Unwanted Mystery Differential

[HELModels]

I've been refining a V-tail design forever it seems. Each iteration gets better in some way, but reveals some new problem even if it is minor. My Tx doesnt have V-tail mixing or any fancy computer mixes, so I cant go there to solve this problem. The problem is when I turn left I need very little up elevator, and when I turn right I dont seem to have enough. It isnt a problem unless I am flying about 10 feet up and trying to stay in close. It happened today and I was able to correct, but it was very noticeable.

Here is what I have examined:

1) control horn holes not over the hinge line - maybe a mm or two difference, but not alot

2) maybe my inputs are skewed - easier to pull the stick accurately for left turn then to push it and pull for a right turn

3) something funky with the servos - I'm using 2 HS55 and one arm is cocked when the trims are neutral, but I figure the servo arm is going to travel the same regardless where neutral is.

Any ideas?

[JetPlaneFlyer]

If the servo arm is 'cocked' then this will be giving you differential movement. With the servo centred the angle between the servo arm and the pushrod must form a 90 Deg angle otherwise differential movement of the pushrod will result.

[HELModels]

Thanks, mystery only to me apparently. I was using JR's and tried HS55 for the first time. The arm on the one servo is square and the other cocked slightly as they came out of the box. I read about centering here and there, is that what I am dealing with here?

[HELModels]

One thing I havent considered is the mixer. The Tx doesnt provide it and maybe the mixer has it built in. It is just strange that the one servo is square while the other is not.

I was out flying today at the recycling center and had some visitors asking about the plane, so I said, I'll land it and give you a look. I was coming in nice and gentle but did not straighten my final, just one big steady turn and the extra nose down in a right turn was there. I landed alright, just not a greaser even if the kid was still going cooool man. So, tonight I unscrewed the servo arm and could not find a spot where the one servo would square up. I didnt have this problem with the JR's, but it is a rebuilt airframe.


What if I were to bias the one arm the opposite direction it is currently biased?

[HELModels]

Screw it, I went ahead and adjusted both servos by removing the arms and resetting them. I moved them both so that when the stick turns either left or right, more up and less down. If the stick is moved for a left turn, the right rudder goes up more than the left goes down. Same with a right turn, except opposite. I figure this biases both servos in the same way and will outweigh any small difference in centering.

Maybe some errors have combined in this airframe that cancelled out with the other one, but it is annoying.

I'll test fly it.

[BMatthews]

On something the size of our models and their control system components even 1/2 a mm is something to be corrected. Close doesn't work when you're down to stuff like this. You don't need a micrometer but things should be as close to right as you can get by eye and if you think your eye is being fooled use a small right angle square item to help eyeball things.

Centering the servos these days is usually more about moving the arms around than the old way of centering the arm with the pot adjustment. Although a lot of servos still have the ability to center them that way. Its fussy work though since you need a quick and light touch with the accent on the quick part. First check to see if you can square up the output arms by moving the + arm around. The splines are not centered on any one arm just so you CAN move it around until you're really close. Likely you can get one that's so close that it only takes a click or two of trim to set it dead on.

You may be able to adjust the centering. Take out the arm screw and put a really small jeweler's screwdriver down the hole. Turn it around and see if it drops into a slot down there. If it does then you can adjust the centering a little. This takes a really find and fast touch as mentioned because as soon as you move it even a hair it runs away to that side. You're playing with a reverse reaction here. You need to move it and then immediately let go of the screwdriver shaft. If it's not right then try again. Keep "kicking" it until it's centered. The servo needs to be powered and with the trim at dead neutral when you do this. If it runs hard over and stalls you'll need to move the screwdriver the other way really a little and really quick so it comes back into the normal travel range.

[HELModels]

Bruce, I've got some jewelers screw drivers and they are too big. You mention trimming it out and I believe that is what I did after testing it with the intentional differential. I hit a huge, I really mean huge, wind sheer and there wasnt enough rudder to avoid a tree limb. That broke the tail off clean. No problem there, just a dab of glue.

After the glue setup, I reset the servo arms the way they were and then set the trim tabs so both were square, then readjusted the surfaces. That is what you mean by trimming it out, correct?

[BMatthews]

Sounds like you got it. But there's more to it than that although I didn't go into it above. The angles that the pushrods meet the arms as much as the angles of the arms. Oddball setups of the pushrods that have them at funny angles can require that the servo output arms be offset to avoid differential. Without seeing some good pics of the servo setups and control horn setups and pushrod routing it's hard to really say if anything is wrong or not or if you can fix what you have with just a tweak and scooch here and there.

[HELModels]

Bruce, I took another look at the control horns on the tail. I actually measured this time and found a 3 mm difference. The left horn was sitting back from the hinge line 3 mm further than the right. It was also further out toward the tip of the tail by 1/8 inch. I removed and reset that. Amazing, the beater this plane replaced was ugly as sin but had no quirks. I examined the old tail and its horns were lined up right over the hinge line. So, I will test it again and see if it is better. It might end up the servo center was the least of it.

[BMatthews]

Those amounts will result in quite a big difference. There's no "close enough" when it comes to setting up the controls equally from side to side. Do it all to the best that you're eye and some simple rulers and squares will allow. As I mentioned before there's no need for the laser survey transit or the digital verniers but as good as your eyes and simple tools will give is what you're after.

[HELModels]

Bruce, thanks! I reset the one control horn on the tail, but it snowed pretty good here and I didnt want to rust the pushrods. I can now assume the tail is right, but that servo... I went ahead and took a really bad picture. I thought this camera had a macro function, but flash doesnt work with it. It is good enough to show the from factory difference.

[HELModels]

The control horn being off the hinge line was the big cause of the differential. I flew it today, nice clear skies, no wind, and I was able to bank sharply going right and left at slow speed and low level without worry. I looked for the nose down tendency and it was still just a hint but very flyable. The remaining hint must be from the servo being skewed?

[HELModels]

The original problem was that I rebuilt an airframe and the new plane had differential V tail that showed up as a tendency to nose down in right turns, requiring way more up elevator than normal, more than the old plane, and more than for left turns. I found a control horn sitting too far back from the hinge line and fixed that with a significant improvement in handling. The problem was still lurking, although it was barely perceptible.

So, to solve the mystery I examined the mixer. I also looked at the Tx. When I installed the servos and plugged them into the mixer, I used different servos as inputs into the mixer. This time it was rudder going into the mixer where elevator had gone before. This required I use the reverse functions on the Tx. I swapped the inputs to the mixer and flipped the reverse switches on the Tx, and Bingo, the centering on the left servo squared right up and now the right servo has the arm biased/skewed/off center.

The error in the installation of the control horn had combined with the mixer's built in differential and I ended up with excessive differential.

[BMatthews]

I'd fix that last bit up with a little subtrim added in.

Quite the mystery but it got solved in the end.

[HELModels]

I found the solution over in Electric motor design.

The solution to the slightly off center servo arm is to locate the arm on the servo so it is square - Duh. The HS-55 has a 4 arm servo arm thing and if you rotate the arm 90 degrees you can use the next position. If that position is not square, rotate another 90 until it is square. When I say rotate, I mean unscrew the arm and position it as if it has been rotated. Dont go torqueing a servo by hand. For one, it will not change the position, and two, you can damage a servo. Apparently, the splines are arranged so that you can set them square or off center, depending on which spline tooth the arm is set.

This might seem rudimentary, but it was something I never grasped by just fiddling.

The unwanted differential effect which I originally was trying to eliminate had multiple causes, all of which have now been fixed.

[Texas Buzzard]

YOU NEED A MIXER. It mixes elev. and aileron. It weighs about 0.5 oz and costs about $14.00.

I have two and use them for "V-Tailed Gliders and also power ships.

Go to this addresss and find mixers.

http://www.rcpowers.com/accessories.htm This is Dave Powers in CA. He is a good guy who is customer friendly. I have delt with him for over a year and he has good stuff....go see his videos, they are instructive.

YOU NEED A MIXER - GET RID OF DIFFERENTIAL.

[HELModels]

Duuuude, I have a mixer. http://www.dionysusdesign.com/product_info.php/cPath/44/products_id/174

I had unwanted differential and I didnt know why. I found each source.

1) control horn was off the hinge line - relocated it and was the biggest cause

2) mixer seemed to have some - switched inputs and flipped servo switches on Tx which improved

3) servo arm was not square on one servo - fixed it by relocating the arm 90 degrees to the next arm thing. This was the final correction.

All 3 sources added up instead of cancelling each other out. It handles identically now in left and right turns.