Re: [CAD_CAM_EDM_DRO] Re: dual function lathe spindle

This sounds like a possible use for one of those in-wheel motors used for electric cars like PML Fairlight makes - I believe they are servomotors and torroidal to boot- could give you a really nice big spindle thru hole with tons of torque I think. The large perimeter also could do high resolution encoding. Because of the large speed range I reckon you would want two encoder strips- one high resolution for low speed, and one low resolution for high spreed so the electronics could keep up. Make sense?

----- Original Message -----
From: Jon Elson
To: CAD_CAM_EDM_DRO@yahoogroups.com
Sent: Friday, July 30, 2010 1:02 PM
Subject: Re: [CAD_CAM_EDM_DRO] Re: dual function lathe spindle



scyvt wrote:
> Jon, Peter, et al
>
> So, I take it that the consensus is that a single drive CAN perform both functions, perhaps with a brake added for good measure.
>
> Then, for a lathe that originally driven by a 1 hp induction motor thru a variable speed belt reduction drive (a Rockwell 11"), in your opinion what would be the most suitable servo motor, encoder, and servo drive components, assuming EMC2 as the motion controller. (You may pitch your own products!)
>
>
Well, it isn't so simple. But, I'll give it a try. First, you can't
have a gearbox between the motor and the spindle. The backlash in the
gears would prevent it from working. So, you need to use belts, and
they have to tight enough to not slip. Maybe timing belts would be the
right thing, here. A stack of several timing pulleys would be needed to
accommodate the speed range. For the same gear lash reason, you can't
use a back gear.
Then, you need an encoder directly driven by the spindle, so it always
has a 1 pulse/rev index signal for threading. Finally, you probably
need a motor MUCH bigger than normal to provide sufficient torque at
zero speed without overheating. What are you going to do with this?
Will there be a mill head with XY axis that can mill on the workpiece
mounted on the chuck, or will you just be drilling into the workpiece?

You could use any eBay encoder with index pulse and good bearings.
Drive it off the spindle with a 1:1 set of timing pulleys and a really
small belt, Get the encoder as close as possible to the spindle to
reduce flapping of the belt.

Depending on the motor selected, my PWM servo amps might be suitable,
and my PWM servo controller. You would need a little bit of HAL code to
make the switch from C axis to spindle, I think. We have batted this
around a number of times on the EMC lists, and I don't think there is a
definitive solution. One of the problems is when you run as a spindle
for a while, the C position winds up to thousands of turns. Then, when
you go back to C axis, you;d want to re-index the count to somewhere in
the first revolution - kind of like re-homing. This could be done
automatically all in HAL, by turning on the spindle's index-enable bit
and turning the motor until it passes the index location. When in
spindle mode, you want to disconnect the encoder from the C axis logic
so it doesn't cause a following error.

I have heard rumors, but I don't know if anyone is actually using a
setup like this.

Jon





[Non-text portions of this message have been removed]