autox
[Top] [All Lists]

RE: Datalogger Story (now shorter)

To: "Jay Mitchell" <jemitchell@compuserve.com>, "Jim Carr" <jac@scri.fsu.edu>
Subject: RE: Datalogger Story (now shorter)
From: "Brent DeWitt" <bdewitt@ix.netcom.com>
Date: Sun, 8 Aug 1999 22:42:57 -0600
Hi all,

I've laid out of this for a while, but I thought now was a good time to
pitch in.  I spent a good deal of time designing and implementing an
acceleration logging system, as well as collecting real data, so here goes
from a very opinionated state of mind:

First, the third accelerometer is of marginal value for auto-x.  Check out
the equations.  If you have _only_ roll or _only_ pitch, it's great.  This
is what Byron does in his "G-Dyno" or "G-Latdyno" modes.  Assume the
unmeasured axis is steady state and calculate the others.  In "real life
auto-x", there are six degrees of freedom; x,y and z acceleration
(displacement) and pitch, roll and yaw.  Z acceleration, roll and pitch are
small factors, adding only tenths of Gs.  Yaw is a biggie for those of us
who don't keep all four behaving to the laws of static friction!  Jay is
quite correct, you need not only three axes of accelerometer, but three axes
of gyro to complete the equation.

With that said, I wish to express my opinion about Geez!.  First, I'm
biased.  Until this year, Byron was a Colorado local and a great resource
for me as a new driver.  Divorced of the Geez logger and system,  I respect
his thoughts and talent as one of those folks who have earned their black
and red jackets.

Now about Geez!  I'm an engineer (ok, they have a 12 step program for that),
and I did the typical engineer thing of diving straight in without checking
out so much as the built in tutorial.  I immediately started embarrassing
myself by asking/complaining about Geez issues to Byron.  In my defense,
there were a couple of _minor_ issues (don't make the map 0% of the
display), but the vast majority of the problems were due to me diving in
expecting the software to work like the software I had written for my own
hardware.  If you are using Geez!, run through the tutorial start to finish!
Even with my typical "over drive it till it melts" style,  I have been able
to construct amazingly life like track maps with little effort (just get the
"end speed" bit right).  I still think Byron has that wording wrong though.
It should be "end acceleration".  You really can't measure speed in Gs.

Bottom line, it works well.  I've used both a laptop and a Palm III as the
loggers and the "autostart" feature is great.

BIG NOTE!!!!!

For those of you who haven't used the Palm software:  Geez! does not
currently support data display on the Palm!  Some people seem to have
inferred that this is available.  In defense of Byron and Geez!, they have
never made this claim and it remains on the "wish list" for the future.  As
a Palm programmer I can say that the graphic intensive nature of Geez! would
swamp the  poor Dragonball 68000 in the Palm if you wanted anything close to
"real time" playback.

Sorry for the long post, but consider it the summation of five to ten
shorter ones (digest version?).

Best regards to all,

Brent DeWitt

> -----Original Message-----
> From: owner-autox@autox.team.net [mailto:owner-autox@autox.team.net]On
> Behalf Of Jay Mitchell
> Sent: Sunday, August 08, 1999 10:02 PM
> To: Jim Carr
> Cc: autox@autox.team.net
> Subject: Re: Datalogger Story (now shorter)

> There are two sets of alignment issues: 1) accelerometer axes with
> vehicle axes, and 2) vehicle axes relative to level. The g-analyst's
> calibration procedure takes care of the first one, and three
> accelerometers are required to accomplish this. I don't believe you can
> perfectly account internally for pitch, roll, or yaw with only linear
> accelerometers.
>
> Jay
>
>


<Prev in Thread] Current Thread [Next in Thread>