Thread: Mach 3 / CS Labs Smoothing
-
21-08-2019 #1
On my concrete mill (Thor) I run CS Labs IPA with Panasonic servos. Ive always noticed that on circles and similar I get jerky movement, specifically where there are I and J type G Code involved.
I did a test last night, took the same code, ran it on my Bridgeport, running on Centroid Acorn, same Panasonic setup but in Step / Direction (Position mode) versus Speed move as per Thor. The same code was smooth.
Ive seen a lot of complaints about Mach 3's smoothness.
Is my issue a limitation with Mach 3 (or badly configured), related to not having good PID on the servos or something else?
I might try the new SimCNC software from CS Labs to see if its Mach 3 or not. A newer PC might also help.
Thoughts?Last edited by Chaz; 21-08-2019 at 09:52 AM.
-
-
21-08-2019 #3
-
21-08-2019 #4
I'm looking at a screenshot of the General settings page, I dont see it. Any idea where it is?
https://www.machsupport.com/wp-conte...all_Config.pdf - page 5-35 is the General Settings Page.
-
21-08-2019 #5
I use Mach3 and an IP/M on my router and have never noticed any jerkiness on curves. Don't know if that is luck or the right settings! I've been using F360 with various cutting strategies recently, some of which use a large number of short straight line segments which Mach3 seems to handle ok. So it seems ok whether or not G2/3 or straight line segments are used. I am using steppers rather than servos and it does seem possible that servo tuning would have an effect here.
-
21-08-2019 #6
-
21-08-2019 #7
I was thinking of the case where F360 is approximating non-circular curves (e.g. adaptive clearing strategies) using many very short straight line segments so there are a lot of rapid direction changes. Mach3 is supposed to smooth these, as I understand it. This kind of tool path is bad news if Mach3 is configured to "exact stop" mode but in constant velocity mode it's fine.
On reflection, though, probably not very relevant as you are using analogue servo controls which introduces a whole bunch of other variables. Good luck!
-
21-08-2019 #8
-
21-08-2019 #9
Have a look at 5.6.6.2, I/J Mode: You can set the interpretation to be placed on I & J in arc moves. This is provided for compatibility
with different CAM post-processors and to emulate other machine controllers. In Inc IJ
mode, I and J (the center point) are interpreted as relative to the starting point of a center format arc.
This is compatible with NIST EMC. In Absolute IJ mode, I and J are the coordinates of the center in
the current coordinate system (i.e. after application of work, tool and G92 offsets). If circles always
fail to display or to cut properly (especially obvious by them being too big if they are far from the origin),
then the IJ mode is not compatible with your part program. An error in this setting is the most
frequent cause of questions from users when trying to cut circles.
Phill
-
The Following User Says Thank You to phill05 For This Useful Post:
-
21-08-2019 #10
Thread Information
Users Browsing this Thread
There are currently 1 users browsing this thread. (0 members and 1 guests)
Similar Threads
-
cs-labs.eu website problems
By chillybo in forum CS-Lab (CSMIO)Replies: 2Last Post: 15-04-2019, 05:42 PM -
FOR SALE: Cs labs csmio ip-s
By Web Goblin in forum Items For SaleReplies: 3Last Post: 22-07-2018, 08:45 PM -
FOR SALE: Denford Triac VMC with ATC / Servos / Mach 3 / Single Phase / CS Labs
By Chaz in forum Items For SaleReplies: 10Last Post: 28-08-2017, 10:15 PM -
Denford Orac and CS Labs controller
By swdiver in forum Denford LathesReplies: 4Last Post: 07-08-2017, 10:59 PM -
FOR SALE: PSU Rectifier + Capacitor smoothing board for toroidal transformer
By Jonathan in forum Items For SaleReplies: 3Last Post: 02-02-2017, 10:47 PM
Bookmarks