Machsupport Forum
Mach Discussion => Mach4 General Discussion => Topic started by: rodm717 on June 03, 2018, 11:53:40 AM
-
As title indicates, i am having a repeating issue with motor control on y axis of my milling machine. Both x and y are using same motors and settings. I have never had any stalling issue on X, but for some reason, y stalls sporadically, at the worst of times. I have replaced the motor 2 times now and cant figure out what is causing the issue. I have troubled shot the machine itself and am not finding any binding or others issues that would cause this. I've tried changing velocity and acceleration. Any thoughts. Could it still be a motor issue? The motors i am using are used and i have a bunch of them. I find it odd only y axis has issue. It doe not happen in jogging.
It does only seem to happen during rapids. Can i limit or set rapid speed within Mach4?
-
I went back through cam and did find 1 issue with 2d adaptive feedrate, just to take that out of the mix, however, still trying to find issue with rapid. I've ran through jogging at 100% and have not been able to replicate. Still haven't found a way to set rapid to say 75% and see if that will fix it. I really dont see a problem decreasing rapids as a work around, for my use.
-
Your velocity setting is the rapid speed.
-
Thanks, Ger21. I have been trying to get this machine finalized. It's the things that are most simple, that can sometimes have you stuck. Lol. I will try lowering velocity. Its just odd in most any other scenario, it seems fine. Fusion360 and trying to get accustomed to how it outputs code with Mach4 post processor. Some of the initial code seems overly complex for what it needs to do.
Any idea if Mach3 post running on Mach4 would be better? Or a bit more simple?
Lowering velocity down to 45 and still having issues. Looks like a motor swap is a must. Hopefully not a defective driver.
-
It's highly unlikely that the motor is bad.
Accel usually has more of an affect than velocity. Try lowering the accel.
What kind of drives are you using, and what machine is it?
I'm not familiar with the Fusion 360 Mach3 post, but most of the Fusion posts I've seen have a lot of things in them that I'd remove.
-
Hi,
you say the X and Y axis motors are the same, but are the axes?. Is the Y axis heavier for instance?.
A heavier axis will require more torque to accelerate it. If you can eliminate other potential causes like power supplies, driver settings,
control settings the you may need to increase the size of the motor.
What motors are you using?
What voltage power supply are you using?
What drivers are you using?
What microstepping regime are you using?
Craig
-
Be sure you have smoothing checked on your adaptive toolpaths.
-
Hello was there a time when it worked perfect and for how long. or have you had this problem from the begining. Then i would go after my machanicals check oilers especailly they can screw up allot. Can you swap drives X and Y. Might also be good to make sure the power going to each drive is up to spec i have had contactors go bad and deliver bad power when under load. More the exception but i have seen it.
Good luck
mark
-
Thanks for the interest and info, and questions. Problem did turn out to be motor.
I am using a Pokeys57cnc controller, Pokeys 60-256 drivers, 2510 ballscrews, and a7.2a 48v power supply on a round column conversion mill. All but Z is on a 2-1 ratio, while z inherently with worm gear is 10-1 use same pulley. I was initially using 3nm nema 34 3a EAD motors i had bought used, but seemed to run well in all previous testing. I swapped out the Y axis motor for a 4nm 80mm 4a Sumtor motor that was brand new and had the light go off in my head. 110 velocity is so quiet and smooth in comparison. I will be switching out the other 2 axis with 3nm 66mm 4a Sumtor motors that i already had on hand. Luckily, i had bought these over a year ago when the price was exceptional, literally less than half, compared to current pricing on these motors.
While the electronics and configuraion are usually the main suspect, used motors i have learned can be as well.
-
Unfortunately i had to order an additional 22t pulley and for time being placed a 12t on Z and the old EAD motor, while I bored out the 22t to fit the new motors 14mm shaft for X. Doing this raised the resolution even higher on Z. Thinking that limiting micro stepping to 1/2 while still maintaining steps to under 8000 for mach 4 to be a good thing. However, i immediately realized velocity and acceleration issues once again. So, i have 10 to 1 on the worm gearing to Z, a 12t pulley on the motor and a 22t pulley on the worm.
I changed to 1/4 stepping and roughly 14500 steps in mach 4 and immediately resolved the velocity and acceleration issues, while the motor also smoothed out quite a bit. What am i not understanding here?? By increasing micro-stepping and step count, am i not decreasing torque and motor stability??? I had always believed, from various posts, that keeping mach step count under 8000 to be recommended.
-
Hi,
microstepping very significantly reduces torque between adjacent microsteps but changes not at all
the torque between full (or half) steps. Maybe this will help:
http://www.machsupport.com/forum/index.php/topic,37413.msg255913.html#msg255913 (http://www.machsupport.com/forum/index.php/topic,37413.msg255913.html#msg255913)
Craig
-
Hi,
microstepping very significantly reduces torque between adjacent microsteps but changes not at all
the torque between full (or half) steps. Maybe this will help:
http://www.machsupport.com/forum/index.php/topic,37413.msg255913.html#msg255913 (http://www.machsupport.com/forum/index.php/topic,37413.msg255913.html#msg255913)
Craig
Exactly what i thought in terms of torque, but again, on a used motor maybe not a good indication. Received the pulley i was waiting on, so out goes that motor.
Is there a formula to figure effective torque? Utilizing or including the gear ratio, screw, and given torque of motor?
-
Hi,
there are some predictive formulas but they require an accurate assessment of motor variables beyond those the are made available to us by the manufacturer.
The best figure of merit to determine whether a stepper will perform well at speed is its winding inductance. The higher the inductance the longer it takes for the current to
increase and as current and torque are closely related, the greater the drop off of torque with speed.
You might check out the Leadshine website, they don't as a rule publish an inductance figure but they do publish Speed-Torque diagrams which give you an excellent means of
comparison between products and for expected performance in practice. It is interesting to note that some of what appear to be the best and strongest performing steppers behave
very poorly at high speed, maybe as little as 5-10% of rated torque at 1000 rpm whereas a smaller and lesser unit will retain 25-30% at 1000 rpm and be better than its supposedly
bigger and better sibling.
In the attached pic you can see that the applied voltage is a major determinant in torque at speed, the higher the applied voltage the better the stepper will go. Consider then Geckos's drivers
which are 72V capable or Leadshines AM882's which are 80V capable.
I use Vexta 5 phase steppers for my mill and chose the genuine Vexta drivers as well. The drivers are hooked directly to the 230VAC line. Their peak output to the stepper is about 150V and
man do those steppers sing! I paid quite a premium for them though.
The bottom line is that stepper loses torque bigtime at speed, you can expect even good and well designed units with high voltage drivers to get no better than 30% rated torque at 1000 rpm.
If your machine is such that you need greater rotational speeds then you want proper servos.
Don't be taken in by the advertising hype about closed loop steppers, that is to say steppers with encoders and smart drives, they would have you believe that they are as good as servos, not even
close. They will also tell you that they don't lose steps, also fallacious. They are still steppers and a closed loop stepper loses steps (and faults 'following error') in exactly the same circumstances that
open loop stepper loses steps. Provided a stepper is used WITHIN its performance envelope it DOES NOT lose steps. If its used outside its envelope it will, closed loop or not.
Genuine servos have rated torque right from zero speed to rated speed. Some manufacturers publish the peak torque only, it looks really impressive but what you should base your judgment on is
'continuous stall torque'. When you do so you might be surprised that a servo has less torque than a stepper of the same physical size, and of course may well be two or three times more expensive.
IT IS NOT A TYPO! Steppers perform very well at low speed, better than servos even, but servos come into their own above 500-1000 rpm and will carry on giving rated torque where a stepper has long since
stalled.
AC Servos and servo drives are decreasing in price but are still a considerable premium over steppers. I think it may be worthwhile that you experiment some more with steppers until you are familiar
with their strengths and their limitations. Do yourself a favour and get some decent high voltage drivers and matching power supply, you'll never get the best from any stepper until you do. I have been
told that steppers magnets can weaken with age and so you might need to purchase new to ensure that you get rated torque.
The link I gave you previously tells you that microstepping is about making steppers run smoothly not increased resolution, as seductive and logical as it sounds. A microstepping regime of 8 to 16
microstepps per full step is about right, more is giving you only marginally better smoothing but upping your signaling rate maybe beyond your controller/BoBs capability.
Craig
-
Hi,
sorry forgot to attach the pic after all of that!
Craig