Roll steering question

There seem to be two ways to achieve roll steering, also known as
GPSS.

1) The GPS (which contains the planned route, or the IAP waypoints
etc) generates the turning parameters internally, and transmits them
directly to the autopilot; typically using ARINC. In this case, the
autopilot is in NAV mode to get GPSS.

2) There is an add-on box, which takes the data stream coming out of
the GPS (which contains the route, as in 1) above) and this box works
out when to turn. It then generates a faked heading bug signal (which
I believe is a quadrature sine/cosine signal - what does a normal HSI
output for the heading bug setting?). In this case, the autopilot is
in HDG mode to get GPSS.

Method 2) can be used with a GPS, or with an autopilot, if the two
won't talk to each other. But it still needs the add-on box to be fed
with the currently active waypoint details, and the one that follows
it, otherwise how could it calculate the turn?

Is the above description correct? Are there any other variations?

Next Q is: How does GPSS happen if you have an EHSI? Perhaps this is
method 3). In this case, the EHSI course pointer flips automatically
at each waypoint, to the new track. But something somewhere has to do
a turn calculation, otherwise the autopilot would not do a smooth turn
onto the new track. In this case, does the EHSI work like the add-on
box in case 2) and the course pointer flipping to the new track is not
related to the info fed to the autopilot? If this is true, the AP
would be in HDG mode with an EHSI but I believe this is wrong; NAV
mode is used. Maybe the EHSI takes the route data from the GPS, does
the turn calculation, and sends the turning data to the (compatible)
AP using ARINC, as in case 1) above.. ??

"bill_hale@verigy.com" <33bonanza@**********> wrote

>The Sandel and King EHSI simulate the GPSS function by having the
>course
>pointer always slew around to the desired track as the flight plan
>advances.
>
>The autopilot uses
>that course arrow as its basic raw heading info, modifying it by left-
>right needle
>indications in NAV or APR modes.

I am trying to work out how the operation of an EHSI differs from the
other methods.

I am aware of course pointer flips round to the next track, but this
can hardly be used alone by the autopilot, to get accurate turns
especially on approaches.

You mention that the AP uses the CP value modified by the track error,
but how does this work accurately? Does the EHSI somehow generate
artificial track error values to get the AP to correctly fly fly-by or
fly-through waypoints?

"John Collins" <johncollins@carolina.*******> wrote

>Some GPS'es don't have ARINC 429 outputs but provide the same data on a
>serial bus. Most of the GPSS converters support both the serial bus and
>ARINC 429 bus, but not all do. For the KLN94 GPS it outputs an analog
>left/right steering signal for the autopilot to use, but its performance is
>marginal at best.

Thank you for the details - very useful.

The KLN94 (which I have) seems to have an ARINC output, enabled by a
later firmware. It then talks direct to a KFC225 for achieve GPSS. I
haven't yet got this connection made but will have soon. From other
pilots I know, it seems to work OK, correctly anticipating turns.

The software is 'mod 01/02' available around 2005.