Dr.Ben Elite Veteran Location: Richmond, VA, USA
| Vinnie,
Go back and check those models at around 1/2 dozen points in the collective range after the servos have had several cases or more of fuel through them. I promise you the swash won't still be in perfect track. I've seen 9252's go out of perfect match after just several gallons, and servos from the other mfg's are no less prone. It's all about a carbon wiper pot; the part wears, and centering and total throws change. Also, it's not all about keeping that swash dead level. Interaction also concerns what is occurring when cyclic commands are introduced. Go to the higher or lower extremes of collective range and enter a pure ail or ele command. If the model has any time on it and has not be recently retracked, I virtually guarantee you see interaction bewteen the two cyclics. Unless you want to replace servos after only a few cases of fuel, you have no choice but to use the radio to get the servos back in line. There will be some interaction in any case, because rotary servos are being used, and ball links on the servos are moving further away or closer to the centerline of the servo with extremes of collective throw. Even the finest eCCPM models have these liabilities. It is the best designs, though, that minimize the manifestations of these liabilities.
In reality, the trick with any eCCPM model is trying to determine how much error in the system you can tolerate before flight performance really suffers. For some sport flying where the majority of the control inputs are broad and using up much of the collective range back and forth, you won't notice subtle changes in swash tracking along the way. Try some FAI where you have to have the model slowly climb perfectly straight out over a cone while pirouetting, and a little mistracking in the swash can drive you nuts. In point of fact, the FIRST place I notice eCCPM "issues" is in precision rolls. Formerly well tweaked swash phase compensation mixers gradually go all to hell. I have to remember to recognize the problem for what it is rather than try to retrim the whole model. |