In a message dated 06/03/03, karl@............ writes:

> I have just built and setup a Lehman design seismometer. I built it for my 
> school science fair, so keep in mind when replying that I am not in tune to 
> all of the technical stuff. My questions have to do with adjusting the 
> seismometer to record correctly.

Hi Karl,

       It would be a great help if you could put a few Webcam or better 
pictures of the seismometer on the Website and some idea of the dimensions. 
Otherwise, we can only give you the most general help. It is sometimes 
possible to spot mistakes in construction, etc. this way and to make helpful 
suggestions for improvements and adjustments.

> Attached is a sample of the data that I am currently getting. I was 
> wondering if I need to damp more or if I need to turn down the gain or 
> adjust the offset. I am using Larry Cochrane's Amp/Filter board and his 16 
> bit A/D converter board. I have also read about changing the period length, 
> but I am not sure how to change it. Any help getting this data to look 
> 

       The data covers ~8,000 counts, or 1/8 of the total range, but there 
seems to be no indication of the total time, or whether this was a local 
seismic event, etc. We need to know what we are looking at! A very 
underdamped pendulum can move quite a lot. If your trace is just enviromental 
noise, you have either much too little damping, much too high a gain, a great 
deal of local noise, or all three! Do you have a magnet mounted on the moving 
arm?

       You need to set the period first and then set the damping to match. If 
the centre of mass to pivot is length L, the period P = 2 x Pi x Sqrt(L / (g 
x SinA)), where A is the angle between the axis of rotation of the arm and 
the vertical (may be less than 1 degree). You can adjust this by changing the 
lengthways slope of the baseplate with an adjusting screw, or by adjusting 
the position of the nozzle and the length of the top wire, if you have this 
setup. Reduce or remove the damping, deflect the arm maybe 1/2" and time it's 
oscillations. Try for the slower end of 12 to 20 seconds. You may find that 
beyond some period, like 20 seconds, the arm becomes unstable or erratic in 
it's balance position, so it needs to be set for a lower period. Having 
adjusted the period, you replace the damping, displace the arm a small 
fraction of an inch, release it and see what it does. If it returns to the 
central position with hardly any overshoot, this is fine. If it only very 
slowly returns to zero with no overshoot at all, it is overdamped. If it 
swings through one or more oscillations, it is underdamped. It is quickest to 
do this visually at first and when you get it nearly right, reduce the 
amplifier gain and take a recording. 

       Hope that this is of some help,

       Regards,

       Chris Chapman
In a message dated 06/03/=
03, karl@............ writes:



I have just built and setup=
 a Lehman design seismometer. I built it for my school science fair, so keep=
 in mind when replying that I am not in tune to all of the technical stuff.=20=
My questions have to do with adjusting the seismometer to record correctly.<=
/FONT>




Hi Karl,



       It would be a great help if you cou=
ld put a few Webcam or better pictures of the seismometer on the Website and=
 some idea of the dimensions. Otherwise, we can only give you the most gener=
al help. It is sometimes possible to spot mistakes in construction, etc. thi=
s way and to make helpful suggestions for improvements and adjustments.



Attached is a sample of the=
 data that I am currently getting. I was wondering if I need to damp more or=
 if I need to turn down the gain or adjust the offset. I am using Larry Coch=
rane's Amp/Filter board and his 16 bit A/D converter board. I have also read=
 about changing the period length, but I am not sure how to change it. Any h=
elp getting this data to look right would be appreciated.



       The data covers ~8,000 counts, or 1=
/8 of the total range, but there seems to be no indication of the total time=
, or whether this was a local seismic event, etc. We need to know what we ar=
e looking at! A very underdamped pendulum can move quite a lot. If your trac=
e is just enviromental noise, you have either much too little damping, much=20=
too high a gain, a great deal of local noise, or all three! Do you have a ma=
gnet mounted on the moving arm?



       You need to set the period first an=
d then set the damping to match. If the centre of mass to pivot is length L,=
 the period P =3D 2 x Pi x Sqrt(L / (g x SinA)), where A is the angle betwee=
n the axis of rotation of the arm and the vertical (may be less than 1 degre=
e). You can adjust this by changing the lengthways slope of the baseplate wi=
th an adjusting screw, or by adjusting the position of the nozzle and the le=
ngth of the top wire, if you have this setup. Reduce or remove the damping,=20=
deflect the arm maybe 1/2" and time it's oscillations. Try for the slower en=
d of 12 to 20 seconds. You may find that beyond some period, like 20 seconds=
, the arm becomes unstable or erratic in it's balance position, so it needs=20=
to be set for a lower period. Having adjusted the period, you replace the da=
mping, displace the arm a small fraction of an inch, release it and see what=
 it does. If it returns to the central position with hardly any overshoot, t=
his is fine. If it only very slowly returns to zero with no overshoot at all=
, it is overdamped. If it swings through one or more oscillations, it is und=
erdamped. It is quickest to do this visually at first and when you get it ne=
arly right, reduce the amplifier gain and take a recording.=20



       Hope that this is of some help,



       Regards,



       Chris Chapman