## PSN-L Email List Message

Subject: Re: Seismograph for school
From: ChrisAtUpw@.......
Date: Tue, 3 Oct 2006 21:00:29 EDT

```In a message dated 2006/10/03, tchannel@.............. writes:

> 1. You used solder for the mass,  is the mass volume fixed or is there a
> wide range of weights, one could use. I guess the mass is to overcome the
> friction of the hinge and stay stationary. The reason I ask is I see a wide range
> of masses, but most around 5#.  What is yours? and what is your beam length?

Hi Ted,

The period of a simple pendulum T = 2 x Pi x sqrt(L / g) where L is t
he length in m and g = 9.81 m / sec^2. With a garden gate arrangement you can
extend the period by x20 or more, but this gets increasingly difficult since
the suspension angles get tiny and you may get dimensional / temperature
associated lack of stability. Make the cross bar width ~1/3 the base length. There
are pendulum design sheets on John Lahr's website. If A is the angle between the
vertical and the rotation axis of the gate, you multiply g in the equation
above by sine A.  You will be using tilts of less than 1 degree. A 22" arm (1.5
ec) set to 20 sec requires an angle of 0.322 deg; set to 30 sec it requires
0.143 deg.
The period is independent of the mass of the pendulum, but depends on
the 'radius of gyration' of the whole moving system. A heavy connecting beam
can significantly reduce this. A reasonable mass is 1 to 2 lbs with about a 2
ft arm - 1.5 sec. If you use more weight than this, you will need a strong top
suspension. I suggest that you work out this force and look up the strength of
The idea behind the suspension is that there should be as little
friction as possible. Pre electronic seismometers used mechanical lever gain
linkages to move the pen and hence needed a large mass to overcome the friction.

> 2. Again the solder, is non magnetic is this important as I see many things
> use as mass including iron?

I don't use any magnetic materials on the boom 'in principle'. You
won't know for sure what is a magnetic interaction and what is seismic. If I am
taking the time and effort to build a seismometer, avoiding obvious problems
just makes good sense. Like put the pickup coil on the arm and the magnets on
the baseplate!

> 3.  I see you used a resistor as a damp, and not a external damper device.
> I have read about doing this, but could not understand the values for the
> resistor.  You used 94k (in parallel?) acrossed the coil?  How did you arrive
> at that value?

Experiment. Exactly the same principle - any relative movement induces
a voltage in the coil proportionate to the magnetic field, the number of
turns on the coil, the velocity and the circuit resistance provides the damping
loss. The lower this resistance, the greater the loss, but this also reduces the
output voltage. Not significantly in this design, since the magnetic field
and coupling are very high and the mass is low.

> 4. You indicate to use a 100k in place of the 10k when using Larry's amp,
> which I will be using. Why is that?

Because an input resistor of 10 K is less than the required damping
resistor for this particular sensor design.

> 5.  The placement of the sensor: Could it be on other than a concrete
> floor, a normal wood floor? Could it be on carpet? I know that is not the best
> choice, but many schools have such floors. Or should one not bother with any
> other surface but a concrete floor?

You are likely to find wood floors both noisy and lacking tilt
stability - can be critical for a long period Lehman.  I can't say no, but try it
with little expectation of success? You can put wide melamine shelving on a
carpet and add weight to hold it firm, but the thicker the carpet the lower the
stability. Expect drift with temperature, humidity and time. Bricks? On a
concrete floor, I stick on 2" squares of 1/8" SS plate with pool adhesive to provide
a good base for the adjusting screws.

Regards,

Chris Chapman
In a me=
ssage dated 2006/10/03, tchannel@.............. writes:

1. You used solder for the mass=
,  is the mass volume fixed or is there a wide range of weights, one co=
uld use. I guess the mass is to overcome the friction of the hinge and stay=20=
stationary. The reason I ask is I see a wide range of masses, but most aroun=
d 5#.  What is yours? and what is your beam length?

Hi Ted,

The period of a simple pendulum T =3D 2=
x Pi x sqrt(L / g) where L is the length in m and g =3D 9.81 m / sec^2. Wit=
h a garden gate arrangement you can extend the period by x20 or more, but th=
is gets increasingly difficult since the suspension angles get tiny and you=20=
may get dimensional / temperature associated lack of stability. Make the cro=
ss bar width ~1/3 the base length. There are pendulum design sheets on John=20=
Lahr's website. If A is the angle between the vertical and the rotation axis=
of the gate, you multiply g in the equation above by sine A.  You will=
be using tilts of less than 1 degree. A 22" arm (1.5 ec) set to 20 sec requ=
ires an angle of 0.322 deg; set to 30 sec it requires 0.143 deg.
The period is independent of the mass o=
f the pendulum, but depends on the 'radius of gyration' of the whole moving=20=
system. A heavy connecting beam can significantly reduce this. A reasonable=20=
mass is 1 to 2 lbs with about a 2 ft arm - 1.5 sec. If you use more weight t=
han this, you will need a strong top suspension. I suggest that you work out=
this force and look up the strength of your wire?
The idea behind the suspension is that=20=
there should be as little friction as possible. Pre electronic seismometers=20=
used mechanical lever gain linkages to move the pen and hence needed a large=
mass to overcome the friction.

2. Again the solder, is non mag=
netic is this important as I see many things use as mass including iron?

I don't use any magnetic materials on=20=
the boom 'in principle'. You won't know for sure what is a magnetic interact=
ion and what is seismic. If I am taking the time and effort to build a seism=
ometer, avoiding obvious problems just makes good sense. Like put the pickup=
coil on the arm and the magnets on the baseplate!

3.  I see you used a resis=
tor as a damp, and not a external damper device.  I have read about doi=
ng this, but could not understand the values for the resistor.  You use=
d 94k (in parallel?) acrossed the coil?  How did you arrive at that val=
ue?

Experiment. Exactly the same principle=
- any relative movement induces a voltage in the coil proportionate to the=20=
magnetic field, the number of turns on the coil, the velocity and the circui=
t resistance provides the damping loss. The lower this resistance, the great=
er the loss, but this also reduces the output voltage. Not significantly in=20=
this design, since the magnetic field and coupling are very high and the mas=
s is low.

4. You indicate to use a 100k i=
n place of the 10k when using Larry's amp, which I will be using. Why is tha=
t?

Because an input resistor of 10 K is l=
ess than the required damping resistor for this particular sensor design.

5.  The placement of the=20=
sensor: Could it be on other than a concrete floor, a normal wood floor? Cou=
ld it be on carpet? I know that is not the best choice, but many schools hav=
e such floors. Or should one not bother with any other surface but a concret=
e floor?

You are likely to find wood floors both=
noisy and lacking tilt stability - can be critical for a long period Lehman=
..  I can't say no, but try it with little expectation of success? You c=
an put wide melamine shelving on a carpet and add weight to hold it firm, bu=
t the thicker the carpet the lower the stability. Expect drift with temperat=
ure, humidity and time. Bricks? On a concrete floor, I stick on 2" squares o=
f 1/8" SS plate with pool adhesive to provide a good base for the adjusting=20=
screws.

Regards,

Chris Chapman
```

[ Top ] [ Back ] [ Home Page ]