servos get a position command and then they move the servo to the commanded position (comes from receiver). when the command doesnt change its still trying to hold the last position commanded. in the servo is a program that someone wrote that tells it to use for example 1 amp of current to correct the difference of position from the command and what it reads on the internal feedback (pot).

the guy who wrote the program had to make some educated guesses about the mechanical assembly and that is when he decided 1 amp of current is the right amount to bump the motor with to make small corrections.

2 things go wrong commonly.

1-the friction and mass that the programmer made his guess from has changed. so 1 amp actually over corrects the difference in command versus actual. so it overshoots the correction move and then it overshoots again and gain. this is an oscillation from a bad correction move. increase friction and it will stop.

2-pots wear out or experience noise. the receiver sends a command and the servo moves. it watches the pot to confirm it made it to the commanded position. if the pot doesnt read the expected value it will try to correct it until it does. if the pot has a bad spot or their is noise on the pot the voltage from the pot fluctuates and the purpose of the servo is to stabilize the position (the fluctuating voltage). if the voltage isnt stable the servo will move to stabilize it.

^ exactly what is going on with the Proteks

very excellent information. With that being said what manufacturer has made the best educated guess in your opinion

great question. I only use servos that I can program the response myself. Im not paid by them but I’ll provide an address if they want to send me a thank you. Sanwa PGS line allows you to change the compensation power at low, mid and high position.

Explanation a little off bc current is drawn by load not some guys educated guess. The BEC voltage is constant and amps are increased as a result of the load in the circuit at any time.

what does acceleration use? Is the acceleration of the servo always the same in low, med, and high position? If the pot says its at 1 degree but the servo thinks it should be at 1.5 does the servo respond the same acceleration as if it was supposed to be at 90? What if its making a small compensation near an end point?

You are partially correct. The load does increase the current but most of these servos dont have a current loop to add friction current to acceleration current and then math it out when at velocity. Someone determined a range of masses and did some testing to understand how much more or less current to use for small deviations, large deviations, and constant velocity.

If you are lucky there is a lookup table that creates a couple different acceleration/deceleration ramps like you can see in some servos when the reach the end point. You might also hear the term feed forward model but its fundamentally the same. No matter what somebody took a guess at the acceleration current to add to the work current so that it can be taken back out when you hit the end point. Testing or a guess or whatever doesnt account for the servo being in a mechanical system that is under or over the range that the model was designed for.

im not trying to throw dirt on guys that took an educated guess on how much current to use anymore than Im saying guys misapply products. But both situations exist and if you understand you can make a better choice or even control the variables yourself and screw it up or make it really good,

Ive taken probably 10000 guesses at current for work, acceleration, positioning and synchronization and they werent always right so I might be insulting myself more than anyone. Its just some foundational info to help someone make a decision about what to think about next.

However it is true that some servos always accelerate at max available current and max current is available at constant velocity. Then someone took a guess as to what position to slam the brakes on. Which is still an educated guess to minimize damage to the gear train and not under or overshoot the end point or neutral point.