At 07:32 PM 9/15/2008 -0700, you wrote:
>Hi Chris
>STM also describes a fabricated coil using a rare earth ring magnet from 
>mcmaster carr, thick washer and cap screw. It was definitely easier for me 
>to make than taking apart and machining and putting back together a woofer 
>speaker.
>
>Hi Brett
>Using STM equations I have been wondering for quite a while how to lower 
>the gain to prevent clipping. It seemed the whole idea was to have a large 
>gain in the equations for the desired frequency range. I put a voltage 
>divider at the output but that won't prevent clipping of the amp before 
>the divider. Any ideas? I would like two output levels to expand my 
>sensors range. It seems recording two levels would be easier to achieve 
>than a higher resolution A/D , however it would double the number of 
>recording channels.
>

Hi Barry,

As you suggest, the circuit itself will have clipping limits, which will 
come into play when you increase the A/D clipping level.  In order to 
determine what those limits are, I'd need to know quite a bit more about 
exact circuit.  The most likely limit will be in both the voltage and 
current capabilities of the amplifier which drives the derivative branch 
capacitor, though several other factors will also relate to that.  Some of 
those are the mass value, the forcing coil constant, N/A, the coil 
resistance and just possibly its inductance, and the derivative capacitor 
value to name the most important factors.  Given those (except possibly for 
the inductance) combined with the derivative branch driver amplifier peak 
voltage and current capabilities, one should be able to plot clipping 
velocity vs frequency, which I think is what you are looking for.

You might want to consider downloading the Excel workbook 
http://bnordgren.org/seismo/loop7.zip  which has a lot of tools for 
relating component values and instrument performance.  Although it doesn't 
directly relate to clipping levels, it does provide some useful tools 
related to them.  Sorry for the lack of documentation, but there are three 
worksheets at the beginning which define what the variables mean and show a 
little of how to use the workbook.  You will want to let Excel run macros 
if you are planning to do more than just look at the example data, which is 
currently set to be an approximate model of Allan Coleman's MkXX 
design.  Please feel free to ask about anything that needs 
explaining.  Someday I'd like to write some more about how to use it to do 
a feedback design and any questions or comments will help me to do that better.

Regards,
Brett










__________________________________________________________

Public Seismic Network Mailing List (PSN-L)