Author Topic: Mechanical corrections from software  (Read 8769 times)

0 Members and 1 Guest are viewing this topic.

Offline hunserv

  • Active Member
  • Posts: 77
    • View Profile
Mechanical corrections from software
« on: November 21, 2009, 06:58:28 AM »
Hello,

The recent days I was wondering about a few features which have been mentioned earlier, but I do not know if it is possible to realize them.

Background story:
I have a High speed CNC machine with precision linear guides and ballscrew, integrated cooling for thermal stability, etc... However, this is not new.
Screws are renewed, accuracy is as it is... (Not the highest grade)
Linear guides are OK, but it had once a crash, and by measuring axis perpendicularity, at one point one axis has a small bend at a certain range.

I'd like to get some answers on 2 of them:

1. Screw mapping.
How is it nowadays? How well does it work?
Can I do the followings:
I mount an accurate measuring device on the axis, and go through the whole range of the axis. By comparing actaul and theoretical positions, and making a deailed screw map for each axis fore each "step" I could make it more accurate without mounting expensive glass scales...

2. Axis "linearization"
This one is for the bent axis...
Solution 1:  I change the wrong axis's components (quite expensive solution)
Solution 2: make comepnsation from software... Is it possible to somehow do this?
e.g give offset values for axis X when axis Y is at the range of 200-300... even different at different Z levels...
The offset equation is quite easy to define, I can make measurements to know the exact amount of deflections...
If Mach keeps track of absolute positions, then it should not be too difficult, no?

Thanks in advance!
« Last Edit: November 21, 2009, 07:02:17 AM by hunserv »

Offline Dan13

  • Active Member
  • Posts: 1,244
    • View Profile
    • DY Engineering
Re: Mechanical corrections from software
« Reply #1 on: November 21, 2009, 07:46:40 AM »
As far as I know, screw mapping in Mach3 should work fine. Never heard of any issues reported regarding it. So this one should be no problem if you say you can make the measurements. BTW, I am very interested how you can make the precise measurements for the whole axis range without a glass scale. I think it will be very time consuming to do the measurement without a glass scale and some kind of auto position feedback.

Regarding your second question, I am not sure how to do it...

Daniel

Offline ger21

  • Global Moderator
  • *
  • Posts: 6,127
    • View Profile
    • The CNC Woodworker
Re: Mechanical corrections from software
« Reply #2 on: November 21, 2009, 10:06:16 AM »
Try using the axis formulas to correct for the linearisation. I've done it with an out of square gantry and it worked fine.
Gerry

2010 Screenset
http://www.thecncwoodworker.com/2010.html

JointCAM Dovetail and Box Joint software
http://www.g-forcecnc.com/jointcam.html

Offline hunserv

  • Active Member
  • Posts: 77
    • View Profile
Re: Mechanical corrections from software
« Reply #3 on: November 21, 2009, 12:40:47 PM »
Dan, I`d do the measurements with a borrowed glass scale... Or borrow an interferometer or whatever from a tech. university nearby :) whatever, measuring it once would cost me far less than buying one single scale (and for 3 axis I`d need 3)

No reports on screw mapping might be due to noone is using it?
Whatever, then when time comes I will try it, and give report if I find problems. (but before the measurements I should make some dummy trials to see how it works for real)

Ger, thanks for the hint, I`ll have a look at that!
 

Offline Dan13

  • Active Member
  • Posts: 1,244
    • View Profile
    • DY Engineering
Re: Mechanical corrections from software
« Reply #4 on: November 21, 2009, 01:06:34 PM »
Dan, I`d do the measurements with a borrowed glass scale... Or borrow an interferometer or whatever from a tech. university nearby :) whatever, measuring it once would cost me far less than buying one single scale (and for 3 axis I`d need 3)

Unless you have good home switches and don't move them under any circumstances or find another way of tying the measurement to the ballscrew, you might need to borrow the glass scale again ;)

No reports on screw mapping might be due to noone is using it?
 

I think it was on the other Mach group that I saw posts from guys who were (successfully?) using screw mapping with Mach3

Offline hunserv

  • Active Member
  • Posts: 77
    • View Profile
Re: Mechanical corrections from software
« Reply #5 on: November 21, 2009, 02:11:39 PM »
Well, the machine I want to use it with has industrial home + limit switches, plus index on the servos, so that would be not an issue, if service, or maintenance is needed (e.g. on bearings) I am aware that re-measurement might be needed...

Offline rgsparber

  • Active Member
  • Posts: 45
    • View Profile
Re: Mechanical corrections from software
« Reply #6 on: October 31, 2015, 07:28:41 PM »
I got screw mapping to work but it required a few work-arounds because of various bugs. See http://rick.sparber.org/ma.htm#13 for two screw mapping articles. It sure would be nice if someone would fix these bugs!

Rick

Offline BR549

  • Mach4 Alpha
  • Posts: 6,855
    • View Profile
Re: Mechanical corrections from software
« Reply #7 on: October 31, 2015, 09:58:01 PM »
Sorry fellows but with Mach3 NOTHING will ever be fixed beyond what it is now . It will Be Mach4 from now on. Or whatever you can figure out and create a work around in Mach3.

So you MAY want to make your OWN depository of versions to work with JUST in case in becomes unavailable in the near future.  

An out of square correction would a simple linear equation BUT a bent (curved) axis would be hard to do without a table lookup. I am not sure you could do it with a simple equation.

I would seriously take a look at fixing the mechanical problem . linear ways Can be carefully  straightened (;-)

YEP Bummer, (;-) TP

« Last Edit: October 31, 2015, 10:08:05 PM by BR549 »

Offline rgsparber

  • Active Member
  • Posts: 45
    • View Profile
Re: Mechanical corrections from software
« Reply #8 on: October 31, 2015, 10:23:12 PM »
Don't get me wrong, enough of Mach3 works well that I am happy with it. But I now see that this program was sold to me after it was no longer supported and, in fact, is still being sold with no warning. I don't know if it is illegal to this but it is not honorable. Mach4? Not likely.


Offline RICH

  • Global Moderator
  • *
  • Posts: 7,302
    • View Profile
Re: Mechanical corrections from software
« Reply #9 on: October 31, 2015, 11:40:36 PM »
You can use optics to measure long axis travel.
I use a jig transit ( 40x) which has an optical micrometer and can measure to 0.0001" from 3' away.
A "calibrated" optical scale ( could be 1' long but a 3 or 4 foot long scale is nice) is used which is set perpendicular to jig transit sight  line.
Thus you are measuring movement to a standard, is accurate,  and can be rather easy to do.

Just a way if you can borrow the equipment,

RICH