Hello Guest it is August 03, 2021, 11:16:39 PM

Author Topic: HELP.........Motor Tuning problem  (Read 6523 times)

0 Members and 1 Guest are viewing this topic.

HELP.........Motor Tuning problem
« on: February 11, 2007, 10:43:11 PM »
Hello,

I'm using Mach 3, and I've got a problem in which I have no
idea what to do......

My X-axis has been a little off ever since I built my small
router, I just never fixed it because it wasn't enough
to worry about (like 1/16" on a 29"span)

So finally I decided to fix it because I need to
machine some close tolerance aluminum parts. So I got
it accurate within a few Thousandths over a 29" span.
I figured the closer I had it over the longest span,
the smaller the margin for error on a shorter
distance. Right?

That's not proving to be the case though. Its actually
MORE off on a 17" span now (like maybe a 32nd) than on 29" (few thou). How can that be? Any idea?? In reality I need it closest on the
smaller parts. Thanks in advance for any help! I'm using a Ballscrew with Zero backlash by the way.

Best Regards,
Nardis
270-576-7030

Hood

*
Re: HELP.........Motor Tuning problem
« Reply #1 on: February 12, 2007, 03:02:04 AM »
Check for backlash in the bearings also is there any chance that your setup on the X Axis is flexing?
Hood

Offline Graham Waterworth

*
  • *
  •  2,348 2,348
  • Yorkshire Dales, England
    • View Profile
Re: HELP.........Motor Tuning problem
« Reply #2 on: February 12, 2007, 08:32:54 AM »
It sounds like a steps per issue to me,  what is the pitch of the ball screw and the ratio of the motor to the screw, also how many pulses per rev are the motors.

Graham.


Without engineers the world stops
Re: HELP.........Motor Tuning problem
« Reply #3 on: February 12, 2007, 10:10:58 PM »
how are you measuring that exact over sucha abig distance?
Re: HELP.........Motor Tuning problem
« Reply #4 on: February 13, 2007, 01:48:48 AM »
Thank you all VERY much for your comments! As I was finishing up a nice long paragraph regarding all the perimeters of my setup process and machine design, I noticed that the ruler I was using in my office to measure parts was coincidently about a 32nd off from my steel rules in the shop. What a piece of crap!! I measured my parts with the same steel ruler I used to set up the machine and they were dead on.

Lesson learned... NEVER EVER ASSUME a measuring device is accurate. I was fit to be tied. Tomorrow I'm ordering a 24 and/or 36 inch Starrett tempered steel scale. Thanks again for the input.

Best Regards,
Nardis
Re: HELP.........Motor Tuning problem
« Reply #5 on: February 13, 2007, 02:00:38 AM »
get a digital caliper....works better ;D
Re: HELP.........Motor Tuning problem
« Reply #6 on: February 13, 2007, 03:10:46 AM »
You mean for setting up and tuning the motors?? Should I cut a profile and measure it +/- 0.001?? I already have some accurate 6" dial calipers. I need a longer rule for the type of design work I do anyway. Thanks for the suggestion.

Nardis 
Re: HELP.........Motor Tuning problem
« Reply #7 on: February 13, 2007, 11:01:53 AM »
well, when i check accuracy on a machine I dont check it over long distance, I will go and cut something that i can measure with a very precise measuring device (like digital or dial calipers, i still rather use digital). Also remember that when cutting somethin with a milling bit, say if your using a .250" bit the bit doesnt have to be ".250 it can be smaller (wont be bigger though). If you have any double flute milling bits i would do a test cut with that because you can actually measure the diameter. Anyways thats what i would do.

edit: after reading that i realized I have to add that after checking over short distance and tuning. I then would check over long distance. I gues you could do it with a precise steel ruler (starret or something). If it looks dead on it should be ???
« Last Edit: February 13, 2007, 11:06:08 AM by xjdubber »

Offline Graham Waterworth

*
  • *
  •  2,348 2,348
  • Yorkshire Dales, England
    • View Profile
Re: HELP.........Motor Tuning problem
« Reply #8 on: February 13, 2007, 02:57:07 PM »
Hi xjdubber,

I will apologise now for what I am about to say.

Sorry but I don't agree with what you are saying.

1. cutters can be over size, they can run out of true in the collet, the head may be out of square so the sides of the cut are tapered.

2. depending on the type of machine configuration cutter pull and push can come into play.

3. less than perfect cutters can cut under size in certain materials and over size in others.

4. when you are setting up a machine with a working area of 29 inches test cutting some thing an inch in size is not going to tell you the thing is stepping accurately over the entire length of travel.  A 1 thou error over 1" is nothing but multiplied over 29 inch is rubbish.

5. if a machine is setup correctly using ballscrews the error over 36" should be no more than .006" with a standard grade ballscrew and no more than .001" with precision screws.

6. the machine should be clocked square and level, the table should be trepanned with a clock on a long arm mounted in the spindle to make sure the head is square to the table in both x and y, any slop and back lash should be removed and then the machine should be set up using a .0001" resolution dial gauge and slip blocks to get it to position accurately. The steps per unit should be to as many decimal places as possible, more the better. 

7. if after all this is done, the machine is not accurate then the construction of the machine is too flimsy for the work to be done and should be beefed up or redesigned.

Graham.
Without engineers the world stops
Re: HELP.........Motor Tuning problem
« Reply #9 on: February 13, 2007, 08:32:45 PM »
Hi xjdubber,

I will apologise now for what I am about to say.

Don't apologise  ;D

Agree with you on every point, except for the first (and some others:)). Im sorry but I've never seen an over size cutter yet? as for the rest, I agree but then those points should already have been taken care of before setting accuracy. As for the acuracy on the ballscrews that really depends on how new they are. Also squaring the spindle is not going to change the lenght of a cut. On your point #4 what do you mean 1" even normal size digital calipers go up to 6", You think i would measure error out of a 1" cut?

I have setup 6'x20' routers, that run anything up to alluminum that hold tolerance of .005 over 10 feet, so far the way i do things has been working for me ???
« Last Edit: February 13, 2007, 08:46:32 PM by xjdubber »