In a message dated 20/02/2010, dickthomas01@............. writes:

Any  suggestions?



Hi Tom,
 
    Sure. Put the amplifier and ADC inside an Al box  with a screwed on lid 
/ folds. Use twin core input cable with a braided copper  screen. Connect 
the amplifier input zero line to the box and to the braided  screen and earth 
the box separately. Connect the seismometer magnet frame to the  braided 
screen to shield the coil. Preferably also earth the seismometer  frame. You 
may find 0.01 mu F ceramic capacitor and a 50 Ohm resistor in series  with 
the coil at the coil terminals will remove most of the RF pickup. Check  that 
the cable length is not at a resonant frequency of the transmitter.
    You can also make quite good LC filters  for the AC house power supply 
lines. I tend to make my own using larger L  ferrite values than commercial 
filters. You can get feedback through these  lines.
    Recognise that filters only give a reduction in the  error signals. If 
you have too high a signal level, some will inevitably get  through. You 
might try using soldered / earthed fencing netting on the ceiling  of your 
cellar where the seismometers are located?
    I once tried to heat a 6" rotating test disk with a  3.5 MHz 3 kW power 
oscillator. It caused chart recorders with shorted inputs to  go off scale 
on the other side of the laboratory. This convinced me to try  heating the 
disk with a spiral heating coil off a kitchen cooker, which  avoided the 
problems.  
 
    Good Luck!
    
    Chris






In a message dated 20/02/2010, dickthomas01@............. writes:
Any=20
  suggestions?


Hi Tom,
 
    Sure. Put the amplifier and ADC inside an Al=
 box=20
with a screwed on lid / folds. Use twin core input cable with a braided co=
pper=20
screen. Connect the amplifier input zero line to the box and to the=
 braided=20
screen and earth the box separately. Connect the seismometer magnet frame=
 to the=20
braided screen to shield the coil. Preferably also earth the seismome=
ter=20
frame. You may find 0.01 mu F ceramic capacitor and a 50 Ohm resistor in=
 series=20
with the coil at the coil terminals will remove most of the RF pickup. Che=
ck=20
that the cable length is not at a resonant frequency of the transmitter.
    You can also make quite good LC fil=
ters=20
for the AC house power supply lines. I tend to make my own using larg=
er L=20
ferrite values than commercial filters. You can get feedback through=
 these=20
lines.
    Recognise that filters only give a reduction=
 in the=20
error signals. If you have too high a signal level, some will inevitably=
 get=20
through. You might try using soldered / earthed fencing netting on the cei=
ling=20
of your cellar where the seismometers are located?
    I once tried to heat a 6" rotating test disk=
 with a=20
3.5 MHz 3 kW power oscillator. It caused chart recorders with shorted inpu=
ts to=20
go off scale on the other side of the laboratory. This convinced me to try=
=20
heating the disk with a spiral heating coil off a kitchen cooker, whi=
ch=20
avoided the problems.  
 
    Good Luck!
    
    Chris