Machsupport Forum
Mach Discussion => General Mach Discussion => Topic started by: Workinforwood on October 23, 2011, 09:42:03 AM
-
I am having a problem where the tool goes crazy. It was working, but I've been playing a bit with the settings because I could not physically see the depth, the Z was not showing it even though it was there. Now the Z is showing the depth and it is the correct depth on the first pass at least. I have 4 passes to do. The machine ran the entire first pass perfect. When it started over at the beginning, the first pocket went perfect. Then it jumped to second pocket and it went deeper than in the first pocket..that's where the trouble began. Then it came up but did not come up high enough, probably because it was extra deep in second pocket, it shot over to 3 pocket scoring across the top of the piece ruining it and tried to go triple deep, busting the bit. I know for sure, the programming is good. I have had a successful cutting. I know for sure the a axis is tuned and the z axis is tuned and the x axis is tuned perfect.
My guesses of the problem might lay in one or both of these areas.
Safe Z setting perhaps.
Says allow safe z, no
safe z machine coordinates yes
safe z work coordinates no
safe z incremental rise no
safe z is zero
go safe z when stop button hit, no
Toolpath settings
Yes, 3d compass
Machine boundries yes
tool position yes <<<since checking this box is when the z axis reading started to display I think
Axis rotation is checked X axis
A rotations enabled is yes
use radius feedrate is yes, and in settings page the radius of the rod is entered correctly.
-
Then it jumped to second pocket and it went deeper than in the first pocket..that's where the trouble began. Then it came up but did not come up high enough, probably because it was extra deep in second pocket, it shot over to 3 pocket scoring across the top of the piece ruining it and tried to go triple deep, busting the bit.
Hi, I like to try to look at the easy user mistakes first (how do you think I got this way? :D ).
Did it show going deeper on the Z DRO ? I've had this happen where the DRO was correct but the cutter sucked/pulled out of the collet and therefore was lower in Z height. Had you checked the Z height of the cutter after this run?
Dave
-
Sounds like your Z axis is losing position and getting progressively deeper. Try lowering your acceleration settings by 25-50% for the Z axis.
-
Do my posted setting look correct?
The tool did not suck out at all. The machine said it was at .275 on the second pass the entire time. The Z axis certainly was getting progressively deeper even though it said on the screen it was constant, and only did this on layer 2 not layer 1. I can certainly try lowering the acceleration setting on the Z. I will do that. I can change the toolpath and safe Z setting too if you guys think any of that is wrong. I can't remember the setting exactly as to what it was on the successful one, I do not think "A rotations enabled" was checked as a yes.
I probably have to wait till tomorrow to reload, as all my bits are trashed but I do have more on order coming real soon. Last 5 attempts have ended pretty much like this. It's a little frustrating, seeing it go from failure to success and then back to failure.
-
Could it be something really simple like the tool is not properly locked into the collet and it is working it's way out during use ?
Tweakie.
-
I sure wish it was as simple as the tool coming loose. I can repeat the error over and over again, happens at start of layer 2 every single time in the same place. The tool is secure, the workpiece is secure.
-
happens at start of layer 2 every single time
Nice if it's repeatable, since if you do as Gerry replied,you may see a change for the better.
At least you know where and when it is happening.
Post the gcode file and you xml.
RICH
-
I'd be happy to do so, how do I upload the NC file?
-
Click the Preview button, when the screen refreshes click the Additional Options button, the click the Browse button to select the file you want to post.
RICH
-
ok, sounds good. The file is .5 mm flat mill bit, .3125 radius <.625 diameter rod>
I tried cutting again with z axis reduced by 50%. I only made it 80% through layer one. Then the machine stopped rotating. It was moving the numbers for rotations, but not actually rotating, when the rotation numbers would stop, then x would kick in and it would go across, then do the a rotation again but not physically moving, then x would move across again, over and over again.
I checked diagnostics, had yellow home light on. I went to inputs and checked off A home output. Light went away. I jogged bit up. I jogged A, it won't move up. I jogged a down, moves down fine. I sped up jog rate. I can jog down the a axis all the way around and keep going. Then I can jog it all the way back up around until it gets to 315 degree mark, then it won't go more. It will go all the way around one way but not the other. I checked homing limits, A is set at 9999, so that looks good to me, kinda says to me there is no limit on A. But A is acting like it's at the end. If I jog Down the A all the way around and keep going, so lets say 500 degrees. I can jog it back up 500 degrees until it gets to that 315 point and then it's stuck. Makes me wonder, is there a limit as to how many rotations the A can go physically inside the unit? Like it's got a worm gear inside that is at the end, and so I have to just jog it back a couple full revolutions to get that gear in the middle? Makes sense but makes no sense to me at all, I would think the A could rotate in either direction indefinitely.
-
Using the possible theory that the rotary table has a physical limitation as to how many times it can rotate, I have backed it up 2000 degrees, reset everything to zero and am now running again to see what happens. I found that by rotating back 2000 degrees, I could rotate forward 2000 degrees before it gets to that spot it won't go any more, so maybe that will solve this new problem. I do know that after ever failed test cut, the machine did rotate, so it is possible that after many attempts it has rotated forward to the point of no more forward if that's a physical possibility.
-
9999 is only 27.775 revs between min and max softlimits , you may have the soft limits enabled on rotational axis, its General Config page. Remember softlimits are Machine Coords so it may look like you are no where near them but you are.
Hood
-
yes, i had seen that setting and it is not checked. So there should be no soft limits on rotation. And in the limits area it is 99999, 5 digits. A mechanical limit inside the rotary table is all i can think of that would stop it. It will make many rotations before it finally won't continue. The degrees of rotation are accurate, in settings I can ask for 100 degrees and it hit is dead on. So that is good. I had to start over after messing with the rotary table stopping so I do not still know if layer two will be botched or not.
-
IF you can make more than a 360 deg move then there is no hard limit on A. It would not make any sense to have one. The theoreticle limit is like a zillion revs in each direction. WAY UP THERE. So I dought you have hit it.
One thing to remember when working with the A axis is SOMETIME you are going to have to unwind all the degrees you called for. Either by runing the code or refhome the axis to reset the machine A0.
In the old days you only had 0-360 degrees in a rotation and you had to account for that in code. Today you have it better with Mach3 letting you wind up the axis.
HOW do you have Mach setup for on the A is it 360deg or continous roll over????
Set up your machine then do a test. Refhome the A axis so the machine dro vlaue is Zero and then zero the work dro. Now the A is at machine home and work home.
Now from the MDI try a Gcode like G00 A 3600
The A should rapid around 10 times in the positive direction.
BUT even then if you are doing carving on a tube you should never go beyond 360 anyway(;-) (oldschool)
I think yo have something going on in your MachinePC that is causing MACH3 to stall or hang up from time to time.
Easy enought to test post a copy of the Cut file and we can test the Gcode.
Just a thought, (;-) TP
-
Ok, i will try that, thanks. I did post copy of the file, check on page 1 you will find it.
-
You could be right about the PC too. I had it built for this purpose. Last several times when I was done in shop, i hit windows button to shut down and the computer goes into some hardware error, or hardware set up error. I should have that looked at, it's brand new, but perhaps some setting got messed up. I just figured because the CNC is hooked up to the computer it is what's causing the thing to flip out. I have had Mach 3 crash a few times with cnc running but not moving and I couldn't even exit the program, had to do a hard shut down on the PC. But..only had that happen a couple times, not all the time.
-
None of the 3 rotational option boxes are checked in the gen config. so limits is off, so is continuous roll over. I dare not mess with any setting right now while it's running. Machine is 80% done on level one and part is absolutely perfect so far. Hopefully when it starts level two it won't do what it has been doing that last several times that started this post. I'll know within an hour. It's running real slow because the bit is a little smaller than before. but, the smaller bit will do the details better. This dragon is really fine with tiny pockets.
-
What version of Mach are you running. I just tested some Gcode in V.049 and there are problems with ABS(G90) Gcode on the A axis.
(;-) TP
-
What is the problem with it, I do not know what ABS G90 is? Is it the rotation center not lying on the X or Y axis? In visual mill, it gives me this error, but according the the support staff it is correct with this error. It is because the top of the rod is at Z0 and the center is at X0 and Z-radius. So in the posted code the center of the rod is x0 and Z -0.3125 , the software asks if this is ok will the machine handle these conditions and I say yes, because the people at visual mill say it is correct and runs perfect on their equipment. When I look at the machining conditions in the software that created the code, it shows the true center with a post and revolves correctly. I will go look at what Mach version I am running.
-
I can not see how to tell what version I have. I only know I downloaded it from Artsoft 3 months ago and I have the registration codes for it from Syil America. I am using windows 7.
I just went and checked machine and it's all messed up different now. On this test run, The change was to reduce acceleration of Z. The result was that the error that kept happening before no longer happened. Now it's a completely new problem. It looks like the machine did not start layer 2 in the exact same spot as layer one. It is maybe a 1/16th inch off, no way to be exactly certain just how off. Of course, I checked the rod and it is solid in the chuck and can't be moved, and the bit is still solid. It is cutting layer two just slightly over from layer 1 now, so the piece is destroyed. uggh!!!
This is what it is supposed to look like. How it ever came off the machine like this is beyond me, cause I can't repeat it for the life of me!
(http://i121.photobucket.com/albums/o202/workinforwood/alumidragon1.jpg)
This is what that one good one I made looks like when cast in resin.
(http://i121.photobucket.com/albums/o202/workinforwood/alumidragon1-1.jpg)
(http://i121.photobucket.com/albums/o202/workinforwood/alumidragon2-1.jpg)
-
Reduce the accelleration on xy as well. Sounds like you are loosing steps on them as well.
(;-) TP
-
Help / About .... for the version.
-
ok, version is vr3.042.038
x was 31,31
y was 100,40
I changed the motor tuning on both to 31,20 now
I will reload a rod in the morning and see if that helps, of course y wasn't running anyhow, but x was and I reduced the acceleration on that by 1/3
-
Manufacturer says speeds for x,y,z are 1500 and 200. I'm sure that's metric. So the speeds should be 59 and 7.5 in standard. That's quite different from what I had. I will try this in the morning and see if it fixes things. A is correct, I know that as it's at the manufacturer recommended settings and in degrees. Say a prayer for me! lol
-
Wow...now, 99% sure this looks like a victory dance! I have x,y,z at 58 and 7.5 in motor control. Lap 2 is cutting and appears to be dead on target and all pocket jumps have 100% cleared the work piece! I suppose the summary is that the motor tuning was way too high causing missed steps.
-
Good Job(;-)
(;-) TP