[Fwd: Re: G-Tech meter acceleration calculations]

[Dan]

>Tom Cloud wrote:
 >
> > >I bought a G-Tech meter and it pretty much works.  You have to be on a
> > >flat, level road for the thing to work correctly, though.  Since it uses
> > >an accelerometer for all of its calculations, it can get thrown off
> > >pretty quickly.  For example, if you are going up a hill, the
> > >accelerometer will think you are going faster than you really are.  You
> > >could even park your car on a hill, and G-Tech will think you are
> > >accelerating continuously.  The device is also only accurate for HP,
> >
> > I'm confused .... an accelerometer is an "AC-coupled" device.
> > I.e. it only detects a "change" in velocity.  If it reads while
> > sitting still or when the speed is constant, it must be a weight
> > hung on a pot, not a 'real' accelerometer.
> >
> > Tom Cloud <cloud at peaches.ph.utexas.edu>
> 
> If you turn the GTech meter 90 degress, you measure the constant
> acceleration of gravity (1.0G).  May be strange to think of, but, on
> earth, you are always under acceleration even sitting still.  So,
> parking on a hill, gives you an acceleration if you had previously set
> your accelerometer to a level surface.  By turning the accelerometer 90
> degrees, it makes a pretty easy check to see if it is set right also
> since it has to read 1.0G's to be correct.
> 
> You can get much better results by including a vehicle speed sensor
> (like tapping into the cruise control) and using the vehicle velocity as
> an input to the HP calculations along with the accelerometer
> measurements.  That way you can eliminate the cumulative errors that
> occur using only the accelerometer.  You can also do some coast down
> tests to see drag; if you do a low speed and a high speed coast down,
> you can separate the rolling frictional from the air drag also because
> the air drag increases non-linearly while the frictional is pretty much
> linear.  Lets you figure out how your spoilers, tires and what not are
> eating your power.
> 
> I bought a two axis temperature accelerometer +/- 2.5G for front to
> back, and +/- 1.5G for side to side.  It also has a temperature sensor
> so that you can do temperature compensation since they drift some with
> temperature.  Also had 5VDC regulation so I could run it off a 9V
> battery and not have to worry about the power supply.  Whole package
> potted ran $150 from NGT, 914-223-3359 in LaGrangeville, NY.  Owner is
> George VanderGheynst (gvanderghe at aol.com) ang he was very helpful.
> Outputs are zero to 5VDC so measurement is easy.  Cheap A/D converters
> that plug into the printer port of a laptop are available from lots of
> sources for $100 or so.  Can't remember at the moment where I bought
> mine but can look for the papers if you want.
> 
> Dan