Hello all,

Please explain what a "wide-band" seismo is. Is it 20Hz?
100Hz? I still don't see where there a necessity to resolve
time intervals of tens of nanoseconds..... Analog sensors
such as LVDTs do work well, but it is extremely difficult to

get one with 15-20 bits of resolution.  Look at your analog
seismo
signal amplifier's output with the seismo locked  at  zero
mechanically. Use a decent o'scope aand compare your normal
full-scale voltage with what is there at rest.  If it were 1
volt FS and
you saw 2 millivolts of combined noise, then it will be
nearly
impossible to resolve better than 2  parts out of 1000,
which
is  just under 9 bits of resolution. This is a major
limiting factor
in analog circuitry.  All the CCD does is get you directly
from
seismic mass motion to a digital word representing the
DISPLACEMENT from rest of the mass, as opposed to the
coil/magnet approach whose response  is a function of
the VELOCITY of the mass, not its' absolute mechanical
displacement. The digital count value is all but noise
immune, so the CCD resolution is really there and not lost
in low-level amplifier noise. You also no longer need any
A/D converter
chip to get the displacement signal into a PC-it is already
in
binary form.
No matter how good a seismo you have, there is a limit
to real resolution and dynamic range using simlpe analog
circuits. Not impossible, but extremely difficult and
expensive.
The CCD approach was offered only as food for thought, not
to imply that we should all toss out  our analog stuff. I
would
never want to see that happen...................

DL


ChrisAtUpw@....... wrote:

> In a message dated 03/06/01, blottobear@.......... writes:
>
>
>
>> I am used to dealing with teraohms, femtowatts, and
>> picoseconds in my long
>> and
>> illustrious dealings with those magical electrons - but
>> how did an analog
>> signal with <10hz or so of bandwidth get involved with
>> those fleeting
>> nanoseconds?
>
>       It is the bandwidth of the seismometer signal that
> we want to record...
>
>
>> The antique seismos used a moving mirror and an
>> INCOHERENT light source and
>> wrote on a rotating drum located in a small vault-not
>> hundreds of feet
>> away! The trace on the film was 1-2 inches or so for a
>> typical (M7-9) large
>> quake.
>> Yea, I sing the praises of the CCD, for it will set you
>> free!!
>
> Dear Dave,
>
>       The antique seismometers used a large coil to
> generate a signal which
> was then input into a sensitive galvanometer. A mirror on
> the galvo coil
> reflected a beam of light. You could read that out with a
> CCD, but there is
> currently more interest in wide band instruments with
> direct readouts. A
> reluctance bridge or a LVDT do this very nicely, thank
> you. So do tell us
> just how you plan to set us free with CCD?!
>
>       Regards,
>
>       Chris Chapman


Hello all,
Please explain what a "wide-band" seismo is. Is it 20Hz?

100Hz? I still don't see where there a necessity to resolve

time intervals of tens of nanoseconds..... Analog sensors

such as LVDTs do work well, but it is extremely difficult to

get one with 15-20 bits of resolution.  Look at your analog seismo

signal amplifier's output with the seismo locked  at  zero

mechanically. Use a decent o'scope aand compare your normal

full-scale voltage with what is there at rest.  If it were 1 volt
FS and

you saw 2 millivolts of combined noise, then it will be nearly

impossible to resolve better than 2  parts out of 1000, which

is  just under 9 bits of resolution. This is a major limiting
factor

in analog circuitry.  All the CCD does is get you directly from

seismic mass motion to a digital word representing the

DISPLACEMENT from rest of the mass, as opposed to the

coil/magnet approach whose response  is a function of

the VELOCITY of the mass, not its' absolute mechanical

displacement. The digital count value is all but noise

immune, so the CCD resolution is really there and not lost

in low-level amplifier noise. You also no longer need any A/D converter

chip to get the displacement signal into a PC-it is already in

binary form.

No matter how good a seismo you have, there is a limit

to real resolution and dynamic range using simlpe analog

circuits. Not impossible, but extremely difficult and expensive.

The CCD approach was offered only as food for thought, not

to imply that we should all toss out  our analog stuff. I would

never want to see that happen...................
DL

 
ChrisAtUpw@....... wrote:
In a message
dated 03/06/01, blottobear@.......... writes:

 
I
am used to dealing with teraohms, femtowatts, and picoseconds in my long

and

illustrious dealings with
those magical electrons - but how did an analog

signal with <10hz or
so of bandwidth get involved with those fleeting

nanoseconds?

     
It is the bandwidth of the seismometer signal that we want to record...

 
The
antique seismos used a moving mirror and an INCOHERENT light source and

wrote on a rotating drum
located in a small vault-not hundreds of feet

away! The trace on the film
was 1-2 inches or so for a typical (M7-9) large

quake.

Yea, I sing the praises
of the CCD, for it will set you free!!

Dear Dave,
     
The antique seismometers used a large coil to generate a signal which

was then input into a sensitive
galvanometer. A mirror on the galvo coil

reflected a beam of light.
You could read that out with a CCD, but there is

currently more interest
in wide band instruments with direct readouts. A

reluctance bridge or a LVDT
do this very nicely, thank you. So do tell us

just how you plan to set
us free with CCD?!
     
Regards,
     
Chris Chapman