Machsupport Forum
Mach Discussion => General Mach Discussion => Topic started by: sjbjcl on April 23, 2006, 11:50:53 PM
-
I am using the Low Cost CNC 4 axis stepper system from the guys in Australia to run a CNC router and I am having trouble sorting out an issue with my motor calibration in Mach 2. Or at least I think it's a calibration issue. It seams that my stepper motors sometime (unpredictably) lock up when a rapid movement command is initiated. It will run fine for long periods but then all of a sudden a motor freezes and makes awful sounds. I have had it happen individually on all three axis on separate occasions. Can anyone provide suggestions or offer advice on how to trouble shoot this?
Steve
-
Could you please run a driver test and tell me what you see?
-
Hi Champion,
Thanks for your reply. Here is what I found during the OCX Driver Test
Interupt Used = 0
Pulse per Second = varying between 24394 to 24389
CPU Speed = 1694
time between pulses = mean is at about 40 but fluctuates between 20 and 60.
Is this what you were looking for?
Steve
-
sorry I meant to adress my last post to Barker806
-
That looks good...
-
Barker806,
When you say it looks good do you mean that the CPU and Mach 2 are working properly? Are there any other ways for me to trouble shoot this problem?
Do you think it might be an issue with the motor tuning settings? The setup I used for each axis is as follows:
Velocity 0.5
Units / sec^2 = 4
Steps per unit = 20000
Kernel speed set at 35000 Hz
Theorectically, with the system flat out the maximum Velocity should be 1.75 inches/second or 105 inches per minute. To be conservative I limit the rapid movements to 30 IPM. Do you have any thoughts on this?
Steve
-
OKay try running the system at 25KHz and set you step to 5us under the motor tuning. If you have trouble with this please post you your XML and I can do it for you.
-
It will run fine for long periods but then all of a sudden a motor freezes and makes awful sounds. I have had it happen individually on all three axis on separate occasions. Can anyone provide suggestions or offer advice on how to trouble shoot this?
Steve
That is a classic step motor Stall. You are probably trying to run them faster than they are able to.
Tell us about your setup. Are the motors on the machine and driving a load? Steppers on the bench unloaded will nearly always stall at a lower rpm than when under load.
Also steppers prefer to have some damping between them and the load. Never use rigid couplings. I tend to use fenner couplings with a nitrile insert acting as a damper.
Wayne....
-
Bryan, I will give your suggestion a try in the morning. Can you explain how kernel speed cause a motor to stall? I appreciate all your input.
Wayne, The motors are attached to the CNC router and dirve ballscrews using timing belts with pulleys and a 2:1 ratio.
Steve
-
Wayne,
I attached the specs on the stepper motors. I have size 34 P/N PF4913A-05 on all three axis.
Steve
-
taking it down to 25KHz will take a little stress of the computer... If you don't need it up that high then why not :) Please be sure to set the step to 5us long.
Thanks
Brian
Bryan, I will give your suggestion a try in the morning. Can you explain how kernel speed cause a motor to stall? I appreciate all your input.
Steve
-
I attached the specs on the stepper motors. I have size 34 P/N PF4913A-05 on all three axis.
Steve...
Looking at that there are two 1280 oz/in motors. Looks like you have been wise and bought the lower inductance one. Although they dont mention it at all. But my guess is the lower current version 34 P/N PF4913A-11 is the higher inductance versoin.
It may be that you are just not driving them with enough voltage. I know nothing of the drives you are using..
I personally wouldnt of had the 2:1 reduction with steppers. It generally makes matters worse, and they need to spin faster with it.
I'm currently retrofitting a small vertical machining centre and have 920oz/in steppers driving 5mm pitch ballscrews 1:1 through compliant couplings. I get over 5000mm/min rapids with no stalling. I'm using 72v supply through Parker drives. Motors rated at 4.2A.
The same motors struggled to spin at an equivilent of 3000mm/min on the bench.
Anyway you can try it without the reduction?
I really dont think this is a signal jitter problem..
Wayne...
-
Wayne, I believe that the package I bought has Prodrive 2000 drives but they don't advertize or publish this in their documentation. They stated that the power supply is a 48V DC 13.8 amp unit. They recommended using a 2:1 ratio based on their past experience with this combination on larger routers. My router has a fairly large working area of 48" x 96" x 24". To remove the gear redcution I would need to buy some more parts. Your speeds are much higher than anything I have ever seen with my machine. That might be because of the gear ratio setup requiring the motors to spin faster to turn the screw. Does this new info give you any other ideas?
Brian, It might be hard for me to reproduce the problem since it only occurs once in a while. I will give your suggestions a try and get back to you on what I find.
I really appreciate all the feed back. This forum has really help me achieve things that I could never of imagined or figured out on my own. I am in debted to you guys.
Steve
-
I believe that the package I bought has Prodrive 2000 drives but they don't advertize or publish this in their documentation. They stated that the power supply is a 48V DC 13.8 amp unit. They recommended using a 2:1 ratio based on their past experience with this combination on larger routers. My router has a fairly large working area of 48" x 96" x 24". To remove the gear redcution I would need to buy some more parts. Your speeds are much higher than anything I have ever seen with my machine. That might be because of the gear ratio setup requiring the motors to spin faster to turn the screw.
Steve. Well if it does have Prodrives in there, they are ripoff's of the Gecko G201. The Gecko's at least are rated to 80v.
Re: reductions. It's all about tradeoff's, with 2:1 reduction you probably more than half the torque at your target rapid speeds, as the stepper looses torque most at higher speeds. I dont quite get their logic! If it's a wood router then it needs much less torque than my machine cutting metals. You will be wanting to cut at higher speeds than me. If you were using servo's I'd understand it.
My vmc yesterday was cutting Al-alloy during testing at 1000mm/min (40"/min) 3/4" endmill 40% stepover and 3mm (1/8") deep cuts, no problem with it at all. and thats direct drive steppers with lower holding torque than yours. No idea of their torque when actually moving though. These are chinese square body motors.
Steppers can be a pain to get right, but very possible! Doesnt help that there's so much duff information out there!
-
Steve....
Just had an idea for something you can try..
Take the motors off the machine, and run them one at a time on the bench.
Increase your max velocity in mach until they will stall on the bench, the motor will just stop turning an squeal.
They probably wont all behave the same, but that soesnt matter.
Now lower velocity until they no longer stall on long rapids.
Now the figure you get will be LOWER than what you SHOULD be able to achieve loaded by the machine.
In my case I was able to increase max velocity by over 50% when loaded by the machines mass and friction..
Let us know what you find..
Wayne....
-
Wayne,
Once I establisht the maximum speed on the bench are you suggesting that I increase the velocity after I put them back on?
Steve
-
Steve..
You should be able to! They should go faster than the speed at which they stalled on the bench.
If they dont, then I'd expect that they are not seeing enough load. Or if they stall at a lower speed you got some serious friction, but I very much doubt that will happen.
I still reckon you might be better off at 1:1. You have some big motors there.
What are you intending to cut?
Wayne...
-
Wayne,
I just order the new pulleys and belts to give the 1:1 ratio a try. I am cutting hard maple blocks, thin gage aluminum sheet, and fiberglass. We make aircraft composite and sheet metal parts for general aviation aircraft.
Steve
-
Hi Guys,
I ran a binch of bench level tests with the motors removed from the machine and here is what I found:
Started with a basic motor tuning / mach 2 setup of:
Units / sec^2 = 5
Steps per unit = 20000
Kernel speed set at 25000 Hz
Checked effect of Velocity Variations:
Max = immediate stall
1.0 = immediate stall
0.95 = runs a little then stalls
0.90 = runs with missed steps every 2 or 3 seconds (sputters and jerks motor)
0.85 = runs with less missed steps (counted 7 or 8 sputters and less of a jerk)
0.80 = runs a little better than 0.85 but still misses steps (counted 6 sputters)
0.75 = even better but still sputters (jerking is gone)
0.70 = even better but still misses once in a while
0.65 = only misses a couple times on a long run
0.60 = runs without missing steps but still a little loud
0.55 = runs smoother and less noisey
0.5 = runs quietly and smooth
The sputter or missed step is basically a change in the noise emmited from the motor and sudden vibration. At higher velocities it is visible but at lower velocities it is not visible (can only feel and hear it). The best way I can describe it is like its about to stall but keeps on going. My original setup used a velocity of 0.5 when I had a motor stall for five to ten seconds (still mounted to the machine).
Does this provide you any insight that could lead to a solution or do you have any other suggestions?
Steve
-
As a test try Sherline mode and tell us if that makes it better.
-
Hi Brian,
I am currently using Mach 2 but I guess I could install Mach 3 and try it in Sherline Mode. Going over your past reponses I noticed something that I might have misunderstood. When you said to set the step to 5uS long did you mean the minimum pulse width? RIght now that is set to 1 and the direction prechange is set to 0.
Steve
-
Yes the min pulse width, set it to 5 and see how that runs.. you can insatll Mach3 nd 2 on the same computer too :)
-
Great I will give that a try in the morning. Is there a way to setup mach 3 using the settings for mach 2 without going through the whole process of sorting out the setup again?
Steve
-
Nope you are going to get to set everything up again :(
-
All my active low pins are checked off for my X/Y/Z axis in my output pins section. How do I know if this is correct?
-
It should be okay...
-
That's cool. I will work with this for now and see if either the 5uS width or if Sherline mode works better. Cheers!
-
Great, i hope this gets you fixed up!
-
Hi Brian / Wayne,
I spent most of the afternoon running tests on my router. I changed the minimum step width to 5uS and reduced the kernal speed to 25000 Hz and reduced the aceeleration to 1.1 U/S^2. I repeated the bench tests starting at max velocity and kept slowing it down until the motor sounded smooth. I came to the conclusion that a velocity setting of 0.6 in/sec felt right. The motors run fairly well (sputtering occaisionally) at speeds up to 0.75 in/sec but on long runs with all the motors running at 50in/sec they almost always stall after about 5 to 10 seconds. I also noted that my problem of sputtering starts to show up at all speeds above 0.6 in/sec.
The 0.6 in/sec setting is a little better than what I had with my original settings. So I hooked up the motors to the router and ran them through a series of long run cycles both individually and simultaneously. The good news is that with this setup I did not experience a stall. To see if I could improve the speed on the machine, I tried increasing the velocity. Again at anything over 0.6 in/sec the motors start to sputter and stall at 50 in/sec just like on the bench tests.
My feeling is that the issue will likely repeat itself because I really did not notice a change. Although after about 1 hour of cycls with the 0.6 in/sec settings I realize that my original problem of stalling once or twice on a long cycle (4 to 9 hours) may still occur.
I think that my drives are either bad or not capable of driving these motors properly. I know that they should be able to cope with mid band resonance stabilize the motor but from what am seeing it sure looks like it is not doing ti very well.
Any ideas?
Steve
-
have you looked inside this driver box you have to see which drives area in there?
-
PROTOBYTE PRODRIVE 2000's
-
PROTOBYTE PRODRIVE 2000's
OK do they have a pot for tuning the low speed like the gecko's have? if so have they been tuned?
What is the supply voltage to these drives?
Cant remember if you mentioned your ballscrew pitch?
I notice you are still quoting speeds in inch/sec. Do I read into that you are still on Mach2?
Not quite 20 questions ;-)
Wayne...
-
PROTOBYTE PRODRIVE 2000's
Set up Mach 3 and use the sherline mode I think iti will fix it up...
-
Brian,
I ran the machine all day without a hitch. I started setting up mach 3 as you suggested but had to stop to run a couple jobs. Is there a new users manual or update for Mach 3? There are some differences with Mach 2 and I am not sure what they mean or what I should select.
Thanks,
Steve
-
There is a new user manual in the works... John P. is writing it but it is a hard job to keep up with Art! What are you having trouble with?
-
Steve
Whats your current "reliable" feedrate?
-
There are quite a few more settings available in Mach3 that are not the same in Mach2. I was hoping to be able to decipher them as I go through it by reviewing the manual. I guess the changes are not major but I always try to understand what I am doing before I implement a change. I am currently having trouble getting Mach 3 to talk to the controller. I should be able to sort it out but I have been so busy lately that I have not been able to take the time to go through the process of checkin all my settings. I will get just need to spend some more time on it.
My CNC router is running well with the 25000 Hz and 5uS settings. My rapid movements are now set at 35 IPM. I face milled, roughed out, and finished four large 3D Milled molds at cutting speeds of 20 IPM and did not have any stalls. The finishing cycles took between 9 to 14 hours a piece so I know it is wooking better since it did not stall once.
I will let you know the results of the sherline settings as soon as I get Mach 3 up and running. Has Art though of just putting out a quick start guide for Mach 3 that could explain some of the differences with Mach 2?