Hello Guest it is June 19, 2019, 05:58:47 PM

Author Topic: Help with a Mill  (Read 3578 times)

0 Members and 1 Guest are viewing this topic.

KTB100

*
Help with a Mill
« on: June 17, 2008, 01:26:47 PM »
I have a mill acting funny.  I can use MDI and send the table 10" away from an indicator, come back and it comes back to zero. 
I tell it to feed out 1", then back, it moves 1.000 and comes back to zero. Same for "Y".

While running a program, it seems to be scaling by a factor of 1.027% in "X" and 1.023% in "Y", thus over cutting. 
The scaling is turned off and I feel certain I have eliminated mechanical problems....
Anything else that could cause this? 

It is being programmed using Surfcam with about .002" moves in 3D.... could the control be "rounding up" the 4 place decimal?
Just guessing....

I am using 3000 RPM AC servos with 2500 CPR encoders. Ground ball screws. 

Any help would be appreciated.

Thanks   

Offline Hood

*
  •  25,844 25,844
  • Carnoustie, Scotland
    • View Profile
Re: Help with a Mill
« Reply #1 on: June 17, 2008, 01:49:05 PM »
The control wont round up, well certainly not anything like 0.0001. I would think it more likely that your code is the problem, are there any settings in surfcam for accuracy? Is there a possibility that your cutter is slightly oversize from what you have told SurfCAM it is?
Hood

KTB100

*
Re: Help with a Mill
« Reply #2 on: June 17, 2008, 02:14:33 PM »
Surfcam only posts G01 G00, we insert G43, T, H, etc.  no cutter comp used.  This exact same program has run in a different machine with no problems at all.  Cutter is 1/4" ball mill.  It is cutting .100" too much over 6" but as I mentioned with the indicator, it will come back to zero....... This is what confuses me.   

Offline jimpinder

*
  •  1,233 1,233
  • Wakefield, West Yorks, UK
    • View Profile
Re: Help with a Mill
« Reply #3 on: June 18, 2008, 04:01:26 AM »
You do not say if this is an old mill that has just developed a fault , or a new mill you have just set up.

The fact that you can send the table out X inches and bring it back to zero makes no difference to Mach 3, since any faults will be cancelled out in the return.

What you should do is move right 1/2 inch, to take up any backlash, set up a DTI then send the machine 1 inch to the right and see what that distance is. If it is incorrect, send the machine another inch to the right and check that. If the fault over 2 inches is double the fault over one inch, then your steps per unit are not correct. You can carry on checking up to the extent of your DTI or calipers.



Not me driving the engine - I'm better looking.

Offline Hood

*
  •  25,844 25,844
  • Carnoustie, Scotland
    • View Profile
Re: Help with a Mill
« Reply #4 on: June 18, 2008, 04:19:53 AM »
Yes Jim is right, I never picked thgat up. When I read it I thought you were measuring the first travel and it was correct then you were coming back again to measure the backlash. Measurement needs to be taken in one direction and also even though you have good ground ballscrews with little backlash there may be backlash in other  areas such as the  bearings or brackets. However your return to zero test has shown Backlash is not your problem but you need to confirm the initial movement as Jim suggested.
 Hood

KTB100

*
Re: Help with a Mill
« Reply #5 on: June 18, 2008, 10:46:48 AM »
Thanks for the reply.

This is a brand new mill.  The lead screws taken out have never cut a part. 
I used a 1" indicator as you suggested.  In MDI, told it to travel 1.0000"  it did exactly that.  I commanded X-.250"
it traveled -.250".  I then centered a part and cut a 2" square part.  "X" axis measured 2.0001" and "Y" 2.0000
I then cut a pocket 10" in "X" and 4" in "Y".  it measured 10.001" x 4.000.  (using 12" calipers).  The problem only seems to
exist when I use 3D programming with the small moves surfcam puts out......

Can the servo drive acceleration  / deceleration setting can cause this?

Offline Hood

*
  •  25,844 25,844
  • Carnoustie, Scotland
    • View Profile
Re: Help with a Mill
« Reply #6 on: June 18, 2008, 02:08:45 PM »
When you are doing the tests are you doing it with a G1 or a G0 move?
What is the following error on your drives and is there any way of monitoring what error there is? Sounds like your motor tuning may not be perfect and with the quick acell/decell of the 3D file you are getting a following error.
 What is the actual error you are getting when doing the 3D?
 

Hood

KTB100

*
Re: Help with a Mill
« Reply #7 on: June 18, 2008, 06:21:40 PM »
I have not tried it with a G0.  All has been done with G1. 
AAAHHHH!!!!!!     
I went to go see what the following error was set to and while moving the cable to the Breakout Board to see the drive, it fell out of the socket.  I am out of material for the moment however when I pick some up tomorrow, I think I am going to see a different machine.  Man....I thought I had checked them all.....  We will see.

Offline Sam

*
  • *
  •  988 988
    • View Profile
    • hillbillyhilton.com
Re: Help with a Mill
« Reply #8 on: June 18, 2008, 09:28:17 PM »
Sounds to me like you've got a ricket problem. By any chance do you live in Canada? Rickets are bad up there this time of year. Soon as you turn your back...WHAMMO !!! they do some dastardly deed.
"CONFIDENCE: it's the feeling you experience before you fully understand the situation."