ORIGINAL: woodie
Speed of the servo has nothing to do with 'accuracy' from an electronic viewpoint. The servo may resolve the receiver signal into x number of steps which might affect the rotational degrees of a 'step', however, the accuracy is determined by the 'dead band' allowed by the servo. The 'dead band' is how close does the servo make the signal from the pot match the signal from the receiver. The broader the dead band, the bigger the window of rotation that the servo considers to be 'correct'. This accuracy is for the servo only, it ignores and linkage or mechnical slop.
Speed is a matter of gear ratio, voltage and motor strength combinations in the servo. You can take a servo at 4.8V and it will have a certain time frame for x degrees of rotation and have a certain amount of torque. Up the voltage and the time to rotate that same x degrees decreases, the torque probably increases, but the accuracy remains the same.
The tighter the dead band, the more the servo works to find 'center', and the closer to center it will return each time. When you hear a servo 'buzzing', that is the servo trying to find center. The weight of a control surface, drag in the linkage, etc, may inhibit the servo ability to land on 'zero' and stay there. It is easier for a servo manufacturer to build a servo with broader deadband, the components can have a little more tolerance, etc.
Woodie
Hi Woodie
I am a bit confused by what you say here. If I read you correctly then servo accuracy is tied to the deadband width and the effect of the number of "steps" is relatively inconsequential.My thinking was that the more "steps" there are in the rotational travel of a servo, the more accuracy is attained. This of course is not taking into consideration the mechanical link to the flying surface. Don't get me wrong, I am really leaning toward the JR8411 for my coming 2M but just for my ownedification I would like tounderstand the reasoning of why people are/are not using a certain servo. In the JR literature they state(for the 8411) " increased resolution of 5900 steps per 120 degrees and a less than 1 msdeadband for extreme accuracy". Is the number of steps quoted just another gimmicky selling point ordoes it havemerit when they include it in their information? In practical application is this accuracy noticable? On the Hitec websiteI could not get a number for the steps involved but they give a 1500 usec number for theoperating window. It does state however thattheir deadband is 1ms (this is for the 7966). Both servos are operating at 250hz. I am also a little confused as to which is the best gear train in a servo. Some people say the metal gear train inherently has too much "slop" but the preferred JR 8411 has just that. All a compromise I suppose. Maybe the difference of the metal gear train is inthe quality of a higher priced servo (JR) as opposed to a middle cost servo (Hitec) giving credence to the fellow that said "you get what you pay for". I am just 7 months into the hobby again after a hiatus from it for 15 years. I did compete in pattern almost 25 years ago and amdetermined to get back into itbut everything has obviously changed. Back then there were no computers!! Now even the little servos in our airplanes have them!! At that time the concensus was that electric flight was going nowhere as they were just barely getting off the ground hauling around 12 or 16 nicad cells. Now it seems everything is going electric. The new 2m will be just that. Learning curve is still very steep. There are 'way more choices to make now than there were back then and I would like to wind up with a machine I will be happy with (for the time being). Now to get some more stick time!!
Thanks again for the input.
Cheers
MJ