I’d like to try and resolve this conversation if I can.
If I understand it right, Jaka’s point is that you’ll feel less wiggle in the pushrod when you connect it to the inner hole of the servo. He’s right about that.
Pylonracr’s point is that more pushrod movement for any given surface throw means minimizing the effect of linkage slop. He’s also right.
Where I see a disagreement, and correct me if I’m wrong on this guys, is in the idea that one must pick one benefit over another. I say you don’t. There is an absolutely best way to do it, and I’m going to try to explain why.
In an RC system, we start with angular motion at the servo, it gets converted to linear movement at the pushrod, then converted back to angular movement at the control surface. So let’s talk about gear slop. Gear slop would be measured in degrees, so let’s say a fairly loose servo had ½ a degree of slop in its gears plus the amount of deadband it has designed in. You could do some basic trigonometry and figure out how much linear movement that ½ degree would become at the pushrod, but we really don’t need to be that precise. It’s easy enough to see that the farther out you go on the servo arm the more linear movement that ½ degree of slop is going to create. If I understand it right, that’s Jaka’s point. But that’s not the end of it.
Since we convert that linear movement back to angular movement at the surface, the only thing that really matters is the angular ratio between the servo and the control surface. I’ll admit it took me a day or two to think this though myself, but I’m sure about it. If you were in a servo arm hole 15mm from the center of the arm (axis of rotation to be technical about it) and connected the other end of the pushrod 30mm away from the center of the hinge, you’d have a ½ ratio. So that ½ degree of servo gear slop would translate to ¼ degree of slop at the control surface. It would change absolutely nothing about the effect of gear slop to switch it to 10mm and 20mm, or 40mm and 80mm. The lengths don’t matter. It’s all about the ratio.
But then there is another source of slop-linkages. Linkages create linear slop that gets added onto servo gear slop. An understanding of ratios is helpful here too. If you use cheap linkages and get 1mm of total slop, and your linear pushrod movement is 20mm, then 1/20[SUP]th[/SUP] of the actual control surface’s movement will be seen as slop at the control surface. But if you use those same linkages but with longer servo arms and control horns to create the same ratio (minimizing gear slop too), then maybe you could wind up with 40mm of pushrod movement. That would mean only 1/40[SUP]th[/SUP] of the control surface’s movement would be slop, resulting in an overall tighter setup.
Bottom line: Use the longest control horns on your surfaces that are practical, use all of your available servo movement to minimize the effect of gear slop, then adjust the pushrod at the servo for the throw you need. That will give you the tightest possible setup and will also minimize the loads on your pushrod leading to less flex and possibly even minimizing the chances of having linkage fail.