G-Tech meter

[Peter Shoebridge]

I bought some accelerometer's with the view of doing just that. The ones I
got were from Analog Devices. I was going to hook them up to the efi332
board when we get the io board going.

I'd be very interested in the calculations.

Peter
 ----
From: M HILL <EAXMJHI at ean2.mecheng.nottingham.ac.uk>
To: diy_efi at coulomb.eng.ohio-state.edu
Date: Wednesday, April 30, 1997 10:46 AM
Subject: Re: G-Tech meter

I sawthese advertised and thought that it would be simple to make up
my own.  Sure enough within a week I had a fully built and working
accelerometer.  The one that I built is programmable for the
acceleration speeds, i.e. you could time 50 -70mph or 0-100 if you
wanted.  It uses a 24 * 2 character LCD display and does maximum hold
values for acceleration and braking as well as cornering force.  The
distance calculation is done from the acceleration, the calculations
are very simple to do.   The most expensive parts for it were the
accelerometers which were u25 each but are good quality accurate
ones.  There will be problems with going up and down hills and when
you do not accelerate in a straight line.  The distance that the
measurement is performed over is usually quite small so that errors
that come about from the double integration required are quite small.
I am going to start some work on version two which will have more
features such as storing the data at various points and automatic
recalibration for accelerating round corners.  As for the
performance.  I can't speak from experience of the G-Tech meter but
mine is very good on consistency.  The timing is atarted
automatically when an acceleration of more the 0.05G is detected so
that false starts from vehicle vibrations are not a problem.  If you
are interested in this and would like some help with the calculations
I would be glad to help.  The work on the microcontroller was done on
an 87C752 which is basically an 8052 core with built in EPROM and RAM
and ADCs.  From what I've heard the G-Tech meters and accurate to
within 0.4seconds and repeatable to within 0.2 seconds.  This is from
some review in a magazine.

Martin