Newsgroups: comp.robotics
Path: cantaloupe.srv.cs.cmu.edu!das-news2.harvard.edu!news2.near.net!howland.reston.ans.net!ix.netcom.com!netcom.com!nagle
From: nagle@netcom.com (John Nagle)
Subject: Re: More Gyro details. (Was Inertial Gyro Specs)
Message-ID: <nagleCzoJ12.IDA@netcom.com>
Keywords: Gyros
Organization: NETCOM On-line Communication Services (408 261-4700 guest)
References: <calvinCznKqw.A9C@netcom.com>
Date: Tue, 22 Nov 1994 17:24:38 GMT
Lines: 75

calvin@netcom.com (John Calvin) writes:
>Greetings again.
>I've had a suprising number of inquires regarding the $2.00 Gyro's I 
>located in Haltek in Mountain View Ca. (Phone (415)-969-0510).  

      Here's the spec on these things.

THE ETAK/BOSCH RATE GYRO

	The ETAK gyro is a six-terminal device.  Pinout is as follows:

          1 (clear)	unused
          2 (Brown)	+5V, 50mA   
          3 (Yellow)	GYRO YAW (5v digital pulse output)
          4 (Green)	GND
          5 (Red)	PITCH INCLINOMETER  (0..5v analog output)
          6 (Black)	ROLL INCLINOMETER  (0..5v analog output)
          7 (Orange)	GYRO PITCH (0..5v analog output)
          8 (Blue)	unused

The colors apply to the cable from the compass.  These colors can be seen
inside the transparent modular connector.  The pin numbers above read
from right to left viewing the plug from the contact side.

	There are thus three analog outputs and one pulse train.  The
frequency of the pulse train varies from 0 to 70KhZ.  The pulse train
is a strange beast.  It is not a variable frequency; it is a fixed
frequency pulse train with some pulses missing.  Counting the pulses
produces an angular rate.

If the rate limits of the gyro are exceeded, the pulse train output
will appear as a solid 1 (for one direction) or a solid 0 (for the
other).  

The pulse train output is an open-collector TTL output, and requires
a pull-up resistor or other suitable input.  1M appears to work.

      The terminology above assumes that the long axis of the gyro is
aligned with the vehicle's long axis.  (ETAK's drawings are labelled
differently, assuming a different mounting).

      Physically, the gyro/inclinometer unit is a box with a circular
projection housing the motor on one end.  A screw hole for mounting
appears on the bottom.  The motor housing should face fore or aft, (we
can compensate in the software) and the bottom should face down.  
The inclinometers actually appear to work with the unit inverted,
but the design is intended for use in the standard orientation.

      Internally, the gyro is simply a motor spinning a thin plate
which flexes as the gyro is turned.  The deflection of the plate is
measured capacitively at two points, producing two angular rates.
The inclinometers are a cylinder of liquid whose position is, again,
sensed capacitively.

      The inclinometers are good to about 1 percent accuracy and
produce meaningful values only for rotations under 45 degrees or so.
Response of the inclinometers is about 1Hz.  The poise of the liquid
used in the inclinometer has been carefully chosen to obtain good
response under vibration.  If a 1Hz low-pass filter is used on the
inclinometer outputs, the output values should be good, although they
will, of course, lag the current situation slightly.

      The gyro has some serious limitations.  At best, it is only
about 1% accurate.  This degrades sharply under some conditions.  The
unit is temperature sensitive, enough so that holding the unit in hand
will induce errors.  (This may improve in later units).  The unit is
limited to angular rates of about 45 to 60 degrees per second.
(This, too, may improve in later units.)

     This specification applies to board 03-1204-02 rev X2, which is a 
prototype version dated 8 FEB 1988. Production boards may be slightly 
different.

					John Nagle

