I crashed a plane a few months ago == a TufFlight tiger EDF doing a low pass for a radar gun about 20 feet off the deck. Ground speed was 129 mph. At the end of the run, the plane did a fast "down" elevator and planted itself in the dirt. It may have tried to roll too (ie, snap due to full up on one servo). It happened too fast and was about 400 feet away so I couldn't really tell.

At the time I blamed the action on a control linkage actuator slipping on the taileron axle. I've since silver soldered the linkage for bulletproof security.

I repaired the plane and flew it again several times without issue the latest at the NEAT fair this weekend. After a few low passes, I pulled up high and the plane did an uncommanded snap roll and then spiral before gaining complete control again. I landed immediately.

I'm flying a spektrum AR500, and two HS 85 MG servos, and a Shulman Fury 60 controller which has a switching BEC with more capacity than the servos need. Also keep in mind the loss of control happened when the servos were NOT highly loaded (level flight in one case, gradual pull up and immelman in the other) so I doubt very much a brown out could cause the issue especially since I've done much more radical maneuvers which would draw more current without problems. I also doubt loss of signal is to blame given my failsafe should hold the last known position not a full up condition on only 1 servo.

I now suspect a servo to be the issue and will replace both of the taileron servos before flying again.

Perhaps a servo is to blame in your case as well? Does your F20 use only 1 servo for elevator control? How about the Saab? That might explain unusual sensitivity. Did you guys get the same kind of servo from the same source? (ie maybe a bad batch?)

I didn't think it was possible for a spektrum RX to issue an uncommanded "full up" command during an error condition. It' s much easier for me to suspect a bad servo, but of course I'd like to hear any evidence in this case.

Joe