Re: [CAD_CAM_EDM_DRO] Lovejoy Couplings

jmw@... wrote:

> Are Lovejoy couplings (jaw on each end w/ an elastomer cross-shaped
> chunk in the middle) suitable for directly connecting a servo motor
> shaft to a leadscrew, or will wind-up of the coupling cause
> positioning problems? If the encoder was on the other end of the
> leadscrew (rather than the other end of the motor shaft), wind-up
> wouldn't matter, would it?

In a wood router, it is probably OK. In a CNC machine tool, I doubt you
would
have good performance over time, as the compliant coupling element wears
or
gets softer.

I have NO couplings between the motor and the leadscrew, so there are
no couplings subjected to motor torque. A toothed belt is used there.
There is a one-piece, helical-slit coupling between the leadscrew and
the
encoder. It only needs to take the torque required to accelerate the
encoder, which is negligible.

I believe the Lovejoy couplings are not recommended for intermittent
torque loads, either.

With the encoder on one end and motor on the other, it is possible for
high performance servo drives to 'ring' the leadscrew into resonance
as a torsional spring. If this ever happens and the gain at that
frequency
is much above unity, things will get exciting VERY quickly! A low power

servo drive would probably be OK like this, but it is definitely NOT
recommended practice. I have had a few resonances develop while
developing my servo systems, and it is amazing how much vibration
a modest servo motor and servo amp can produce!

Putting a stepper on one end of a leadscrew and an encoder at the
other can be even worse because the steppers tend to generate a
lot of torque pulses that can excite leadscrew torsional resonances.

Jon