Unless all things are equal and it’s a perfect model airplane world any effort expended to align the servo arms perpendicular to the servo case and or parallel to the hinge-line are meaningless. As previously noted it’s paramount to realize equal servo angle or arm rotation and again equal surface throw or travel arcs. It’s highly unlikely both of these travel arcs will be of the same volume due to the linkage geometry associated with unequal servo arm and control arm length at the very least.

Remember sub-trim on modern computer TX’s does NOT affect the ATV travel volume either side of center, however trims adversely lengthen and or shorten travel arcs.

Before you ever get out a TX, Hitec Programmer, MatchBox or the like mechanically matched surfaces i.e., dual elevators are a prerequisite for the elusive matched control throws. Once you introduce offsets in ATV/end-point values (TX’s/MatchBoxes) a scalar relationship with the TX and surface are all but lost.

More common than not; multiple linkage ratios are skewed thus creating unequal surface travel arcs and speed with control arm pivot points not being centered over the hinge-line, miss-matched physical locations associated with servo mounting heights, control arm height, and or control arm mounting locations require that the servo arm be centered wherever required to realize the center of the surfaces travel arc. In nearly all cases the servo arm will be centered several degrees aft with the proper mechanical setup realized of a perpendicular axis pushrod setup (typical of ailerons) and then there is a short pushrod configuration typical of servos in the fuse sides for elevators.

Bob,

If you’ll post the dimension information of the specific elevators, servo arms, control arms, control horn locations and such I’d be glad to run math that will net a nearly perfect initial setup.

the hingelines on my Aeroworks 35% 260's elevators are slightly off, placing one elevator half slightly higher than the other. To most, this is probably never even looked at. Then, to top it all off, I misdrilled the control horn by about 1/8" on one elevator half. Both control horns' fulcrum is directly over the hingeline still, but one is 1/8" outboard compared to the other. It doesn't seem like a big deal, but, like Mglavin said, no amount of additional mechanical tinkering or servo programming will ever fix it, not even a Hitec servo programmer. The other thing I noticed (I use the two rods coming to a point behind the rudder trick, too) is that, while I can adjust my ATV's to achieve the exact same throws for a given rate, they still don't move at the exact same rate all the way up to their endpoints[:@]. Besides the obvious geometry problems one can encounter, other factors, such as variences in the servo or servo arms themselves, or a little extra friction in the hinging of one control surface or ball joints campaired to the other, can also reek havok. The best thing to do is to get rich and get sponsored. That way, you can afford a pro-build instead of the chinese stuff, and can have buckets of servos lying around so you can mix and match them to get exactly what you want!

Yep. You are right, it has to be mechanically perfect if you want things to really match up.

Key is to take your time and make absolutely certain that everything is mounted in exactly the right place. If the servos and control horns are mounted right or at least exactly the same on both sides, getting the surfaces to match up is not that hard.

I thought all of the holes were predrilled for the control horns on the QB. Was yours not a QB?

Perhaps my comments were absorbed out of context. You CAN remedy ill installed linkage scenarios with the proper approach mechanically; iteration is the key to resolving problematic miss-matched linkage/surface/geometry. For the most part you can match servo angle with like ATV values and surface deflection, while the ratios maybe slightly different the end result is what were after.

Are you implying your control arm is offset span-wise 1/8" different than the opposing control arm? If you’ll provide the numbers I can likely work out a best case linkage setup. At the very least we can see the numbers your realizing.

This whole thread had been a REALLY outstanding tutorial. After working through this on TWO Yak's, (the BME 30%, and this morning the QQ 73"), I am now TOTALLY convinced that the angular variation of travel due to rotation asymmetry is THE culprit. And of course, that is clearly made WORSE when OTHER mechanical variations creep into it.

For the VERY FIRST TIME.... it is absolutely CLEAR to me now WHY "Programmable Servos" is "The Solution", and I honestly never bought into that before.

OTOH - we are splitting some pretty fine hairs now, when using 12 or 18" pointers on elevators moving 35*....it is ALSO clear that we MUST run ATV's at the highest settings possible, which is also somewhat of a "new" revelation to me, as I was always one of those "stay around 100%" guys.

MikeGlavin - if you really want to do the math, it will be fine, but I have really come to understand what's going on now, and have corrected the mechanical aberrations of both of these fine airplanes. They will only be better now!!!!!!!!

Finally - assuming that all of this is ABSOLUTELY the same on ailerons, even if only one servo connected per side???
Believe it or not, guys, although a little painful, this has been a GREAT learning growth for me, and I appreciate all of you helping me out!!!!!!!!! THANK YOU!!!!!!!!!

Yep, it's the same for all surfaces.

You can even do this on the throttle. It's much easier as all you really need to worry about is the connection points on each arm. If the throttle rod is connected on the engine at 1 inch from the center, then connect the rod at the servo end 1 inch from the center. You'll end up with a linear throttle but I still use curves for the inevitable gas engine power band issues. IE...full power at half of the throttle butterfly movement.

What you are going to end up seeing is both of those fine Yaks are going to roll much more axial once you get the ailerons done and hard pulls into the vertical will be exactly that, the elevators will be in sync and the plane will transition easily.

If you have any aileron differential in the radio, remember remove it before you go to work on the aileron linkages

Bill - interestinly enough, I *had* figured out the throttle thing a few planes back.... but didn't transition that knowledge to control systems. This is cool.

Ya know, I gotta tell you, you had me thinking really hard on how to explain it. I've gotten so used to just doing it. It actually gave me a migraine (just kidding). I nomally use Hitec's now but on a whim I used 8611's on my AW Yak this last time and I was "lost". Granted it was still easy to do since this thing is all figured out for you, but I did have trouble with the H9 aluminum arms trying to get a matching set for the elevator halves. I finally said the heck with it and went to the SWB arms that I use on the Hitec's.

I just pulled the plug on the new WH 33% Edge and in two weeks I'll be ordering the 5955 Hitec's for this bad boy. I'm getting really tired of replacing gear sets at 30 to 50 bucks a pop after 75 or so flights because I wore them out on the normal metal gear servos

I've heard nothing but good things on the 5955 since they use titanium gearsets. Guys are in the area of a couple of hundred flights with no slop in the geartrain and the guys over at TBM have been using 8.4 volts on them succesfully for over 440 ounces of torque

Bill,

Take another look at the throttle linkage scenario; if you utilize a 1" arm mono-mono, resolution is less than desired and or available. The range of our carbs throttle body plate is approximately 80 degrees, for maximum resolution we want to utilize MAX ATV together with MAX servo angle. So what were after is a linkage ratio that will net same. My math suggests a .75” servo arm and a 1.0” control arm will suffice.

Yep... what he said!!!

I like the exercise; if nothing else we can share the good and the bad and maybe learn something along the way.

See the attached picture for the required measurements. There is one critical dimension that is not depicted this is the spanwise offset between the servo output shaft and the control arm.

That is almost exactly what I have setup and I am getting 120/120 on the ATV's. Is there anything that can be done to improve? The pushrod is parallel to the servo arm and the throttle arm the way I have it set up. I just ran the engine this morning after making some adjustments to match the ATV's exactly. The throttle trasition is very linear and smooth. Makes really REALLY low hovering a snap when the airplane is not bouncing up and down with 1 click of the throttle.

Mike,

Good info. That would help in cutting down on the throttle curve. I'll give that a shot. [sm=thumbs_up.gif]

Yeah, I mentioned it a couple of posts back in one of my rambles. I did not know if you guys had seen it or not. My weakness is technical jargain. I know how to get the results by techniques I have learned, but I do not know how to calculate it mathematically. I think I could do it if someone showed me (again, re school) but I have forgotten all of the formulas years ago.

Bill, remember I had a BME 110 come apart on me last week?. I got it back yesterday and ran it this morning andit was ripping a 26x10 right away. I think it is much smoother now than it ever was before. Throttle transition is absolutely flawless, even though the engine is brand new. Only problem is it ran a little hot. I am hoping after it breaks in a little it will cool down. I forgot my temp gun but I am guessing the jugs were in the high 200's when i landed.

This was the cool part. I rocked the spinner about 15 times before I flipped it the 1st time. I choked it and it fired on the 1st flip. Then it started on the 3 flip. Unbelieveable.
I cant believe Keith got the engine, repaired and shipped it all in the same day. Again unbelievable.[8D]

Now that elevators are "fixed", and they are sweet, now....since the ailerons aren't side-by-side, how do I do this setup to make sure the rate/deflection is the same for each of them??

The downonthedeck tutorial talks about matching 'servo deflection' angles to "control movement", but that sounds like that is simple ATV issues. How do I ensure proportional tracking BETWEEN the two wings??

Well yes, but as usual I will talk in simlpe terms and then let the mechanical engineers clean it up.

A question about the basic setup criteria.

Do you know about setting the control horn heights so that they are at the same height off the aileron relative to the center of the wing. As you know, the wing gets thinner at the tip, so as you move out on the aileron away from the root, the control horns will each be a little higher off of the aileron surface. The inside control horn may be 1/4" off the surface and the outer may be 1/2". It just depends on how far apart the servos are and how much the wing tapers.

Once I get that set I turn on the aileron servos and get the servo arms at or as close to 90 degrees to the hingeline as I can get using the protractors. I will subtrim what I cannot get out.

Then I connect the pushrods to the control horn and servo horn with the aileron centered.

After that I turn off the radio system and move the aileron by hand, adjusting the pushrod lengths until both servos are moving at the exact same deflection per the protractor with the aileron hitting the stops on full deflection up and down. It does not matter if the servo arm move more in one direction as long as both servos move the same. If your control horns are in exactly the same place relative to the servo you can get both servos on each aileron exactly the same. If you have to move one control horn in or out to increase decrease throw at the control surface, you have to move both control horns the exact same amount.

Once you get it perfect, go to the other aileron and using the protractors setup so that you get the exact same deflection on the servos on both wings with the ailerons hitting the stops. And they should match the other aileron.

Once both sides are mechanically identical per the protractors, I turn on the radio system and carefully move the ailerons. LEAVE THE ENDPOINTS AT 140 (or max) Adjust the control horns or servo arm to get the ailerons to move to the maximum deflection you want at max resolution. . It should be pretty easy to change both servos exactly the same. And the same on both ailerons. You have to do this without moving the endoints. They have to be the same. IF you absolutely must reduce teh endpoints, make sure that all of the numbers are the same.

Once you get that then you can make tiny adjustments to the pushrod lengths to get rid of any binding in the servos.
This is where the Rocket City or Hangar 9 titanium turnbuckles (prolinks) come in handy. They allow you to make the smallest changes you want, and you are not limited to 1/2 turn adjustments.

http://www.chiefaircraft.com/rcmsec/...s/Hangar9.html

Again, we now use the programmables and its so much easier. You just hookup the control linkages with the control horns set to the right height. Then 1 servo at a time, put the servo arms on, set the center, program the endpoint of the servo and adjust the servo arm control horn to get the max deflection you want. Then go to the other servo and using your notes, setup the deflections the exact same. Adjust the linkage to eliminate the last little bit of binding. Then do the other aileron the exact same using the same servo deflection numbers and go fly.

Figuring out how to use the programmer is the hardest part.


Here we go...

Mike, GREAT explanation, and made sense. If I had dual servos, but I don't. Right now, the birds I am forced to fly are one aileron per panel, so I assume everything you wrote stays the same, just apply to a single-setup?

You did not mention subtrim use.....is that taken care of by the "radio off" mechanical travel checking???

bodywerks - my BME setup WAS nearly identical to yours, error-wise, but messing with the pushrod length and sub trims AFTER setting max atv got me there. It was trial and error, and the BME aren't turnbuckles, so I must have installed and removed each elevator ball link from the Airwild arm a dozen times. But -- in the end -- they are VERY, very close. Not actually "perfect" but HUGELY better than at start up.

Knowing the "subtrims v.s. pushrod length" deal unlocked the puzzle for me!

Yep, just like setting up a pattern plane. [8D]

As far as subtrim, sure. If you are not programming you might have to use a little subtrim to get the centers as close to the same as possible.
also, we are now using a big old protractor glued to a piece of foam to set the throws to be exactly equal. I did not mention it, but it is a very nice tool. Its blown up to about 8" wide and you cut a V in it so that you can slip it onto the wing right near the root of the aileron. Its a great way to get very precise deflection measurements.

GREAT tip on making a tool that will "hold" itself on the wing!! That has always been a PITA problem for me!

MGLAVIN - measurements coming via email!!!!!!

Guys, I wanted to let you all know the result of the effort. The big Yak (had the biggest rate mismatches on elevators) is really NOTICEABLY better. Pulls radiuses more accurately, less rudder workload, less aileron roll-off with pitch...lot more predictable and solid throughout the range.

I was surprised that I would actually "Feel the Difference", seein' as how I *already* fly JR.... but .... I could, and it was much better.

Thanks for all the help.

That's great to hear. Once you do one, the rest of the planes get easier. It really does help a lot

Good deal Bob. Its really nice when you don't have to fight the airplane![8D]

This is a great post and im about to put most of this into practice in a Comp Arf 260 3m i have just finished and had 3 setup flights on. Its all trimmed about right but rocks a bit on harriers and trys to roll out in one direction consistently. I think due to the fact the elevators on high rate only are mismatched by about quarter of an inch full up deflection, normal rates (acutally set more like medium rates) are fine.

What occurs to me in reading the thread is that time and again the gear we use is basically designed inadequately for high end and more extreme use that we modellers some times use it. In the competitive industrial world this as an operational or install process would be intolerable as highly error prone and requiring far more skill and precision than is normally reasonable to expect of 'average' individuals doing the job.

At full deflections of 55deg+ a fraction of a millimeter out on linkage/arms/horns can cause aggregate errors in full deflection that seriously adversely affect flight performance. All of the tools we have in our armour on the radio and even programmable servos operate in a linear fashion and are inadequate as the error is non-linear in character. However, it is not mathematically unpredictable, quite the opposite. Given a set of end points and mid points it is simplicity itself to program a trasnmitter with algorithms (im really talking of the tx operating system now and not user end programming which should be kept simple) to calculate the error effect at all points on the servo resolution. Im no mathematical genius but thinking back to school days i think this is basic Calculus. However, tx operating systems are stuck in the linear world and allow only the most basic of travel and mid-point adjustment on a linear scale. The closest the Tx comes to this is the throttle curve proramming for Heli's, but again this is poor as it is a linear scale on which a serious of points can be set to approximate a curve, and a poor curve at that given the very few points allowed. Given the ongoing development of superb ARF's of great quality, and getting better and more and more ARF, such as the one I have just built there is a need to sort some of these problems out post build, and not take for granted that the builder can mechanically obviate them in all cases.

Hopefully instead of inventing flashy noises, MP3 players, colour displays with simulated 3d models, gold coloured sticks etc etc all of which are of no use to me, the manufacturers will A) make the hobby safer (i still dont understand why they dont build in failsafes for channel clashes, this again is simplicity to do) and B) make the Tx more useful to us with the nature of the problems modern flying brings. Hopefully they will also make it affordable! Since desktop computers and mobile phones now have more processing power and operating system sophistication than your average Tx which can be more than 10 times the price!

There's one other thing I didn't mention. besides being offset, actually about 3/32", spanwise different than the opposing control horn, the geometry is nearly identical. I even put a spacer under one of the control horns to adjust for the one elevator half that had hinges installed slightly off center. I can also get the travels to end at exactly the same spot using ATV's when I am in high rates, but then it is way off in low rates. What I ended up doing was making a compromise between the high and low rates - now they are both slightly off.
My understanding is that some of the top Futaba radios have adjustable end points for each rate. Unfortunately, JR does not...
One thing I might be able to do, though, is use a couple of my multi-point mixes instead of my dual elevator function and set one mix up for my low rate flight mode, and the other mix for my high. I may have to try that.