Machsupport Forum
Mach Discussion => General Mach Discussion => Topic started by: Kevin Steele on May 09, 2009, 12:52:32 AM
-
I'm experiencing an accuracy problem, probably due to losing steps. The machine is a Denford Triac. I have replaced the original stepper drives and power supply with the 3A ones here http://www.arceurotrade.co.uk/Catalogue/Stepper-Motors/Stepper-Drivers and built the power supply to the plans on this site as well. There is a cnc4pc C11 board in the base of the machine where the original controller used to be. The original stepper motors have been left in place. Everything seems fine, but I'm getting a positional accuracy problem. If I run a program by the time it finishes the datum has moved. Not by much, but it is obviously a problem. I think it's missed steps as if I change the stepper drivers to half step or full step instead of microstepping the error increases. I ran a small program to cut a square pocked leaving a circle in the middle. It cut on two levels (1mm and 2mm deep) leaving a larger circle on the lower level. As far as I could measure the circles were round, but they were about 0.05mm out of position to each other. The larger the program the larger the error becomes. I have the step pulse set to 5. I tried the "sherline" setting, but this caused huge errors so was obviously not a good idea. Increasing/decreasing speed doesn't seem to have any effect (I've never run the motor speed over 1000 mmpm). I've tried reducing the speed and acceleration, but without success. It was suggested that the PSU and stepper drivers were probably a bit on the small side for the motors (PSU running about 35V), but before spending a lot of money replacing everything I'd like to be sure. Given the sizes of error it's obviously only missing a few odd steps -but I can't work out where and why, any suggestions or thoughts?
Regards
Kevin
-
Hi Kevin
This could be causing your Problem.
Some Controls Board's are Sensitive to the "Low Active" setting's in Port's and Pin's "Motor Outputs"., If you've used the Low Active's to change your motor direction's.
If the controls board's are all the same, "Low Active" setting's should All be Set the Same "Checked" or "X".
Use "Reversed" in Home and Limit's to change an axis direction If Needed.
Hope this Helps, Chip
-
Chip could well be on to something there but if it doesnt help try doing repeat moves one axis at a time and seeing if it ends up at the same point after a few back and forth moves. This will let you see if its just one axis or more than one. If its one then you could possibly swap hardware between axis to see if the problem stays or follows.
Hood
-
I thought you were on to something there, the low active for direction was ticked and for steps was not -so I'd obviously used that to sort out directions. I ticked both, but then nothing worked so I unticked them both and used the "reverse" to sort out directions. Unfortunately running a test program I'm still experiencing the error. It is in all axis, just depends which one has worked most during the program as to which has the largest error. max speed is set to 1000 and acceleration to 50 -do these sound reasonable figures (mm per minute). I have the step pulse set to 5, but what about the direction pulse (it's at the default of 0 now) -I couldn't find any mention of this parameter in the set up guides. I'm wondering if I could be losing one step each time an axis changes direction.
-
That is the direction prechange time and it may well be affecting your machine, try it at 2and see if that helps.
I would probably be inclined to lower the Step pulse width as well to 2 as if sherline mode stopped it working all together then that would suggest the longer pulse is doing more harm than good.
Hood
-
Hi, Kevin
I'd set them to the same value, It's the time Mach gives the control to see the Pulses.
Some controls have slower response Time's.
Chip
-
Still not cured, but I think we're in the right area. Running the same test program I got different results. With step at 5 and direction at 0 the shift was plus in X, while with step and direction both at 2 the movement was minus in x. I think it is losing a step at direction changes. Any suggestions for my next move, or is it just tweaking the figures and running test programs until I find settings that work?
Regards
Kevin
-
Hi, Kevin
Is your Pulse Frequency stable and Locked in on the Diagnostics Page. ?
Re-Try 5 then 10 and the Sherline 1/2 step Mode again, Strange Issue your having.
Post your XML file for the Profile your using, You'll find it in C:\Mach3 folder, If it's Mach3Mill.xml copy it and Rename It to KevinMach3Mill.xml
Chip
-
Hi Chip,
File attached (I hope). The Pulse Frequency is 24179, which seems an odd number but it's always been that. I cannot alter this in any way -should I be able to?
Regards
Kevin
-
Hi, Kevin
That's good, If the Pulses are locked in, It should be near the Kernel speed you have set Mach to run at.
Chip
-
Just a few comments Kevin,
1. A paper disc with a single fine reference line attached to the motor shaft can provide for comparing
actual steps to commanded steps. Poor mans mechanical encoder for checking if steps are lost ( your trying to isolate the commanded steps to physical movements delivered ).
2. No backlash implemented in your xml. The closer you try to get to perfect the harder it is to find the
what is causing the out of tolerance. AT .05mm ( .002" ) your in the area that a number of things can be
adding up to the tolerance call it the "system influence" . Lead screw tolerance, backlash, your measurements, etc.
Just thinking out of the box FWIW,
RICH
-
Thanks for the help and encouragement. I'm making some progress, but not there yet. While thinking about the pulse frequency it reminded me that I'd been limited on microstepping the drives by my frequency. So I increased it to 35,000 and increased the microstepping on the drives from 8 to 16, which now means 1600 pulses per mm movement. Step and direction have been moved back to 5 (as the Z axis would not drive correctly with lower settings at 1600 pulses per mm). Although this hasn't cured the issue yet (butI've still got room to play with the step/direction values over 5) it has decreased the problem. I think I'm still losing the same number of steps, but because each one now moves the machine by half the amount due to the higher microstepping the problem is reduced. On a simple 2d program I think that this would be virtually un-noticable now. It's just complex programs with lots of code that are an issue.
Regards
Kevin