of best first
    Capacitative. LCDT 10^-12 m
    Inductive. LVDT 10^-10 m
    Optical. 10^-8 m
    Hall Magnetic. 3x10^-8 m
 
    Wavelength optical is sometimes used, but it ge=
ts=20
excessively expensive below 5x10^-7 m
 
    The absolute stability may only be better than=20
~10^-8 m.
 
    Velocity sensors tend to be grouped around=20
Inductive, but are frequency dependant. They are not so effective at long=20
periods.
 
    I am excluding the operation of superconducting=
=20
sensors. They can have extremely high stability.
<=
FONT=20
  style=3D"BACKGROUND-COLOR: transparent" face=3DArial color=3D#000000 size=
=3D2>
  
are there mechanical ways=
 to damp=20
  (remove energy) that might work as well as=20
magnetic?
    Viscosity methods are difficult to use and stro=
ngly=20
temperature dependant.
    You can use phased electrostatic feedback as we=
ll=20
as electromagnetic. 
<=
FONT=20
  style=3D"BACKGROUND-COLOR: transparent" face=3DArial color=3D#000000 size=
=3D2>
  
It might be nice to have=20=
a chord=20
  for a pendulum that will <=
FONT=20
  style=3D"BACKGROUND-COLOR: transparent" face=3DArial color=3D#000000 size=
=3D2>
  
is it possible to levitat=
e two=20
  magnets against each other then measure changes in the field as the two magnets=20
  bounce south a=
gainst=20
  south or vice-versa?
    Yes, but it is non linear and temperature=20
dependant. Solenoid + magnet can be used, but there is still a temperature=20
dependance and a highly stabilised current requirement.
<=
FONT=20
  style=3D"BACKGROUND-COLOR: transparent" face=3DArial color=3D#000000 size=
=3D2>
  
it seems to me that whate=
ver=20
  sensor you use you need two identical for a proper input into a differential=20
  amplifier?
  
that alone would increase=
 your=20
  output by 6Dbv or X2?
    That may not be necessary. A single sensor can=20
often drive a differential amplifier OK. Single sensors drive non differenti=
al=20
amplifiers just fine.
 
    Regards,
 
    Chris Chapman