Machsupport Forum
Mach Discussion => General Mach Discussion => Topic started by: twelti on September 24, 2006, 03:24:00 PM
-
I have a question. I am using mach2 and a taig mill. I am cutting wax for jewelry. So far I have set up the machine and it has worked reliably for several pieces. On the last one though, it abruptly changed its zero reference during roughing. I had one other time when it lost reference during milling. The only thing that ties these two instances together is that I was chanigng the feed rate via the Mach2 feed rate control during the cut. I tend to start at a slow rate and increase up to a higher rate. Or if it is moving to a new location (at the safe z height) I speed it up briefly. Is this a problem?
I have also noticed that sometimes when I zero out the z height, the x and y reference changes. Why is this?
-
Is this in the new 2.0b version of mach3? If so it could be a bug in the Accel . I am working on the problem now
Sorry for the bug
Brian
-
No, actually its mach2. I tried mach 3 but it didn't work so far...
-
O that can't be a bug in mach2 :( I only worked on Mach3
-
If you are changing the feed rate while its running you may be exceeding the acceleration limits of your motors and loosing steps. It has been documented here that feed rate over ride also overrides acceleration and will cause this problem. It is advised to write the program for a high feed rate and override it down when you start up, only taking it to 100% when you are sure its running OK.
-
I usually write the code to move pretty fast (20 mm/sec) and slow it down at the start. Actually I had another go at it, and this time I didn't even change the feed rate. Similar result, after an hour or so of milling using a fine point, it missed some steps or something. Anyway it got missreferenced by about 7 mm in the x axis. I was running at 80% of 20 mm/sec. I will check my driver etc., but it was working fine a few weeks ago. The way it happened, i suspect that the error occured when the tool was picked up and moved to another area of the toolpath. Doesn't it move quicker then?
Aaaargh...
-
I was getting ready to post on the same topic , I am running servos and I am trying
to cut a part and in seems to loose steps the more it cuts .
The part it is cutting has a lip that jumps .0040 and as it hits that shelf it seems to lose a few thous
every time it passes that point .
Can that be something in the dwell time?
It does seem to loose steps as bad if I cut the feed rate down but it would need to slowed down to a crawl
to not loose the steps .
Is that what the dwell time is for ?
Thanks
-
Try increasing the acceleration and deceleration time, it may be skipping on short rapid moves.
Graham.
-
If it only happens after long periods of use, is something getting hot and sticking?
or
is something short of lubrication?
Graham.
-
Interestingly it seemd to happen at (or near) the same place in the toolpath. This is about 1 hour into a 1.5 hour run. Cold be a coincidence.
I checked my settings, perhaps someone can review them...
units are setup in inches ( the toolpath is in mm)
Velocity
x = .4090 in/sec
y = .4031
z = .196
Accel
x = 1.936
y = 3.133
z = 3.304
backlash
x = .007
y = .007
z = 0
Backlash speed % max = 10
min pulse is 1 uS
steps/unit is 32000
y and z axis are reversed (had to do this for some reason to get toolpath to turn out right)
OCX driver test
I get a message "No OCX found in directory" at startup, but it runs ok:
pulse/sec is betwen 25598 and 26000
CPU speed 2590
Apic is present, timing constant 3892
Computer:
2.6 GHz celeron
.5 G ram (.3 G available before running Mach 2)
XP pro SP2
Note i did load a CD rom emulator recently and it was running in the background. Is this an issue? Anyway i removed it.
Any hints would be welcomed...
-
As your velocity in x & y are very similar I would try setting the acceleration for x more like your y figure.
Graham.
-
So, INcreasing acceleration could help?!?
-
not the acceleration but the distance over which it is done. It might need more time to get to full speed. If it can't overcome the mass quick enough then the motors will loose step.
Graham.