<<<<<<<<<<<<<<<<--->>>>>>>>>>>>>>>>
          <<  Lancair Builders' Mail List  >>
          <<<<<<<<<<<<<<<<--->>>>>>>>>>>>>>>>
>>
I was a little confused by the posting regarding the accuracy of terrain data and
GPS position information. It seemed to saying that not having terrain data was a
good thing because the data may not be completely accurate. Bullshit!

In order for the earth's surface to be displayed from a manageable amount of
digital data certain compromises must be made in the absolute accuracy of the
displayed surface. If you imagine that the surface of a mountain is represented by
a series of points then you can see that the more points (data) you have the more
accurately the true shape of the mountain is represented. Even with a relatively
few points, drawing an irregular polygonal surface that is accurate to the true
shape of a mountain within +-100 feet is possible. The problem comes in having a
processor and graphics system with enough heuvos to move the terrain around on the
screen in real time.

When someone says: "Terrain data are currently available and can easily be
incorporated into moving map displays. However, existing terrain data is not
accurate and can  be seriously misleading. The problem is further exacerbated by
GPS error  introduced by the military. Because of the errors, most moving map
display manufacturers have prudently decided to delay displaying the terrain data
until it is reliable."
What they are really saying, without the marketing BS, is:" Sure we can display
terrain data but when we do it brings our hardware to it's knees and during a
standard rate turn the display lags tens of degrees behind the airplane and is
confusing to the pilot, so we don't do it. Other manufacturers agree it isn't
prudent because they can't do it either."

As I recall the introduced (in)accuracy in the GPS signal is on the order of 50
feet.

Inaccuracies due to velocity verses refresh are easy to calculate. If a display is
updating with terrain 4 times a second and you are turning 3 degrees a second then
the terrain will lag the aircraft by 0.75 degrees. If you are traveling at 250 Kts
(422 ft/sec) then the display will lag your true position by 106 feet.

Adding it up you can see that you will know where the boundary between terra firma
and the clear blue yonder is to within 250 feet. Your options are therefore:
1) Look out the window (infinite inaccuracy in IMC)
2) Triangulate your position using VORs or lat longs off the GPS and find that
point on a chart, find the MEA and check your altimeter (BTY that process took at
least 10 seconds so have traveled almost a mile in that time).
3) Have a GPS based terrain system and know where you are within 250 feet.

Even if the accuracy for the terrain system was 10 times worse (2500 feet) it is
still better than 1 or 2 for horizontal navigation.

I recently had the opportunity to fly the Sierra Flight System's EFIS 2000 in a
real plane around real big mountains (Jackson Hole WY) with terrain (and radio
towers and airports and approach fixes and highway in the sky and velocity vectors
and airdata) displayed. I can tell you from first hand experience that I if anyone
tells you that terrain data is "not a prudent thing to have" then they haven't
flown behind the SFS EFIS 2000. I want one and I want it NOW!! So should you.

What pleases me most about all this is that a couple of smart guys (experimental
airplane builders no less) were able to do what NASA, Collins, Honeywell and a
cadre of AGATE ticks has yet to achieve. Saying that "it isn't prudent to do until
WE have done it" is just the voice of envy.

This is a very dynamic time in the avionics world. It is also the most confusing.
The technology is changing rapidly and the bull is unbelievable deep. Assume all
sales people are liars and you must fly it before you buy it. Don't accept the
sales pitch, brochure or canned demo as a basis for making a purchase decision.

Remember that everything I say should be discounted as the ranting's of a fool.
Check it out for yourself or the grater fool be you.

Regards
Brent

>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>
LML homepage:   http://www.olsusa.com/Users/Mkaye/maillist.html