Re: tach vs encoder

Hi,

Let's start with the encoder first. I think you are confusing single
ended vs. single channel. A single channel encoder cannot give
direction of rotation information, which is vital for a servo loop.
It can only give accurate speed information.

Single ended refers to having only one output per channel. It still
takes two seperate "single ended" channels for both position and
direction.

Double ended "differential" outputs have true and complimentary
channels. For a quadrature encoder, they will be marked A (channel
A), A NOT (inverted, or opposite polarity of A), B, and B NOT.

The idea for differential outputs is any noise that affects one
channel will affect them all equally. By having the complimentary
(opposite) outputs, one can cancel this noise since it would subtract
and cancel from each channel while the signal from those channels
would add and reinforce. That's the theory anyway.

Going on to the tach vs. encoder. Some servoes require a tach, but
all servoes require an encoder. The ones that don't require a tach
get the "tach" info from the encoder. This involves some small
compromises, the biggest being able to derive a good "derivative"
from the signal for a PID loop.

As Jon has pointed out numerous times, a tach feedback will have a
very slight advantage in smoothness at very low speeds. This has to
be balanced against the very real and very high cost of a tach
equipped motor, if you can find one.

A non-tach, encoder only servo motor solution will certainly be
smoother than the best microstepping step motor solution, yet people
are perfectly happy and do good work with steppers.

In my opinion it is a decision that comes down in favor of "no tachs"
for all but the most demanding, expensive and esoteric applications.
It costs 90% of the bucks to get the last 10% of performance.

Mariss


--- In CAD_CAM_EDM_DRO@y..., "Multi-Volti Devices \(Murray\)" <multi-
volti@s...> wrote:

> I guess in principle the tach is for velocity characteristics and

the

> encoder is for position, but internally some types can share some

circuit

> similarities (ignore that last remark if it just confused

things...it's a

> 'chicken & egg' thought).
>
> My guess is how they detect the rotation and how they process the

output.

>
> Tachometer typically has a voltage/rpm output signal.
>
> Encoder has a # of pulses/rev output signal.
>
> A motor with both is apparently being used in an application where

both

> velocity and position are important control parameters.
>
> The reason I said there can be internal similarities is that some

tach

> circuits internally look like encoders - they start out producing

at a

> certain number of pulses per turn, but have a detector circuit that

gives an

> indication of how many of those pulses occurred in a given time and
> 'displays' that with a voltage proportional to how fast those

pulses occur -

> a frequency to voltage converter, if you will.
>
> I had some stepper motors with a confusing device on them - it

looked

> internally like a single-ended (non-quadrature) encoder, but maybe

it was a

> tach. Mariss voted for single-ended encoder, and he would know

better than

> me.
>
> Murray