Re: [CAD_CAM_EDM_DRO] Question about Tachs and Encoders on DC motors

sprooney797 wrote:

> I understand how DC Motors with encoders can be servo controlled, but
> I am not sure how DC motors with tachs are used.
>
> I also noticed at work that our Bridgeport EZ-Track has SEM DC servo
> motors that have both a Tach and an Encoder on them. Why are both
> needed? And how is this used in a control scheme?

Classic positioning servo systems use both a tach and an encoder.
The problem with the encoder is that between encoder pulses, no
information is given, other than that the system has moved less than
a full count. So, the encoder is a discontinuous-time sensor from the
very beginning. Classic analog loop control is made VERY complicated
by mating continuous-time and discontinuous-time systems, especially
when the encoder updates at a rate depending on its speed! That
is a disaster for a control system with only linear functions.

So, the fix had been for many years to put a continuous-time velocity
sensor (a DC tach generator) on the motor, and feed this back to
the servo amp. The classic servo amp has two nested control
loops. The inner one compares armature current to an error signal
derived from the outer loop, and forces motor current to match that
error. This error signal is derived from the difference between the
velocity command from the CNC control and the actual velocity from
the DC tach.

There is an outer positioning control loop, that can be as non-linear
as needed, that compares encoder reading to desired position (and
may also add corrections for calculated velocity vs. desired velocity)
and produces a new velocity command to the servo amp. This loop
is sometimes implemented in application-specific ICs, otherwise
in a computer of some sort.

Jon