In a message dated 16/02/03, bromberg@......... writes:

> I am currently a freshman at Colorado School of Mines and last semester fo=
r=20
> one of my classes I was required to design and build a working seismometer=
..=20
>  Our final design consisted of a folded pendulum which used magnetic=20
> induction, an amplifying filter circuit, an A/D converter, and Alan Jones=
=E2=80=99=20
> AmaSeis software.  I am continuing on with my folded pendulum design this=20
> semester and I am trying to make some improvements on it but I have some=20
> questions that I am having trouble finding the answer for so if anyone=20
> could help me find some answers I would really appreciate it. =20
> =20
> Where could I find a thin flexible metal sheet that I could use for the=20
>=20

Hi Brady,

       Look for Model Shops which stock K&S metals http://www.ksmetals.com/=20
They sell 2 thou stainless steel and also 2 thou brass.=20
       Alternatively, look for a tool shop or Engineering supplier who stock=
=20
1.5 thou 'shim strip' or 'feeler gauge' strip. It is about 1/2" wide and 12"=
=20
long.=20
       For fine steel wire, go to a music shop and ask for mandolin strings.=
=20
These are stocked down to 7 thou. dia. You can get nickel plated strings for=
=20
guitars from 8 thou dia. See http://www.daddario.com/

W
> here could I find what I believe is called a PZ electro crystal (the cryst=
al=20
>=20

       Spark igniters for gas appliances / cigarette lighters use them. They=
=20
react to large voltage changes, not to current. Suggest that you do a Web=20
search. I think that I have seen them being sold as PZT crystals.

W
> here could I find information about optimizing the magnet / coil / damper=20
>=20

       I can make suggestions when I know something about the design of your=
=20
seismometer, the mass and the likely drift. I suggest using Neodymium magnet=
s=20
in groups of two or four with mild steel backing plates. See=20
www.wondermagnet.com for rectangular / V shaped magnets. These can give a=20
considerable increase in sensitivity and linearity over Alnico U magnets.

H
> ow would I go about integrating an A/D converter into my circuit that woul=
d=20
> have an output of =E2=80=9C  98156158 =E2=80=9C for the va=
lues 98, 156,=20
>=20

    The data logging programmes usually have this built in as part of the=20
software. Dataq provide suggested programme modules for their systems.=20

    Otherwise, you would have to write a computer programme to do this  / us=
e=20
a dedicated processor like a Basic Stamp or a PIC.

H
> ow many samples per second does the 8 bit have compared to the 10 bit and=20
> the 12 bit, and what exactly does bit mean?

    You look up the max sample rate from the data sheet for the device. You=20
can download these from Manufacturer's Websites. Then you have the problem o=
f=20
access to the computer. For the $24 Dataq 10 bit A/Ds, you have a maximum of=
=20
240 samples / sec total, but you can opt for less and four channels will giv=
e=20
1/4 this. You choose the number of samples / sec from the max frequency you=20
wish to record. For a frequency of f, the minimum would be 2f sps and a more=
=20
reasonable rate would be 4f or 5f.

    Numbers are held in a computer in binary form. Instead of using 0 to 9 i=
n=20
the decimal system for each decade, you use 0 and 1. So 13 decimal =3D 10 +=20=
3;=20
in powers of 2 =3D 8 + 4 + 0 + 1 =3D 1101 in binary. A byte is a term for 8=20=
bits=20
=3D numbers from 0 to 255.

       A 10 bit A/D will have 1023 steps for the total range (1024 levels).=20
For a +/-10V input range, the step size will be 4.88 milli V. A 12 bit A/D=20
will have 4095 steps and a 16 bit will have 65535.

> If you have any questions about my design, advice, or helpful information=20
> you can contact me at bromberg@mines=
..edu

       It is a bit difficult providing good advice when we do not have a=20
drawing or description of your proposed device. John Lahr has a picture on=20
his Website, but it does not show sufficient detail. Is there an online copy=
=20
of your report anywhere?

       Regards,

       Chris Chapman
In a message dated 16/02/03, bromberg@......... writes=
:



I am currently a freshman a=
t Colorado School of Mines and last semester for one of my classes I was req=
uired to design and build a working seismometer.  Our final design cons=
isted of a folded pendulum which used magnetic induction, an amplifying filt=
er circuit, an A/D converter, and Alan Jones=E2=80=99 AmaSeis software. &nbs=
p;I am continuing on with my folded pendulum design this semester and I am t=
rying to make some improvements on it but I have some questions that I am ha=
ving trouble finding the answer for so if anyone could help me find some ans=
wers I would really appreciate it.  

=20

Where could I find a thin flexible metal sheet that I could use for the=20=
inverted pendulum joints?



Hi Brady,



       Look for Model Shops which stock K&=
amp;S metals http://www.ksmetals.com/ They sell 2 thou stainless steel and a=
lso 2 thou brass.=20

       Alternatively, look for a tool shop=
 or Engineering supplier who stock 1.5 thou 'shim strip' or 'feeler gauge' s=
trip. It is about 1/2" wide and 12" long.=20

       For fine steel wire, go to a music=20=
shop and ask for mandolin strings. These are stocked down to 7 thou. dia. Yo=
u can get nickel plated strings for guitars from 8 thou dia. See http://www.=
daddario.com/



W

here could I find what I be=
lieve is called a PZ electro crystal (the crystal that expands when it recei=
ves a current) for testing our unit?



       Spark igniters for gas appliances /=
 cigarette lighters use them. They react to large voltage changes, not to cu=
rrent. Suggest that you do a Web search. I think that I have seen them being=
 sold as PZT crystals.



W

here could I find informati=
on about optimizing the magnet / coil / damper system?



       I can make suggestions when I know=20=
something about the design of your seismometer, the mass and the likely drif=
t. I suggest using Neodymium magnets in groups of two or four with mild stee=
l backing plates. See www.wondermagnet.com for rectangular / V shaped magnet=
s. These can give a considerable increase in sensitivity and linearity over=20=
Alnico U magnets.



H

ow would I go about integra=
ting an A/D converter into my circuit that would have an output of =E2=80=
=9C  98<cr><lf>156<cr><lf>158=20=
=E2=80=9C for the values 98, 156, 158?




    The data logging programmes usually have this built i=
n as part of the software. Dataq provide suggested programme modules for the=
ir systems.=20



    Otherwise, you would have to write a computer program=
me to do this  / use a dedicated processor like a Basic Stamp or a PIC.



H

ow many samples per second=20=
does the 8 bit have compared to the 10 bit and the 12 bit, and what exactly=20=
does bit mean?




    You look up the max sample rate from the data sheet f=
or the device. You can download these from Manufacturer's Websites. Then you=
 have the problem of access to the computer. For the $24 Dataq 10 bit A/Ds,=20=
you have a maximum of 240 samples / sec total, but you can opt for less and=20=
four channels will give 1/4 this. You choose the number of samples / sec fro=
m the max frequency you wish to record. For a frequency of f, the minimum wo=
uld be 2f sps and a more reasonable rate would be 4f or 5f.



    Numbers are held in a computer in binary form. Instea=
d of using 0 to 9 in the decimal system for each decade, you use 0 and 1. So=
 13 decimal =3D 10 + 3; in powers of 2 =3D 8 + 4 + 0 + 1 =3D 1101 in binary.=
 A byte is a term for 8 bits =3D numbers from 0 to 255.



       A 10 bit A/D will have 1023 steps f=
or the total range (1024 levels). For a +/-10V input range, the step size wi=
ll be 4.88 milli V. A 12 bit A/D will have 4095 steps and a 16 bit will have=
 65535.



If you have any=
 questions about my design, advice, or helpful information you can contact m=
e at bromberg@......... 



       It is a bit difficult providing goo=
d advice when we do not have a drawing or description of your proposed devic=
e. John Lahr has a picture on his Website, but it does not show sufficient d=
etail. Is there an online copy of your report anywhere?



       Regards,



       Chris Chapman