View Single Post
  #3  
Old March 31st 04, 03:41 PM
James M. Knox
external usenet poster
 
Posts: n/a
Default

(Ben Jackson) wrote in
newsQjac.138439$_w.1599417@attbi_s53:

I've been playing with one of the Analog solid state gyros. I want to
build a small combination TC + "DG" instrument (no absolute heading,
but you could think of it as automating a timed turn). A single gyro
is just *barely* stable enough (with multi-point temperature
calibration for each individual chip) to provide a heading reference
alone. It's easily stable enough to provide a TC-like display. I
think most of the commercial units you'll see will be 3-axis
gyro/accelerometer combos (like Control Vision's system) which can
self-calibrate (and for that matter automatically "erect" even without
being perfectly square to the airframe) by using the accelerometers as
a long-term stable reference.


I made one up in the lab a few years ago. Interesting experiment, but
there was no way to get it certified (without so much extra stuff that
the basic gyro would have to sell for around the $3K+ range). The
problems are pretty much what you have already figured out.

o Temperature drift.
o Integration error
o Non-static erection (no jokes, please G)

The temperature drift can be helped by characterizing each unit. Long
term integration error is the biggest problem. Most of the Control
Vision type systems "fake it" by assuming you are usually flying
straight and level. So whatever you do long enough becomes straight and
level. [Real AH's have the same problem, but at the time the FAA would
not accept "equivalent level of safety" certification.]

The non-static erection problem still exists, of course. That's why
most of the new glass cockpit systems have a requirement that the plane
remain level (and usually stationary) for several minutes after startup.
A "hiccup" in flight, where the system has to re-initialize itself, is a
real problem if you are not currently straight and level.

Still, I think it could be a big help in an emergency. Just wasn't
possible to convince the FAA.

-----------------------------------------------
James M. Knox
TriSoft ph 512-385-0316
1109-A Shady Lane fax 512-366-4331
Austin, Tx 78721

-----------------------------------------------