As everyone knows, I'm working out the cockpit controls on my DrI. I've figured out how to have the servo drive the rudder bar which in turn acts as the bellcrank for the rudder cables. I can also see how to connect the elevator servo to the joystick to have that move the elevator cables. And at the moment, I have the 2-axis movement on the stick which would allow scale aileron control, if I wanted to go with a full closed-loop system up to and through the wings.

BUT, is this simply impossible to achieve mechanically? The engineering problem, as I see it, is that while it would be simple enough to hook one servo to the stick for front to back motion and another servo to the shaft on the floor for side to side rotational motion, the interaction between these two motions on the stick/torsion bar would mess with the length(s) of the servo linkage(s).

It there a mechanical solution for this? Seems like something a good Yankee engineer from the 1880's could have figured out!

Perdeval has the idea sorted on his Tiger moth.

http://www.rcuniverse.com/forum/m_90...page_2/tm.htm#

While the moving control column & rudder bar idea is cute perhaps the more common sense idea would be to mount the servos upside down & lower, run the pushrods under a false floor & mount the belcranks behind the pilot's seat. - John.

I try to avoid common sense whenever possible! [ ]In terms of the (scale) rudder bar, I don't see any reason why it would be any more sensible to use a longer linkage to connect to a non-scale bellcrank further behind the CoG. The rudder bar IS just a more elaborate bellcrank. Ditto on using the control stick as an (rather more decorative) vertical bellcrank. Would it be any different if I were to install a traditional model bellcrank in that location (such as the one on my Snipe)? I don't really see any advantage in that.

But I really don't see any way to include scale rotation of the aileron lever bar without disconnecting this from the elevator "bellcrank" movement. That is I could have a mock bar the does in fact rotate (for example driven by an arm to a servo on the side) as long as it didn't pass really through the elevator control stick. But then the stick wouldn't move either.

I hadn't really planned to have scale aileron control, for a host of practical reasons, but I'm very curious whether 2-axis servo control is possible.

You know, I think I may have an idea! It's poorly formed and only hazy at the moment, but it involved a movable cuff passing through the control column. The cuff would not rotate, when the aileron control bar rotates.

Nevermind.

In terms of the rudder bar and elevator, what I'm doing is really nothing different mechanically from the way Summerwind did the bellcranks on his Nieuport. My bellcranks are just gussied up a bit.

AbuF,

I'm no engineer, but I wonder if the old sliding servo idea could be adapted to your problem. I personally never used it (a Christy mixer was easier to set up), but as I recall it was used on ruddervators where (I think)the rudder servo was moved back and forth by the elevator servo. The movement did not change the rudder function, but caused them to move up and down in unison for elevator action.

Like I said, I'm no engineer and I can't really visualize your challenge, but them above came to my mind. If it helps, nobody will be more surprised than me (-:
Walt

PS you DO post some of the most intersting challenges. WT

Walt Thyng
Is this what you taking about
[link]http://www.rcmplans.com/issues/requested/content/features/pdf/r-fe-control-011977-1-1.pdf[/link]

That's very interesting and a great exercise in motion geometry. And it certainly is great food for thought. But, ultimately, those solutions allow for mechanical mixing in the days before computer mixing was possible. In the case of this "thought experiment" the elevator and aileron controls need to stay completely separate.

Of course, the problem here was that the original solved this problem by adding a human arm to the top of the stick!

I think you have to remember that the aileron and elevator servos in your illustration have to work together just like MvR's arm would have. There is no movement separation. If the ailerons are moved there will be a small amount of elevator movement, (use ball links to minimize slop and misalignment), that the pilot will compensate for. I find that, while there may be other ways, (and reasons), to route control runs, following the original seems to always work.

Martin

If you went full loop, you could burry the aileron servo in the wing and have the stick part of the loop. but not used directly to move the surfaces.

VERY interesting idea! A loop's a loop and it doesn't matter where in that loop the servo happens to be. Though I must admit I was planning on two independently adjustable servos. Makes like so much easier! But then I'm not very good at doing easy.

Oh, and Nieuport Nut, no need to worry about the unavoidable mixture of elevator and aileron as that just gets transferred to the human fingers at the end of the Tx sticks (for those flying Mode 2). But the system itself has to keep these two controls separate, that is a precise 1-directional movement on the stick should in principle only effect one control surface. It's up to the human to control the amount of interaction.

BTW, TFF, that's a fascinating way you've got your motor mount attached to the firewall! Oh, and I see what you mean now by the side mounting of the elevator servo.

And would love to know that your fuse weights in that state! I'm looking at about 450g on mine (with the brass frame and minus the engine). But I understand that you may not have access to it at the moment. []

Would this work for you? I am only showing the servo for aileron in this sketch.
The 'collector' pulleys for the elevator cables take almost all of the aileron-elevator interaction out of the system.

The problem (as I see it) is with that unshown servo. Assuming that both servos are fixed in their location, as the aileron servo rotates the bar on the floor (which rotates the stick) it also moves the elevator servo arm slightly away from (or towards) the attachment point on the stick. Maybe this is just marginal and a bit of looseness on the elevator linkage would make up for it.

OK. Wait. I get it. By using the "collector pulleys" we're taking out the problem caused by the wide separation of the two connection points on the stick. Is this right?

No, the linkage is still a problem. I think.

BTW, I picked up a couple of these the other day:

You are absolutely correct that there could be an issue between the elevator servo and the stick, which would cause an aileron deflection to input elevator movement. The collector pulleys take out the interaction between the stick and elevator but not necessarily the stick to servo.
This additional sketch shows an aileron servo (probably looking at the wrong side of it here) that has its control arm output IN LINE with the stick's longitudinal axis. This would create a conical movement, without shortening or lengthening of the linkage.

Good luck fitting all three servos in their optimum locations in that little area however.
(Note that the linkage to the elevator servo is at 90 degrees to the servo arm - or you will get differential)

Maybe that's got it! And you know that might not even be that hard to setup. With the elevator servo "inside" the ammo canisters (which are above the rudder bar) and the servo controlling the aileron tube torque would be off to the side. The rudder servo is the least problematic as it doesn't interact with the other two. Might do a working mock-up of this!

The proctor build of the N28 on the RCSB had the aileron servo rotate the roll like allanflower's drawing and the elevator servo piggybacked the whole thing.

Given the above illustrations, and your skill at soldering, is it not just a matter of soldering a servo mounting frame for the elevator servo that is fixed to the aileron tube such that the elevator servo and linkage rotates with the movement of the aileron tube? You may need a high torque servo for aileron to account for the added weight of the setup...but this would work mechanically.

I'll add a sketch if my description is not clear.

In fact, I think this is what TFF means. (I was drafting my post simultaneously, and only saw his after mine posted and refreshed the page...see, lightening can strike at the same time even in different states....lol)

No wonder I smelled something burning! http://www.rcscalebuilder.com/forum/...782&PN=3&TPN=5

Yes, I think we have a winner here! At least this is certainly ONE possible solution! I don't know if I'll be able to use something like this on my little model, but it's nice knowing it CAN be done and that somebody's done it! [sm=thumbs_up.gif]

BTW, I also like the way we've all just nibbled away at this problem until eventually the pieces came together.