Re: Adding inertia to an axis

Hi,

Jon is absolutely right: adding inertia only shifts the resonant
frequency, downwards. It does nothing useful.

Step motors have two resonant features, commonly called low speed
resonance and midband resonance.

The first is treated with first-order damping, i.e. friction or far
better, microstepping.

The second has to be treated by second-order damping, i.e. viscous
damping. Jon very adequately described the mechanical approach which
is commonly called a Manchester damper.

There is also an electronic method of doing the same thing. This is
best viewed in terms of phase margin. Without going into the oft-
repeated details, left alone, a motor / drive combination develops a
180 degrees phase lag. This is a classic setup for an unconstrained
oscillation or resonance.

What is needed is to introduce a phase-lead component, or a
derivative function to the total loop response so as to decrease the
phase shift to significantly less than 180 degrees.

This can be a mechanical rate damper (derivative function) or
electronically by adding a "rate of load change" component to the
drive loop.

Mariss

--- In CAD_CAM_EDM_DRO@y..., Jon Elson <elson@p...> wrote:
> Peter wrote:
>
> > In the recent thread about X-Axis won't do as it's told - there
was a
> > suggestion from Hoyt to try adding inertia to the axis. As it
turned
> > out that wasn't the problem in Richard's case - but it may be the
> > problem for one of my machine axes. But I'm curious as to how
much
> > extra inertia to try adding... I'd like to start in the right
ballpark.
> >
> > I have a NEMA42 motor so I will probably try turning a solid
steel disc
> > around 3" dia and say 1" long as a starting point, recognizing
that the
> > disc's OD is most significant in moment of inertia calculations.
Would
> > appreciate any suggestions other than my wild guess as a starting
point.
>
> Adding true inertia only moves the resonance to a lower speed.
that may
> not help at all, as it may intensify the resonance by adding more
mass to
> vibrate. The professional approach is a viscous damper. Early ones
> were made by attaching a disc to the shaft, and clamping a heavy
metal
> disc to that one with a rubber sheet between. With the rubber
sheet loosely
> clamped, the attached disc would rub against the dead weight and
absorb
> the vibrational energy. getting the clamping pressure right, and
keeping it
> right was not easy. The current approach is to mount a hollow round
> cavity to the shaft, and have a heavy metal disc free floating in
the cavity.
> The cavity is partially or completely filled with a viscous oil to
give the
> desired vibration absorption. I would think it wouldn't be too
hard a
> lathe project to make one of these out of aluminum, with a steel
disc
> inside, and a cover plate with a gasket. SAE 90 gear oil should
provide
> good damping.
>
> Jon