to HSI or not to HSI

It's true. Several years ago when we (the group that owned the Bonanza 36 I
was a member of) decided to replace our maintenance-hog Narco HSI, we chose
the King KNS-55 HSI because it worked on the electrical system only (no
vacuum). That meant that a vacuum failure left the HSI, turn coordinator,
and S-Tec autopilot working which would be pretty much a non-event.
Electrical failure would leave the backup DG (the DG that was replaced by
the original HSI was still in the panel) operating.

"Dave Butler" wrote in message
...
This quote is from the Avionics West
article at http://avionicswest.com/articles/kno..._autopilot.htm
This HSI, slaved or not, MUST have BOTH
vacuum and electrical inputs in order to operate. In other words, if you
lose vacuum or the electrical system, this HSI compass card will quit!

Can anyone confirm the statement above from Avionics West? I consider them
a
reliable source of information, but this seems counterintuitive to me. In
case
of electrical failure, why is the heading information not still available?
Is
the compass card not mechanically connected to the gyro?

Obviously, I can see that the navigation information would not be
available in
case of electric failure, but the quote specifically says the compass card
will
quit.

Thanks,

Dave

"Richard Hertz" no one@no one.com wrote
Can you explain why that is the one advantage (BC)/revers on localizer, and
why that is so?
Do you mean to say that people confuse which color sector they are in on a
localizer due to "reverse needle"?

Yes, that is exactly what I mean to say.

If so then it is a training issue, not a
technology issue.

Oh man, here we go. You've just touched off a religious debate.

In real life, I run a department that designs instrumentation for
process environments. What that means is that engineers design it,
but generally non-engineers (plant operators, meter readers,
technicians) install and use it. These days, most instrumentation has
software in it, so it should not come as a surprise that I rose into
that position from software engineering.

In the process, I learned a lot about user interfaces. There are two
kinds of user interface bug. There is the kind where the user
interface acts contrary to design, in a useless or unpredictable
manner in a given situation (coding error) and there is the kind where
it acts as designed (intentionally or unintentionally), in a manner
that is predictable and useful but, in certain situations,
counter-intuitive to the operator (design error). The first kind is
unusuable in those given situations. The second kind is usable,
provided you read the manual and are aware of how the system will
behave. There are those who believe that this means it's not an error
- that you should simply RTFM. In other words, that it is a training
issue. They are wrong.

The "reverse" indication of a conventional CDI is a design error. You
can work around it. I have. I had to shoot a LOC BC approach with
engine failure at leveloff (simulating a failure of the engine to come
up on the powerup for leveloff) followed by a single engine missed
approach. I passed - meaning I executed the approach and miss to ATP
standards, and I have the certificate to prove it. Nonetheless, a
couple of times I found myself, with the needle half a dot off,
applying the incorrect contol inputs before I "caught" myself. So
clearly the training worked - I corrected before I deviated beyond ATP
standards - but that doesn't mean that the design is correct. It's
not. On an approach, you're used to correcting towards the needle.
Under normal conditions, you should have the situational awareness to
know what you are doing, rather than just correcting by habit.
However, in emergency conditions where the workload becomes high,
there is a tendency to revert to habit. In other words, the operation
becomes counter-intuitive.

Sometimes this is unavoidable, but where this is done for no good
reason, it's simply bad design. It's really quite simple to modify a
conventional CDI for "reverse" sensing - all it takes is the addition
of a simple DPDT switch, and the needle will act correctly on the BC.
Thus I have to say it's done for no good reason. Only in aviation is
somehting like this not done - because this is how we've always done
it (and because the FAA would make such a modification
cost-prohibitive).

The UK (and I believe other nations) will not certify LOC-BC
approaches because the potential for error is believed to be too high.
I don't agree with this - I consider the potential for error to be
adequately low with proper training - but the addition of a cheap,
simple, and reliable part to the CDI (or replacement with an HSI)
eliminates the potential for error - and is thus clearly an advantage.

Michael

The decision the HSI vs. DG is a personal decision. Some people think
that a battery GPS with an electronic HSI is a good compromise to
supplement -- but of course not replace -- a panel-mounted DG. This is
because an HSI on a handheld GPS is in many ways easier to fly than a
"real" HSI because the handheld GPS shows track rather than heading. That
means wind corrections are much easier to establish with a handheld GPS HSI
than with a panel-mount HSI. It is arguable that a panel-mount DG plus
handheld GPS HSI creates a lower workload situation (with more redundancy
and lower cost) compared to a panel-mount HSI.

If you do stick with the DG, there is another change to your panel that I
would suggest. I would suggest moving your turn coordinator to the far
right of your panel and putting the electric AI in your primary field of
view (presumably where your turn coordinator is now). Flying partial panel
with an electric AI on the right side and other instrument on the left can
easily induce vertigo.

In fact, as long as your electric AI has a ball (inclinometer), you do not
even legally need to keep your turn coordinator at all.


--------------------
Richard Kaplan, CFII

www.flyimc.com

"Mike Rapoport" wrote in message
k.net...


under $2000 with AHARS, it seems that someone must be cooking up a
self-contained EHSI with internal gyro. I don't actually know of anyone
working on this but it seems unlikely that it is not being persued

It already exists, i.e. Garmin 195 or anything later.

(Yes, I know, it is not based on a gyro, has a somewhat slower response
time, and is not legal for primary navigational or attitude information.
Nonetheless, it works quite well to supplement a panel DG.)


--------------------
Richard Kaplan, CFII

www.flyimc.com

"Michael" wrote in message
om...

somehting like this not done - because this is how we've always done
it (and because the FAA would make such a modification
cost-prohibitive).

I have such a switch on my airplane -- it is part of my Cessna 400B
autopilot control head. Just flip the switch and the CDI needle reverses
its orientation.

I agree completely that this makes a back course approach easier to fly even
though "training" should be able to solve the problem otherwise.

Let me give another example to support your point. One of the most
challenging situations I occasionally try in my simulator is a demonstration
of reversed ailerons not caught by a pilot on preflight. In theory, once
the pilot realizes what happened, there is no emergency at all -- the
airplane is perfectly controllable. In reality, almost all pilots find this
to be an extremely difficult scenario, and in fact it seems as if the
higher-time the pilot the harder it is to reverse one's thinking and provide
reverse aileron input. The same is likely true when flying a localizer
back-course approach; we are all so used to "normal" sensing on a CDI needle
that our responses become so ingrained/automatic as to make it much harder
to reverse this action on rare back-course approach.

--------------------
Richard Kaplan

www.flyimc.com

Once one accepts the fact that all sensing is "normal", and flies
headings as indicated by the VOR head, instead of the more difficult
methd of flying left and right needles, back course localizer flying
is no more difficult than "ordinary" course tracking.

On Sat, 13 Nov 2004 21:07:48 -0500, "Richard Kaplan"
wrote:


"Michael" wrote in message
. com...

somehting like this not done - because this is how we've always done
it (and because the FAA would make such a modification
cost-prohibitive).

I have such a switch on my airplane -- it is part of my Cessna 400B
autopilot control head. Just flip the switch and the CDI needle reverses
its orientation.

I agree completely that this makes a back course approach easier to fly even
though "training" should be able to solve the problem otherwise.

Let me give another example to support your point. One of the most
challenging situations I occasionally try in my simulator is a demonstration
of reversed ailerons not caught by a pilot on preflight. In theory, once
the pilot realizes what happened, there is no emergency at all -- the
airplane is perfectly controllable. In reality, almost all pilots find this
to be an extremely difficult scenario, and in fact it seems as if the
higher-time the pilot the harder it is to reverse one's thinking and provide
reverse aileron input. The same is likely true when flying a localizer
back-course approach; we are all so used to "normal" sensing on a CDI needle
that our responses become so ingrained/automatic as to make it much harder
to reverse this action on rare back-course approach.

--------------------
Richard Kaplan

www.flyimc.com


d

Once one accepts the fact that all sensing is "normal", and flies
headings as indicated by the VOR head, instead of the more difficult
methd of flying left and right needles, back course localizer flying
is no more difficult than "ordinary" course tracking.

Any way you look at it -- left/right, to/from, yellow/blue -- a backcourse
approach is not a common situation and human nature is such that this means
there is more likelihood for error.


--------------------
Richard Kaplan

www.flyimc.com

Agreed.

So why then teach the solution that requires doing it the hard way?


On Sun, 14 Nov 2004 01:51:32 -0500, "Richard Kaplan"
wrote:


Once one accepts the fact that all sensing is "normal", and flies
headings as indicated by the VOR head, instead of the more difficult
methd of flying left and right needles, back course localizer flying
is no more difficult than "ordinary" course tracking.

Any way you look at it -- left/right, to/from, yellow/blue -- a backcourse
approach is not a common situation and human nature is such that this means
there is more likelihood for error.


--------------------
Richard Kaplan

www.flyimc.com

wrote in message
...


So why then teach the solution that requires doing it the hard way?

The easiest way to interpret a CDI needle 99% of the time is to think of it
in terms of "left" vs "right".

A more generalized solution which thinks in terms of sectors does indeed
make a back course easier for 1% of instrument flying but also makes the
other 99% of instrument flying less intuitive.


--------------------
Richard Kaplan

www.flyimc.com

Only if that is the way it is taught.

It is just as easy ti interpret a CDI needle in terms of headings and
course intercept angles if you are taught properly to do so.

And it is far less prone to error.

And it is consistent, which makes all the navigation, including the
"other 99%", easier and more accurate.





On Sun, 14 Nov 2004 11:16:49 -0500, "Richard Kaplan"
wrote:


wrote in message
.. .


So why then teach the solution that requires doing it the hard way?

The easiest way to interpret a CDI needle 99% of the time is to think of it
in terms of "left" vs "right".

A more generalized solution which thinks in terms of sectors does indeed
make a back course easier for 1% of instrument flying but also makes the
other 99% of instrument flying less intuitive.


--------------------
Richard Kaplan

www.flyimc.com