Hello Guest it is March 28, 2024, 04:49:27 PM

Author Topic: Cutting under or over size depending on software  (Read 1135 times)

0 Members and 1 Guest are viewing this topic.

Offline pdh

*
  •  4 4
    • View Profile
Cutting under or over size depending on software
« on: July 02, 2020, 02:34:03 PM »
Hi, I've got a small mill with Mach3 on it (Warco WM16 - Cast iron, weights about 200KG). I've set it all up and if I measure the movement with verniers (with the head pushing them) I'm within 0.02mm which is not fantastic but is OK. If I use the newfangled wizards for pockets, either round or square, they come out under size, sometimes as much as 0.1mm. If I user dolphincam to generate g-code to cut the same shape it comes out slightly oversize, by around 0.05 to 0.1. This is cutting plastic so fairly soft, so I don't think it's flex. If it was, I'd expect it to be the same. Any ideas?

Also, What are typical tolerances for Mach3 and 1605 ball screws? I'm using 1600 steps per rev and NEMA23s. Not sure if that's a good value of not, but gives reasonable speed and should technically get me 0.00312 per step. Any advice welcomed on how to get it better.

thanks Paul.

Offline Graham Waterworth

*
  • *
  •  2,668 2,668
  • Yorkshire Dales, England
    • View Profile
Re: Cutting under or over size depending on software
« Reply #1 on: July 02, 2020, 07:46:23 PM »
If you are out by .02mm in a 10mm movement then you are out 0.1 in 50mm movement, its not hard to set the steps/per if you know all the data which by the look of it you do.

1600 steps per rev of ballscrew at 5mm pitch would be 320 steps per mm for direct drive on the ballscrew

If its not direct drive the the value needs adjusting by the ratio.
Without engineers the world stops

Offline pdh

*
  •  4 4
    • View Profile
Re: Cutting under or over size depending on software
« Reply #2 on: July 03, 2020, 04:41:47 AM »
Thanks. Any idea why two separate pieces of software have different errors?
Paul.

Offline JG

*
  •  61 61
    • View Profile
Re: Cutting under or over size depending on software
« Reply #3 on: July 20, 2020, 01:42:23 PM »
Have you compared the G-Code generated by the two programs against each other?

Is there a difference in the dimensions in the G-Code?

Offline pdh

*
  •  4 4
    • View Profile
Re: Cutting under or over size depending on software
« Reply #4 on: July 23, 2020, 09:54:48 AM »
I haven't but I'm not that clued up on G-code. I'll have a look to see if they look similar.
Paul.

Offline JG

*
  •  61 61
    • View Profile
Re: Cutting under or over size depending on software
« Reply #5 on: July 23, 2020, 11:36:35 AM »
Hi Paul,  I'm also very new to G-Code but my approach is to get an understanding of what G-Code does by writing short and very simple programs by hand to see EXACTLY what happens.

Over the past week I've learned a great deal by making silly errors and having them corrected by the other learned folk on the forum.

If you don't know what (say) G30 or G17 or G54 does then reading through a program generated by some CAM program is not likely to make a lot of sense.


Offline pdh

*
  •  4 4
    • View Profile
Re: Cutting under or over size depending on software
« Reply #6 on: July 23, 2020, 12:26:46 PM »
Yes, probably a good idea. I'm currently at the 'can the machine work properly' stage using software to drive it rather than coding.
Paul.

Offline TPS

*
  •  2,501 2,501
    • View Profile
Re: Cutting under or over size depending on software
« Reply #7 on: July 23, 2020, 03:08:54 PM »
as Graham has written, if you use ballscrews (does not matter chinese or not) you should be in a tolerance
less then 0,001 mm per Revolution. so the first step would IMHO be to figure out, where your 'relative big tolerance' is comeing from. backslash (witch is not realy real with ballscrews) lose main bearings for the ballcrew
... and so on. first of all you have to get a preceise move of your axis (over the full range), and then talk about
tolerances.
anything is possible, just try to do it.
if you find some mistakes, in my bad bavarian english,they are yours.