Re: [CAD_CAM_EDM_DRO] Re: Power Supply Circuit

Well, I have horrible power quality here and often ride through dips as
well. A computer
without an ups simply will not function around these parts.

What is most important to me in a power dip or loss is not to have continued
machine
motion after the SPINDLE has wound down. It could coast down pretty fast if
it was
doing a heavy cut during the outage. Driving the non-rotating tool into work
with
1000 lb+ forces is going to make real bad things happen before amp
overcurrent
kicks in. I have to keep that overcurrent setting pretty high with a low
time constant
to get the accelerations I need.

Now if the servo supply drops 3 time constants in .25 sec it is more likely
that the spindle
will still be turning when following error initiates an e-stop. That's good
for me.

The bad thing is that even a minor power glitch(something I could ride
through) will halt the machine
and require a restart.

Seems like power failure supply discharge could easily be a switchable
option, perhaps with
selectable time constant.

What I think I really need and might install is some kind of spindle force
sensing system.
That would cover what I described. I need the high forces to accelerate the
axes,
but don't want more than about 100 lb force on the spindle.

Let's see now.... strain gages on spindle mount? Nah too complicated.
Sensing servo
currents? No, those currents could be for acceleration forces. Monitoring
tach or back
emf also? Well, maybe, but I often do a max acceleration from a standstill.
But..... what if there
was a cheap 3 axis accelerometer on the spindle motor area? I have played
around with these
for years- the AD ones. If you get high servo torques (currents) without
corresponding
high accelerations what is left? Friction and spring deflection...and
perhaps a stationary
tool running into work is the ultimate manifestation of that.

magnitude of all servo torques=C1( Ma + spring forces +friction forces )(the
classic 2nd order diff equation)

magnitude of accelerometer signal= C2Ma

Subtract and estimate transmission friction forces (easily measured)


So I will think about that.... the hypothesis is that the acceleration
magnitude measured
at the carriage (assuming it moves as in a gantry) AND the magnitude of all
servo
torques is enough information to simply calculate an estimate of spindle
force.

Les

Leslie Watts
L M Watts Furniture
Tiger, Georgia USA
http://www.rabun.net/~leswatts/wattsfurniturewp.html
engineering page:
http://www.rabun.net/~leswatts/shop.html

----- Original Message -----
From: "Jon Elson" <elson@...>
To: <CAD_CAM_EDM_DRO@yahoogroups.com>
Sent: Saturday, June 15, 2002 1:50 AM
Subject: Re: [CAD_CAM_EDM_DRO] Re: Power Supply Circuit


> mariss92705 wrote:
>
> >
> > What it does is sense the 120Hz cap charge restoring current pulses.
> > If these are absent, such as when the AC power is shut off or if the
> > cap is being backdriven by a drive, a power MOSFET crowbars a wire-
> > wound power resistor across the supply. The resistor is sized so that
> > it draws the full current rating of the supply thru it.
>
> This might make it impossible to ride through even small power dips.
> I had a power outage one night while running my servo-controlled
> milling machine. It was dark long enough that I started reaching for
> where I had a flashlight stored. The lights came back on, and the
> machine was still cutting metal. The spindle VFD, servo drives, and
> computer all rode through the interruption, which had to be a large
> portion of a second in length. I'd hate to give up that capability.
>
> Jon