I had the elevator trim tab installed backwards on a T28 Trojan I was to conduct a post restoration maintenance flight test on. Interestingly the up and down travel on the tab is the same which certainly heightens the risk of getting it backwards. Also the takeoff setting is zero degrees. Because this airplane is quite powerful it needs significant nose down trim as it accelerates in the climb after takeoff. When I started trimming the forward stick force required rapidly increased and I knew something was very wrong. Fortunately I remember a piece of advice I was given by a very experienced pilot. He said if the airplane suddenly starts to misbehave right after you have done something start by undoing your action, so I returned the trim to zero which resulted in a diminution of the stick force required and so I was able to return for an uneventful landing. However the bottom line is simple. It was wholly within my power to have avoided this situation by conducting a full control check.
Since that event I have had 2 further instances of control issues after major work. The first was a post repair check flight after a damaged rudder on a C 421 was repaired an reinstalled. When I cycled the rudder trim to the full left position the trim wheel jammed and could not be moved. It turned out an errant lock wire pig tail got caught in the works and jammed the control.
The second was an instructional flight in a C 172. The student was not my own, it was to help out another instructor who had to take a day off due to a family issue. During the control check the student gave a tentative stir of the wheel while watching the flight control surfaces and called the check complete. I interrupted him and said no, he had to check the 4 corners ( i.e. full forward, left and right and then full back left and right). When he got to full back the wheel jammed. Obviously we returned to the ramp and an investigation showed that a fastener for a wire bundle had come loose and the wire bundle sagged into the control path behind the instrument panel. The edge of the sprocket caught the bundle at full aft travel and jammed the control. There was evidence of distress on the wiring bundle so it was likely that this condition had existed for at least a little while.....