Re: [CAD_CAM_EDM_DRO] Positioning

on 2014-06-03 10:28:10 UTC

In following this dialog about accuracy, it seems that glass scales can have claimed resolution to 0.01 micron (see page 8: http://www.mitutoyo.com/wp-content/uploads/2013/11/E13005_LinerScaleSystem.pdf ). In other words, 10 nm. I guess that level of precision may be important for manufacturing some modern electronic chips and other positioning, but it is more than most of us need for cutting metal or wood on a mill.

The light we see has a wavelength of 400 to 700 nm (i.e., 0.4 to 0.7 micron), approximately. The wavelength of X-rays is generally considered to be 10 nm or less. In other words, you can’t “see” the difference, unless you have X-ray vision. Fifty microns (about 0.001”) is more than adequate for most of what I do. When pushing it, maybe 5 microns (0.0001”) is needed.

Maybe getting back on or off subject...

I have read on another forum that retrofitting ball screws to a convention mill, say a Bridgeport, reduces the power needed to move the table – both by the operator and by the cutter. One comment was to the effect that with ball screws, hand machining became quite difficult. What I am interested in is how does one determine the necessary holding torque of stepper motors (or other drivers) to avoid the table being moved by the cutter?

Regards, John

From: mailto:CAD_CAM_EDM_DRO@yahoogroups.com

Sent: Tuesday, June 03, 2014 4:10 AM

To: CAD_CAM_EDM_DRO@yahoogroups.com

Subject: Re: [CAD_CAM_EDM_DRO] Positioning

The specs on the screw accuracy are irrelevant.
What IS relevant is that there be no slop, as a machine with dual-feedback, ie glass scales, will shake itself to pieces is it has backlash.

Thats why you cannot use the crap chinese scales, because they are off, ie inaccurate, by upto 0.00" or 0.04 mm, in short distance, and the error varies.
So using such, would make the machine jitter (sometimes, in some places).

This causing back-and forth motion, upto hundreds of times a second, depending on how fast the secondary positoning error-correction mechanism is.

It is impossible to make a (working) secondary-feedback mechanism if you have backlash in the system.

Tony..
Unlike some people here, I try to be factual and polite.
Grin.

Re: screw accuracy.
Everything is inaccurate. Its just a question of how much.
One good part about glass scales, is that they are usually accurate (to 1 micron, or down to 0.01 micron with some Mori Seiki machines).

Sure, when you read the specs on the average screw, and that "Dummies Guide To Fantasy Machining" you've got.

Do you know what a drunken thread is?

Do you know what a temperature coefficient is? (Go work out the expansion on a 600mm steel screw for a 1 degree delta.)

Tony

(You and your microns, lol)

-- -hanermo (cnc designs)

Discussion Thread