Path: utzoo!utgpu!water!watmath!clyde!att!osu-cis!tut.cis.ohio-state.edu!mailrus!iuvax!silver!commgrp
From: commgrp@silver.bacs.indiana.edu
Newsgroups: sci.electronics
Subject: Re: Accelerometers on a chip
Message-ID: <7200005@silver>
Date: 3 Aug 88 15:27:00 GMT
References: <740@esl.UUCP>
Organization: Indiana University CSCI, Bloomington
Lines: 62
Nf-ID: #R:esl.UUCP:-74000:silver:7200005:000:2508
Nf-From: silver.bacs.indiana.edu!commgrp    Aug  3 10:27:00 1988




>    I remember reading recently (in the last 3 months) about a 
>design of a small chip based accelerometer.  The principle was 
>that of a variable capacitor, where the movable plate was 
>actually a piece of deposited metal that overhung another metal 
>well.
>
>        ^
>        | Acceleration causes bending of the metal overhang.
>        v
>__________    __________________________/
>XXXXXXXXXX\___________________XXXXXXXXXX
>XXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXX
>
>The amount of acceleration is detected as a change in the 
>capacitance between the plates of the air gap as the metal 
>overhang bends...

>Some of my friends and myself are interested in producing small
>scale seismographs to monitor earthquakes in the California area.
>We currently have two (2) classical mechanical seismographs.  But
>now we are looking into the possibility of using single chip
>accelerometers as the sensing elements in a network of seismographs.

>Thanks
>-- 
>Richard W. Webb                           ecvax!decwrl!borealis!\


I don't know where to get silicon accelerometers; they're 
probably very expensive. I once worked with silicon strain-
gauges, which are extremely sensitive but have enormous 
temperature coefficients.

Seismographs usually require a LARGE mass to detect acceleration 
at extremely low frequencies.  An electronic accelerometer could 
be made using a mass attached to an LVDT (linear variable 
differential transformer), which is a transformer in a bridge 
configuration which is unbalanced by slight movements of its 
core. The associated electronics can be complex but there's a New 
Chip made especially for LVDT interface; the chip includes the AC 
exciter. I can't remember its number or manufacturer; Don 
Lancaster told me about it last month when I met him at the 
National Speleological Society convention. He will describe it in 
an upcoming "Hardware Hacker" column in _Radio Electronics_ 
magazine. Don says this chip should also make a nice bridge-
balance-type metal detector.

Another LVDT configuration is the "force-balance" transducer:  
Displacement of the LVDT core produces an error signal for a 
negative-feedback circuit which drives a DC solenoid which 
opposes the displacement, driving the LVDT back to center. The DC 
output is proportional to the applied force. This system is used 
in electronic scales.  A friend who used to work with these says 
that LVDT's are easy to build; I don't know the details.

--

Frank
reid@gold.bacs.indiana.edu