Re: "Exponential" differential via mechanics

David J. Gall


I never meant to launch a personal attack against you. In my world view,
disagreeing with you does not constitute such, but challenging your
character does. I do not believe that I maligned your character, but if I
did, I apologize and welcome instruction in that which crossed over into

To clarify: My position on the tailcone belcrank is that it is a better
solution than merely splitting the cables somewhere mid-fuselage. Granted,
ideally the rudder belhorn IS the tailcone belcrank, but that puts all the
connecting goodies out in the airstream (=slowness). Besides, the QAC folks
got it exactly backwards by running the cables FIRST to the tailwheel and
then FORWARD to the rudder. Tailspring breaks and all directional control is

So, in order of evolution:

1) Use the rudder belcrank as the terminus for the rudder cables, then run
separate cables to the tailwheel. Much better (as you have done). However,
the geometry is not favorable for the tailwheel cables/springs and as I
said, all the slowness leaks out of the fuselage. (Though I do believe that
the lower rudder hinge IS sufficiently robust for such duty.)

2) Put the slowness inside the fuselage where it belongs. But there's the
rub; simply splitting the cables and running one part to the rudder and one
part to the tailwheel invites various minor and major catastrophes, from
rudder flutter to inability to separate the ratios of the control inputs to
each (wheel vs. rudder) to loss of directional control if the tailspring
breaks (yes, it could happen depending on how the whole thing is arranged).

3) A solution to the problems in 2) is to run separate cables all the way
from the rudder pedals to the rudder and tailwheel. Different attach points
on the rudder pedals can accommodate the need for different ratios, the
tailwheel springs can go anywhere, all is good in the world. This might be
the simplest, most austere, most elegant solution available (how come nobody
promotes this one???!)...


4) There are people who already HAVE a tailcone belcrank. These people are
in the forefront of defining a pseudo-standard for what constitutes an
acceptable Quickie vis-a-vis the JB6Pack. I've fought good and hard to get
my alignment suggestions recognized as psrt of the 6-Pack because I
recognize that you can never win in a contest against superior marketing
savvy. Not to mention that they are RIGHT (for the most part). Likewise, why
fight success by arguing against the tailcone belcrank. These people HAVE
the belcrank, they BELIEVE in it, and it is part of the GOSPEL of how to
make an acceptable airplane. And it works. You might disagree on engineering
grounds, but the marketing machine is what will eventually lift these planes
out of the dustbin of history and bring the resale value up to something
above sub-par. So what if the solution isn't the most pure one possible? In
fact, that might be an advantage because it is IDENTIFIABLE as so VERY
different from the original. As an engineer you know that there is no
optimum solution, only solutions (PLURAL), so pick one. So long as the
solution has been decreed by the marketing department and is justifiably a
bona fide solution, then allow the engineers to massage it into a more
useful, more elegant, more functional solution than the rudimentary initial
device may have been. What's the harm?

5) These planes were meant to be built by avarage guys with little more than
a high school education. The subtleties of design may escape them, but they
can identify that a part is or is not built in accordance with "the
drawing." Sometimes "complicating" the design actually makes it simpler to
build or just to understand. Sometimes maybe people don't want to dig under
the panel to extract their rudder pedals and get new parts welded to them,
so they find it easier to just add a new belcrank and cut the cables to

6) Since you don't have a fuselage split, you may not appreciate the
perceived benefit of having the rudder and tailwheel cables all captured
right there in the mouth of the tailcone. It may not actually be a benefit,
but it is perceived as one.

So, Peter, you see, I'm not at all opposed to what you've done or your
opposition to what others are doing. I'm just greatly in favor of some kind
of standardization such as we would have had if a man of Burt Rutan's
stature had been at the helm of the Quickie Aircraft Corporation. Absent
that leadership, I'll take the JB6Pack Cheerleaders Council as defacto
leadership if only through sheer vocal determination (not that they're wide
of the mark, either) in order to establish some sort of recognizable parity
to yank this fine airplane design out of the doldrums of its well-deserved
reputation. There's no reason for the general consensus among homebuilders
to be to sneer at the world's most efficient two seat airplane just because
it lands like snot on sandpaper. Give it a Kleenex and forget the quibbles
over design economy and eloquence. Fix it, put a happy face on it, and make
it friendly again.

As for your Norton, I was only trying to draw a parallel, noting that you
have "abandoned" the elegance of few moving parts and rotary motion in favor
of a well-polished but old-school design with lots of reciprocating bits, in
analogy to the high parts count and "fail point" count of the "un-needed"
belcrank that we continue to try to refine....

As for Jabiru, well, I've spent enough time under their hood already to know
that that little jewel of an engine is being mistreated by the factory
personnel, whoever they are, who are "advising" the design of cowlings for
the various airplanes, including Jabiru's own, that house the sweet little
engines. The plenum design they've recommended is poorly executed, leaks,
and leaves the crankcase soaking in uncirculated cylinder waste heat. If
you've copied anything like the factory cowl on a Jabiru or Thorpedo, then I
think you'll be buying new crankshaft and camshaft bearings sooner than you
should have to. Evidence: the "loosening up" period recommended by the
factory. Evidence: Five minutes at full throttle in a J230 leads to
excessive oil temps in level flight. Evidence: ground operations lead to
high cylinder head temps on #6 under the stock plenums. Evidence: what the
heck are all those cooling outlets for on a Thorpedo??! I won't join the
battle, though, because they have marketing muscle and market penetration on
their side. And I don't have time. But for you, friend, I'll make time to
hint that perhaps you should investigate the possibility that you could be
cooking your expensive investment into scrap aluminum under the factory's

Cheers, Peter, and carry on. Never mind all this ruckus Stateside.

David J. Gall
P.S. No, I haven't taxied a Q and I don't think I need to. I joined this
discussion because another pilot/engineer had already determined that the
problem existed and had identified the correct solution. His was a question
of implementation. You, however, have questioned the original assertion that
there is even a problem to address. Granted, questioning the assumptions is
what good engineers and armchair quarterbacks do, but all this "diverting
tedious, complex splitting of hairs" has actually been by request. The
necessity thereof was never at issue.

-----Original Message-----
From: Q-LIST@... [mailto:Q-LIST@...]
On Behalf Of Peter Harris
Sent: Tuesday, October 24, 2006 6:30 PM
To: Q-LIST@...
Subject: RE: [Q-LIST] "Exponential" differential via mechanics

Hi David,

I apologise if my posts on this subject have seemed a bit
negative. I should be the last one to discourage experiment
and design as it is my main interest and

I am always impressed by the thorough and highly informed
approach you take, and the mod to the pedals is a simpler
design. But it is just this additional pivoting bellcrank
differential thingo that you are on about that makes me have
to raise my hand to present a different point of view.

1. At high speed on touchdown any missaplication of
steering is accommodated by flex and skid. Crude but very
simple and effective.

2. At intermediate to low speed the plans steering
ratio works fine
(for me )

With a limited amount of experience all of this is imprinted
in the brain stem, just like riding a bike. It is automatic
and cannot be described.

My message is simplificate, simplificate. !

I am not sure if you have flown, landed or fast taxied a Q on
gravel, tarmac or concrete?. It is very good exercise for the
brain stem.

Trouble is now a whole bunch of people are out on a diverting
tedious, complex hair splitting exercise when I really do
think that it is unnecessary.

Just my humble opinion David, but from my experience the
simplest way is usually the best.

My experience with the Norton taught me very nicely how to

And if you are going to start throwing stones at Jabiru I
think you should join their list at jabiruengines@yahoogroups
so you will be better informed.

It would be better not to resort to personal attacks.



From: Q-LIST@... [mailto:Q-LIST@...]
On Behalf Of David J. Gall
Sent: Tuesday, 24 October 2006 7:42 PM
To: Q-LIST@...
Subject: RE: [Q-LIST] "Exponential" differential via mechanics


Allow me to rebut:

Part of the reason it skids is that the plans ratio moves it
too much. At higher speeds, the tailwheel is quickly and
easily pushed past the limiting slip angle and it begins to
skid. Like an all-flying tailplane that has too much throw,
the pilot can push it right past "stall" and it becomes less
effective than it could be were it not stalled. Likewise, the
plans ratio allows the rudder control system to push the
tailwheel right on past the skid limit of the tire (between 3
and 7 degrees, depending on tire type) when the pilot applies
"a little" rudder pressure.

At lower speeds, when the tail can respond and a turning
radius can be accomodated, the amount of deflection can be
more without exceeding the limiting slip angle. So it makes
sense to have a differential system. At higher speeds the
smaller throw near center helps to keep from skidding the
tailwheel, actually increasing effectiveness, and at lower
speeds the large throw needed for ramp maneuvering is still available.

If you think it is a lot of complication, consider that your
airplane actually has the opposite, a DEcreasing
differential, due to the installed angle of the rudder pedals
and the absence of any thought given to the design. The
cables simply attach to the sides of the rudder pedals, so as
the rudder pedal is pressed forward the amount of linear pull
on the cable actually diminishes for increasing angular
displacement of the rudder pedal.

Lack of thought does not imply simplicity; more thoughtful
design does not imply increased complication. I gave a
perfectly valid suggestion that increases the "fail point"
count by exactly ZERO while reversing the bass-ackwards
differential that the fine folks at QAC gave you. Simply
change the shape of the one-piece rudder pedal so that the
cable attach point is aft of the hinge axis. This one change
will give INcreasing differential without all the
monkey-motion of a belcrank, if you wish, while still
addressing the fundamental issue of an ill-executed design.

I'd wager that there aren't many airplane designers who
actually take the time to think about the linkages they
create in their control systems, and we consumers/pilots pay
the price every day. Burt Rutan is not immune from mess-ups
in his designs, and having Tom Jewett do the detail design
work on the Quickie was no guarantee of error detection and
correction. Gary LeGare (the plumber) scaling up the design
to two-place certainly didn't add any particular expertise in
the realm of control system design and the ensuing redesign
for mass production was not about refinement, either. So you
have a minimalist system that had no real thought given to it
adopted in toto and you now wish to defend it on the basis of
its "simplicity" and limited number of "fail points"? Hello?
It doesn't work right, what about that?

The design itself IS a fail point, witness its failure to
prevent the tailwheel from skidding during the fast part of
the ground run. Were it not for the many, more serious design
issues missed by the QAC, we would have been on about this
one, say, twenty years ago, but it has taken us this long to
synthesize a consensus on the JB6Pack to where we can
actually start to talk about fine-tuning. If that warrants
poo-pooing, then so be it. Some things truly aren't worth
persuing; this one is, and I knew it in 1997 when I first
published my thoughts on the matter.

Carry on, O gravel-runway reflexor-less single-data-point friend....

David J. Gall
P.S. And your Norton Rotary wasn't a lot of complication for
dubious results...?
P.P.S. When are Jabiru going to finally hire an internal-flow
consultant to design proper cooling for their engines instead
of telling owners to burn 'em in for 50-100 hrs until they
"loosen up"? Some of the nicest engines, some of the most
pitiful cowls.... :(

Join to automatically receive all group messages.