PSN-L Email List Message

Subject: Re: Verticle Sensor Design
From: ChrisAtUpw@.......
Date: Mon, 25 Aug 2008 01:33:07 EDT


In a message dated 2008/08/25, barry_lotz@............. writes:

> Chris
> I'm not sure what the periods are. The STM8 style ~30" long appears to have 
> a period (with feedback) of 82 sec. I get occasional spikes I think from the 
> concrete slab it rests on. When this happens I get a single cycle which has a 
> period of 82 secs. I have plans in the future to put a secondary coil near 
> the feedback coil to provide a induced force to measure the period etc. 

Hi Barry,

       Can you use a small bar magnet and just flip it through 180 degrees to 
simulate a pulse?
       I had expected the achieved period to be dictated by electronic time 
constants?

       Concrete is not a very good material for making a seismometer. 50:50 
cement and sand is likely to be much better. The problem comes due to the 
cement and the stones in the concrete having very different thermal expansion 
coefficients. 

The 12" STM8 style vertical has a background oscillation of 167 sec but this 
could 
> be from anything including electronics. The Lehman has a damped period 
> around 20 sec. I could make it longer but it's located near an exterior wall in 
> my garage and I would be spending every day adjusting its position in the 
> spring and fall. I have plans for a basement. Hopefully this will assist in 
> stability. The STM sytle verticals need slight motor adjustment every few days In 
> the spring and fall. 

       20 seconds is a fairly good period to use for a Lehman, provided it 
does not have a short arm. Getting a period extension of x10 is fairly easy, 
getting x20 is more difficult, getting anything more renders you very liable to 
tilt shifts. You can provide long period position feedback to stabilise this, 
preferably using photocells, but the system gets more complicated. What 
suspensions are you using, top and bottom?
The traditional limit comes from trying to use knife edge bearings, which 
have a generally lousy performance. I use a 1.5 second pendulum set to 30 
seconds. This is about the practical limit. I have a clamped 8 thou wire at the top, 
a SS plane rolling on a 1/2" SS ball fixed to the column, at the bottom and 
quad magnet damping.

    What is the electronic period of the 12" STM-8?

       I look froward to seeing more pictures in future!

       Regards,

       Chris Chapman   
In a me=
ssage dated 2008/08/25, barry_lotz@............. writes:

Chris
I'm not sure what the periods are. The STM8 style ~30" long appears to have=20= a period (with feedback) of 82 sec. I get occasional spikes I think from the= concrete slab it rests on. When this happens I get a single cycle which has= a period of 82 secs. I have plans in the future to put a secondary coil nea= r the feedback coil to provide a induced force to measure the period etc. <= /BLOCKQUOTE>

Hi Barry,

       Can you use a small bar magnet and just= flip it through 180 degrees to simulate a pulse?
       I had expected the achieved period to b= e dictated by electronic time constants?

       Concrete is not a very good material fo= r making a seismometer. 50:50 cement and sand is likely to be much better. T= he problem comes due to the cement and the stones in the concrete having ver= y different thermal expansion coefficients.

The 12" STM8 style vertical has a background oscillation of 167 sec but this= could

be from anything including ele= ctronics. The Lehman has a damped period around 20 sec. I could make it long= er but it's located near an exterior wall in my garage and I would be spendi= ng every day adjusting its position in the spring and fall. I have plans for= a basement. Hopefully this will assist in stability. The STM sytle vertical= s need slight motor adjustment every few days In the spring and fall.

       20 seconds is a fairly good period to=20= use for a Lehman, provided it does not have a short arm. Getting a period ex= tension of x10 is fairly easy, getting x20 is more difficult, getting anythi= ng more renders you very liable to tilt shifts. You can provide long period=20= position feedback to stabilise this, preferably using photocells, but the sy= stem gets more complicated. What suspensions are you using, top and bottom?<= BR> The traditional limit comes from trying to use knife edge bearings, which h= ave a generally lousy performance. I use a 1.5 second pendulum set to 30 sec= onds. This is about the practical limit. I have a clamped 8 thou wire at the= top, a SS plane rolling on a 1/2" SS ball fixed to the column, at the botto= m and quad magnet damping.

    What is the electronic period of the 12" STM-8?

       I look froward to seeing more pictures=20= in future!

       Regards,

       Chris Chapman

[ Top ] [ Back ] [ Home Page ]