PDA

View Full Version : rough traversing and stalling only in main screen



davethefixer
09-05-2013, 08:26 PM
Hi all
i have a odd problem with my latest build.
i am having problems with traversing smoothly.
my system is as follows
X axis -two RM2010 ballscrews 1500mm long, belt driven (one to one ratio 40 tooth HDT pulleys)
from single 8.7nm nema34 stepper, hiwin 20mm profile rails, CW8060 stepper drivers, 50v 600w power supply.
now from the motorcalc spreadsheet my max rapid is 3800mm/min
i have set up all the various parameters pins/ports ect.
now thw crunch.......
in the motor tuning screen steps are set to 40 steps/unit
at 3500mm/min the carrage moves smoothly with little noise PERFECT........... BUT........
as soon as you save and go back to the main screen and try to traverse it stalls out and is very jerky with a lot of noise from the stepper motor.
it does this on all axis traversing from the keyboard or running gcode.
go back into motor tuning and its smooth as smooth again.!!!!!!!!!!!!
iv been through the Mach3 manual searched the forums but have not found anyone else having ths problem.
im hoping someone has had this problem and knows how to solve it.
heres a link to a shakey video i took(mobile phone) problem cnc - YouTube (http://www.youtube.com/watch?v=K-p0KxcqsoE)
heres a few photos (i will get round to a build log soon)
88948895
88968897889888998900890189028903890489058906
Thanks all

Swarfing
10-05-2013, 11:42 PM
Listening to that vid your velocity is too high. Post up your settings from Mach, do you have your feed over run set high? i don't use Mach3 myself

davethefixer
11-05-2013, 08:57 AM
Thanks Swarfing
Sorted now
Having tried everything i could think of including deleting and reinstalling mach3, running the pulse testing again (excelent) i decided to try a different PC and wow what a difference.
beautifully smooth traversing right up 9000mm/min
it must have been a hardware fault or was being interupted by something else.
i had been using that pc for about 9 months on my first attempt at building a cnc (MDF special) so you have to wonder how much it was effected. although it was monumentally slow anyway due to the rigidity problems inherent in such a build. (but it gave me the bug!!!!)

Jonathan
11-05-2013, 10:45 AM
250mm/s^2 is very low for the acceleration. If you leave it at that then the feedrate when cutting parts with lots of small moves will be severly compromised, so I'd try about 1000mm/s^2. Due to using 20mm ballscrews and large pulleys, the inertia of the X-axis is substantially greater than it could have been, which will limit the acceleration you can get, so you may have to reduce the maximum feedrate from 9000mm/min to keep it stable. You could compensate by running that motor on 75V instead of 50V, since at the highest speeds that will get approximately 50% more torque from the motor. Once you've found settings that seem to work, you could try running random G-code from this spreadsheet (http://www.mycncuk.com/forums/gcode/5843-random-g-code-testing-motor-tuning.html) to test it properly.

davethefixer
12-05-2013, 08:34 AM
Yes 250mm/s^2 is low but as i was having so much trouble getting it to move smoothly at all i figured its best to start low and work my way up. it seems to tolerate 1000mm/s^2 quite well as i ran it through some of my 3D
tool paths yesterday without the spindle attached at 5000mm/min and it didnt seem to have any problems (the test will be in a week or so if it can actually cut at these speeds)
the 40 tooth test gear i bought was made of alluminium so the weight was well dowm but when i went to order them i found that they were not available in ally in HTD.
i have cut the boss off the back of them to suit the lenght of the Translock mini locking bushes im using to hold them to the shaft. (iv marked them up to see if they slip but they shouldent as they are rated at 39Nm wich is over 4 times the torque the motor can give out)
The 75v power supply is at the top of my "to buy asap list" and has been for a while (since i ran the numbers through the motorcalc spreadsheet.
Thanks for the link to the random G-code spreadsheet that helped no end to see just where some tight sopts were.

one of these days im going to get out of my shed and into the sunshine!!!