In a message dated 2007/07/18, gpayton880@....... writes:

> I have always wondered why a mass is needed.  Why could not the coil itself 
> be used as the mass on a pendulum seismometer?

Hi Jerry,

       It could be and it has been. But a coil may not be quite as 
dimensionally stable as a solid mass and you don't really need large coils in these 
days, with NdFeB magnets readily available. One substitute for a mass is to use a 
1/4" to 3/8" thick horizontal copper plate for both the mass and the damping 
blade. Winding a coil of maybe 1 lb of wire takes a lot of time and care - 
apart from the cost! I find coil winding an effort, not a pleasure! The biggest 
coil that I ever wound was 80,000 turns and it took several days to make with 
interleaved doped paper layers.

       You need a certain mass for the motion not to be swamped by thermal or 
environmental agitation / noise. 
    It is the product M x T x Q, which is important for noise considerations, 
not just the mass M. T is the set period and Q is the mechanical analogue of 
the intrinsic damping (w x L / R for an oscillating circuit). 

>  If I understand what I have read in my books (questionable), the size and 
> weight of a mass has nothing to do with the period.  

       Certainly in theory, but you have to suspend the mass at the end of an 
arm, which needs to be light in comparison to the mass - otherwise it 
seriously shortens the period. You are in practice limited by the materials that you 
can easily buy. You need the distance between the centre of mass and the hinge 
to be maximised. You need a certain minimum mass to get over thermal 
excitation, but usually a lot more to avoid the effects of small air currents and tiny 
losses in the suspension system. The arm / suspension needs to be very rigid. 
Very light mass systems may appear unstable, noisy or 'sticky'.
       Not all suspension systems are equal, some have much more inbuilt 
friction / losses than others. The point in a cup and the real knife edge 
suspensions are particularly poor. The ball on a plane and crossed cylinders are much 
better.

Furthermore, I believe the length of the pendulum and supporting spring or 
wire on a 
> horizontal or vertical sensor matters more for obtaining the wanted period, 
> excluding friction loses etc.  The coil is only the desired method of 
> detecting movement. 

       If you hang up the mass by the boom and hinge vertically, it will have 
a period given ~by the usual formula 2 x Pi x Sqrt(L/g). You then reduce the 
restoring force by suspending it a small angle to the vertical to get the 
'garden gate' arrangement. Getting an increase in period of x10 is not difficult 
(1/100 of g), but larger increases get progressively more difficult, as the 
suspension angle gets below 0.3 of a degree. This also makes the system highly 
sensitive to ground tilt effects, more difficult to adjust and keep in 
adjustment and it requires a very good low loss suspension or you will find that you 
can't get stable long periods.

       Electromagnetic damping is usually used. It can be made simple and 
easy to adjust. You can use quad magnets on mild steel backing plates and a 
Copper blade, or you can use larger quad magnets to drive the sensor coil and put a 
variable load resistor across that. I prefer the damping blade, since you can 
use it to limit the swing of the arm and it has the lowest noise.
       Oil damping tends to be relatively difficult to adjust, non linear in 
it's effect and very temperature sensitive. It is also messy. Oil creeps over 
most surfaces and they then collect dust and dirt. The operation looks 
superficially simple, but it is a headache waiting to happen. 

       Verticals use a spring of some sort to counterbalance the 
gravitational force. You can get periods out to about 5 seconds with ordinary steel 
springs, but the thermal drift of the spring constant limits you. The system just 
'collapses'. Professional systems use NiSpanC springs which are extremely 
stable. 

       There are a variety of possible sensors. The simplest is the magnet + 
coil velocity sensor. These are OK for 'normal' use, but are noise limited at 
long periods, say over 60 seconds. Optical and Hall magnetic effect position 
sensors can also be used on amateur equipment. Professional equipment tends to 
use LVDT magnetic or LCDT capacitative sensors. The LVDT types have a noise 
limit of about 0.1 nm. Capacitative types can get an extra factor of 100 on 
this.

       Force feedback may be used with a position sensor to stabilise the 
operating position and to extend the natural period of the seismometer. 

       This is a very rough description of the 'background' to amateur 
seismometer design. Most of the limitations are practical - there are good and not 
so good ways of making things, relatively cheap and quite expensive materials. 
Choosing only the cheapest options are likely to limit the performance.

       Hope that it is of some help / interest.

       Regards,

       Chris Chapman   
In a me=
ssage dated 2007/07/18, gpayton880@....... writes:



I have always wondered why a ma=
ss is needed.  Why could not the coil itself be used as the mass on a p=
endulum seismometer?




Hi Jerry,



       It could be and it has been. But a coil=
 may not be quite as dimensionally stable as a solid mass and you don't real=
ly need large coils in these days, with NdFeB magnets readily available. One=
 substitute for a mass is to use a 1/4" to 3/8" thick horizontal copper plat=
e for both the mass and the damping blade. Winding a coil of maybe 1 lb of w=
ire takes a lot of time and care - apart from the cost! I find coil winding=20=
an effort, not a pleasure! The biggest coil that I ever wound was 80,000 tur=
ns and it took several days to make with interleaved doped paper layers.



       You need a certain mass for the motion=20=
not to be swamped by thermal or environmental agitation / noise. 

    It is the product M x T x Q, which is important for noise=
 considerations, not just the mass M. T is the set period and Q is the mecha=
nical analogue of the intrinsic damping (w x L / R for an oscillating circui=
t). 



 If I understand what I have re=
ad in my books (questionable), the size and weight of a mass has nothing to=20=
do with the period.  




       Certainly in theory, but you have to s=
uspend the mass at the end of an arm, which needs to be light in comparison=20=
to the mass - otherwise it seriously shortens the period. You are in practic=
e limited by the materials that you can easily buy. You need the distance be=
tween the centre of mass and the hinge to be maximised. You need a certain m=
inimum mass to get over thermal excitation, but usually a lot more to avoid=20=
the effects of small air currents and tiny losses in the suspension system.=20=
The arm / suspension needs to be very rigid. Very light mass systems may app=
ear unstable, noisy or 'sticky'.

       Not all suspension systems are equal, s=
ome have much more inbuilt friction / losses than others. The point in a cup=
 and the real knife edge suspensions are particularly poor. The ball on a pl=
ane and crossed cylinders are much better.



Furthermore, I believe the length of the pendulum and supporting spring or w=
ire on a 

horizontal or vertical sensor=20=
matters more for obtaining the wanted period, excluding friction loses etc.&=
nbsp; The coil is only the desired method of detecting movement. 



       If you hang up the mass by the boom and=
 hinge vertically, it will have a period given ~by the usual formula 2 x Pi=20=
x Sqrt(L/g). You then reduce the restoring force by suspending it a small an=
gle to the vertical to get the 'garden gate' arrangement. Getting an increas=
e in period of x10 is not difficult (1/100 of g), but larger increases get p=
rogressively more difficult, as the suspension angle gets below 0.3 of a deg=
ree. This also makes the system highly sensitive to ground tilt effects, mor=
e difficult to adjust and keep in adjustment and it requires a very good low=
 loss suspension or you will find that you can't get stable long periods.


       Electromagnetic damping is usually used=
.. It can be made simple and easy to adjust. You can use quad magnets on mild=
 steel backing plates and a Copper blade, or you can use larger quad magnets=
 to drive the sensor coil and put a variable load resistor across that. I pr=
efer the damping blade, since you can use it to limit the swing of the arm a=
nd it has the lowest noise.

       Oil damping tends to be relatively diff=
icult to adjust, non linear in it's effect and very temperature sensitive. I=
t is also messy. Oil creeps over most surfaces and they then collect dust an=
d dirt. The operation looks superficially simple, but it is a headache waiti=
ng to happen. 



       Verticals use a spring of some sort to=20=
counterbalance the gravitational force. You can get periods out to about 5 s=
econds with ordinary steel springs, but the thermal drift of the spring cons=
tant limits you. The system just 'collapses'. Professional systems use NiSpa=
nC springs which are extremely stable. 



       There are a variety of possible sensors=
.. The simplest is the magnet + coil velocity sensor. These are OK for 'norma=
l' use, but are noise limited at long periods, say over 60 seconds. Optical=20=
and Hall magnetic effect position sensors can also be used on amateur equipm=
ent. Professional equipment tends to use LVDT magnetic or LCDT capacitative=20=
sensors. The LVDT types have a noise limit of about 0.1 nm. Capacitative typ=
es can get an extra factor of 100 on this.



       Force feedback may be used with a posit=
ion sensor to stabilise the operating position and to extend the natural per=
iod of the seismometer. 



       This is a very rough description of the=
 'background' to amateur seismometer design. Most of the limitations are pra=
ctical - there are good and not so good ways of making things, relatively ch=
eap and quite expensive materials. Choosing only the cheapest options are li=
kely to limit the performance.



       Hope that it is of some help / interest=
..



       Regards,



       Chris Chapman