Machsupport Forum
Mach Discussion => General Mach Discussion => Topic started by: kcib on April 28, 2012, 03:17:17 PM
-
I have a problem with the "Z" axis on my router , with the help of Tweakie I have made sure that all my cables are shielded and re-routed them I have also fitted capacitors on all my inputs , when I run the driver test I get a perfect straight line without any spikes , I have changed stepper motors over stepper drives over without any success , what happens is that whilst doing a roughing cut the "Z" axis starts to move very slightly up from the level it should be at then goes back down now this is not a real problem when roughing out but it does become a problem when doing the finish cut as it leaves lines on the work piece which after spending 3 hours or so on a 3D carving is a pain that I could do without.
Before anyone one asks my computer runs Windows 7, Quad 3.4ghz processor , 12gb memory and a 64gb solid state hard drive , apart from the operating system the computer is loaded with Mac3 and the Mach Standardmill screenset nothing else at all and is not connected to the internet , all drawing work etc. is carried out on my main computer and the via a USB key loaded onto machine computer, the software that I use for my carvings etc. is Vectric Aspire and Cut3D.
Any ideas would be greatly appreciated , and my thanks to Tweakie for all his help.
Attached is a pic of the sort of problem on a work piece.
-
Loose coupler on the Z screw.
Or, losing steps from too high accel. Lower accel by 50% and retest.
If it still does it, my vote is a mechanical issue.
-
Coupler is 100% tight and all the "Z" axis all brand new , new ballscrew and nut plus all the other parts required and new Hwin linear rails and bearing blocks , there is for sure no loose fittings backlash etc.As for acceleration I have been up and down the scale with no effect , so if it is mechanical as you suggest where would you be looking ??
Thanks
-
I've heard of couplers appearing tight, but are able to occasionally slip back and forth a little. Especially if they use setscrews.
Also, try going into the bios of the PC and disabling C1E and/or EIST if you have them.
-
Alan,
I would suggest some simple pocketing ie; a few passes in the kind of wood / against the grain and at the cut depths you use for roughing.
If you do a straight cut over say 10" and the Z rises, then you would know that it's a mechanical problem because the code is not changing for the Z. Acceleration would not be factor.
You may want to just plunge directly in as that would put an upward force on the Z and over some travel it should come down as it's cutting the length.
FWIW,
RICH
-
Hi Rich
I have already tried this and am certain that this is not a mechanical issue as the Z axis only moves up never down , only by a small amount but always up , if I watch the DRO for Z then where as it should be cutting at depth of say 5mm then it will alternate from the 5mm that it is supposed to cut to 4.9650 and then back to 5mm , as I say not a great amount , the other thing is that if I remove the stepper motor from the "Z" axis and just hold it in my hand you can see and feel the motor run down to where it would be machining at the required depth and then you can see and feel the motor move a small amount so I am convinced that this is an electrical problem and not mechanical
Thanks
Alan
-
What drives? What BOB?
Do the DROs in Mach show the movement or is it just the motors move and DROs read correctly?
Hood
-
Hi Hood
the DRO's in Mach show the movement , the stepper drives are from www.lamonde.co.uk and are STP-DRV-4850 and can be found here (http://www.lamonde.co.uk/acatalog/Drives_PSU.html) and the stepper motors are from same supplier and are STP-MTR-34066 fitted to the "X" and "Y" axis and STP-MTR-23079 fitted to the "Z" axis and can be found here (http://www.lamonde.co.uk/acatalog/Motors_Cables.html) BOB is a CNC4PC part number C35 and can be found here (http://www.cnc4pc.com/Store/osc/product_info.php?cPath=33&products_id=270).
Thanks
Alan
-
If the DROs show the movement then Mach is commanding it and it is definitely NOT noise. Why Mach is commanding it is what you need to find out, is it definitely not in the code?
Hood
-
the DRO's in Mach show the movement
For that to happen, that would mean the anomaly is in the g-code, and I doubt that is the case. If it were, you'd see it in the preview in Aspire. I've never heard of this occurring in either Aspire or Cut3D.
Also, that looks like much more than .035mm offset.
-
What is the resolution of the Z axis ?
A very coarse setting could possibly produce the effect you are seeing due to rounding.
Possible ? ?
Russ
-
Assuming also that it's a clean licensed install from Vectric.
-
Agree Gerry that those lines look a lot deeper than 0.035mm ( 0.0014") and for the deeper cuts you would see it when
the code is simulated.
What microstepping are you using for the Z axis?
If greater than 10 us your wasting your time. Your 276 in oz stepper probably can't hold that positioning. Note that when you microstsep your torque is is greatly reduced, as much as 50% or greater. As you make all the numerous small moves you can be loosing or gaining steps / skipping in the Z. May not even hear the skipping and it adds up and may not be repeatable for the same parts of code.
The above shows it's ugly head when doing xyz moves. You could do regular machining all day and not see the effect.
Changing acceleration, velocity, depth of cut etc doesn't help. That's my experience here if it's motor related. Again very hard to pinpoint since you think it's a mechnical problem.
BUT
Before you focus in on the above make darn sure there is nothing mechanical that would create what your seeing and definitely make sure it is not the code.
RICH
-
Hi Rich
I have already tried this and am certain that this is not a mechanical issue as the Z axis only moves up never down , only by a small amount but always up , if I watch the DRO for Z then where as it should be cutting at depth of say 5mm then it will alternate from the 5mm that it is supposed to cut to 4.9650 and then back to 5mm , as I say not a great amount , the other thing is that if I remove the stepper motor from the "Z" axis and just hold it in my hand you can see and feel the motor run down to where it would be machining at the required depth and then you can see and feel the motor move a small amount so I am convinced that this is an electrical problem and not mechanical
Thanks
Alan
Hi Alan,
If your Mach3 DRO's are showing the incorrect Z axis movement then the problem is either in you GCode or within Mach3. I trust you are using the fully licensed versions of both Mach3 and Vectric Aspire ??
Tweakie.
-
Tweakie,
Not sure what he is seeing in the DRO ...ie; is the line that is running the same as the what the DRO is indicating...
One can single step through the code or simply put a dwell ( G4 ) in the code and compare
RICH
-
Changing acceleration, velocity, depth of cut etc doesn't help. That's my experience here if it's motor related.
From my experience, if the Z is losing position, then reducing accel will definitely help.
However.
The fact that it appears to lose position, and then get back to position, really points to a mechanical issue to me.
If the Z were to lose steps, it's very unlikely that is will lose in one place, and then regain the exact same amount in another place. Typically, it'll do one of two things. Wither get deeper and deeper. Or slowly lift above the part and stop cutting.
What are the actual accel and velocity settings you're using?
-
Alan,
Can you post a pic of your Z axis and machine?
Hey Jerry, what you say, and I do agree , that reducing accel should help.
What I said in reply #12 needs additional or better explaination, but, for now don't want to go there.
Like yourself, desire to eliminate the mechanical and code end of things first.
RICH
-
Hi all
first of all all my software is fully licensed and genuine , secondly I have never said that it was a mechanical issue it is 100% electrical , to start with if it were mechanical you would no be able to remove the "Z" axis stepper motor from machine and not only feel the stepper move a fraction but see it move. As for it being in the Gcode someone else run this program for me using exactly the same steppers and drives fitted to his machine and it ran perfectly so it has to be something related to my machine , would it be anything to do with the PCI parallel port cards I am using ?? I have two in this computer so I might just swap them over to see if it does make any difference. Also the picture I attached in an earlier mail was the wrong one so attached is the correct one both machine one after the other with no changes made to machine etc. the part on the left you can see the lines on the left hand side and the part on the right has the lines on the right as I say no changes were made as soon as one part was finished second one was done.Also Pic of "Z" axis
Thanks
Alan
-
Need help in loading up the pics to go with my last mail I keep getting this message (Sorry! There is already an attachment with the same filename as the one you tried to upload. Please rename the file and try again.) can anyone help me out ??
Thanks
Alan
-
Hi all
thanks for all your help but have changed over the two parallel port cards and for some reason all is cured not sure why tat hould be but the most important thing is that the problem has disappeared ;D
Thanks
Alan
-
Hi Alan,
My apologies, I did not mean to suggest that your software was not genuine I just wanted to clarify that you were not using trial or demo versions.
Tweakie
-
Good that you found the culprit Alan.........now I can go back to a good snooze. :D
RICH
-
Just found this thread. Glad you got it solved with the new lpt port.
I didnt see anything that listed any mach 3 settings. I am willing to bed that your original problem was the step and direction pulse width. I have seen this before a couple times. Mainly when running very large Gcodes with tons of Z axis moves.
Upping the step and direction pulse widths cured the problems.
-
Hi Fixittt
thanks for your response , as a matter of curiosity could you tell me what settings you use on your "Z" axis ?? velocity and acceleration as well as pulse widths ?
Many thanks
Alan
-
If it shows up again (now that I see it was fixed) I had a problem way back in releases that had to do with mach3 not handling a consecutive sequence of very small moves well. The cam software I was using for 3D surfaces was set to allow scallops of nothing higher than .0001". So there were a lot of very small moves that were stacked-up in the gcode and they were going plus/minus on all three axis. I can't remember what fixed it - maybe something to do with look ahead? I might have had backlash enabled to correct for .002" or something and maybe turned it off? It could have also been a release upgrade that fixed it. You might want to try a test where you make steps or blocks that are only a few mil high and in straight lines.
Jerry