Sender: <marv@lancaironline.net>
To: lml@lancaironline.net
Date: Mon, 14 Jan 2008 12:46:58 -0500
Message-ID: <redirect-2644687@logan.com>
Received-SPF: none
 receiver=logan.com; client-ip=66.150.29.112; envelope-from=n103md@yahoo.com
Subject: Re: [LML] Pitch Servo Connection
From: "bob mackey" <n103md@yahoo.com>
User-Agent: SquirrelMail/1.4.9a
MIME-Version: 1.0
Content-Type: text/plain;charset=iso-8859-1
Content-Transfer-Encoding: 8bit
Importance: Normal

Don Skeele wrote:

>      However you attach the pitch servo rod to the elevator push rod,
> try to eliminate any "slop' or "dead area".  My 235/320 is very sensitive
> in pitch control, so if you allow the push rods to rotate when control
> force is applied, this presents a dead area to the servo/controller,so
> rotate all the rod-end bearngs in one direction to minimise "slop", I
> attached the servo rod end bearing to the elevator pushrod with a bolt
> thru the servo rod end bearing, which was bent 90 deg to parallel the
> elevator push rod and the used two hose clamps around both to hold it
> parallel and tight to the elevator push rod.

While I agree that there should be no slop in the control system,
I don't think clocking the rod end bearings is the right way to
go about removing slop. The distance between the ends of the pushrods
is designed to be independent of bearing rotation, so rotating the
bearing does not contribute to slop. If the servo's action is
causing the elevator pushrod to rotate on its axis, the problem is
that the servo is not located ideally. The servo pushrod should
be very nearly parallel with the elevator pushrod, and the arm
off the side of the elevator pushrod should be in the plane of the
two pushrods. That way, the elevator position is unaffected by
rotation of any bearing around its pushrod axis. The servo motion
does not apply any torque to the pushrods, only pure tension/compression.
If the bearings are defeated by clocking them to take out rotational
slop, then they can bind.


>  The whole trick to good control is to mimic the pitch servos
> response to the A/C's control/altitude changes and with my
> S-Tec 50 it has been a battle. For the most part it is too
> sensitive, maintaining absolute altitude control to the foot
> (almost) but yanking us humans up and down rather uncomfortably.

I wonder whether the stiffness of the servo has defeated the bob weight's
damping effect? The bob weight acts to both increase static stick force
per G, and also to dynamically oppose transient turbulence bumps.
If the servo connection to the elevator pushrod is too stiff, it might
be preventing the bob weight from deflecting the elevator to (partially)
neutralize turbulence.

I think the bob weight is largely responsible for the much smoother
ride in the Lancair 235/320/360 than in a Glasair. Of course there
are other differences as well.

You might try adding some spring to the servo control rod portion of
the system. Some care should be taken to insure that the spring/bob weight
resonance is not near the frequency response limit of the autopilot.
Obviously this should be well-tested below Va before venturing to higher
speeds with the autopilot engaged.

BTW, the S-Tec 50 has internal accelerometers for "short term pitch axis
stabilization".  That would also interact with the bob weight and
any springiness in the system.