Hi mdJohnson, the deadband for that servo is 1 micro second not 1 millisecond.

A servo that can resolve more than 2048 steps is of no use as the best PCM radio's (including 2.4GHz) input signal to the servo is resolved to no more than 2048 steps. My nasty old 2 channel AM set has better resolution and latency (but this brings larger problems ).

I have heard second-hand that some hitech servos, despite the manufactures spec sheet claims, tend to "step" their travel and this may be more to do with mechanical constraints rather than electrictronic. Most cored servos tend to use a 3-pole motor, some hi-end ones used 5-pole motors. One dodgey old servo I had a 3 pole motor and every 120deg of motor position was 0.7deg of arm movement so the most steps the servo was capable of resolving mechanically under load was about 128 steps for +-45 degrees. Analogue controller or Digital that servo mechanically can't do any better than that.

That Hitech servo you mention is coreless however corless motors still have a pole count but the number of poles is not listed. Pole count and gearbox ratios set the mechanical constraints of the servos performance from an accuracy and resolution perspective but these are very rarely given. The digital controller might be able to resolve the input pulse and feedback pot to some level and provide control outputs to the motor, but if the motor is sitting on a "dead-spot" then it isn't going to move until the control signal (read position error) gets big enough.

Where does this leave us mere end users? Until the manufactures publish real performance data it's up to us to spend out money, try it out for ourselves and tell our mates what seems to work well and what doesn't