Machsupport Forum
Mach Discussion => General Mach Discussion => Topic started by: james elder on August 31, 2006, 12:33:47 PM
-
The new forum is much better, the advice is still the same.- excellent
More used to large non critical work, I have been able to live with the error/s shown in the photo, but now I need to sort them. As well as the obvious, I seem to be missing steps, in that after travelling left to right,and back again, for say 100mm, I am 2mm short....and you guys work in thou.. The other axes seem ok.
Where do I start in my search for a cure.. well after reading previous posts, trying a bunch of stuff, and getting confused, that is...
thanks in advance
Jim
-
Are the smaller circles worse than the bigger ones? if so then backlash may be a factor as with the smaller circles you will have the same backlash but with travelling a smaller distance it will show up more I think.
Hood
-
Thanks for reply Hood.
Yes it does show up more on small things, so..it could be backlash, as you say.
I saw something (i am on Mach 2 by the way) a tickbox for backlash on or off, but when checked I started losing steps badly. I always wondered how you could tell by listening if steps were being lost. I guess it's that momentary low groan before the weeeeeeeee
Is it possible to tell from the shape of the circle where the slackness lies? they seem taller on the left than they should be so...the slackness must be in the drive from left to right not kicking in fast enough, because of some play or looseness.
Am I thinking about this right?
thanks again.. it helps loads to have someone else thinking about it.
Jim
-
Easiest way to see what backlash you have is to put mag base with a dial on the table with the stylus on the spindle. Jog one way so the dial reading is near its max travel, zero the glass then go to MDI and do a movement in the opposite direction and see what the difference between the command move and dial is. Repeat for the other axis.
Hood
-
Will do Hood, thanks.
I did read this or a similar method in previous posts. I guess was hoping for an easier answer, but I'm on it now.
Jim
-
Could you please tell me what you are running for drives? Also what is your steps per unit...
-
Mine are just the same as yours, and like you, it hasnt been so criical. It would be nice to see the answer for this problem. I have always written off the reason I had the problem was too few steps per unit. Not enough gear reduction either.
-
Hi:
Those are pretty horrible circles. Sorta chewy parabolas glued together..
First, you need to back up and run some tests.
Zero your machine.
Command G0X10 and measure it. Did it move exactly 10?
Do same for the Y.
Repeat 10 times , is it the same each time.
Now, mark a spot, zero on that spot. Command a file with
G0X10
G0X0
(repeat 10 times)
Did it come back to zero..exactly...
If not, rezero, repeat 10 more times. Is it off the same distance, or different?
We'll go from there.. :)
Regards,
Art
-
Thanks for input guys. I'll get back to you asap with answers.
Jim :)
-
This sure looks like backlash in the X axis ( or what I labeled the X axis.). If you look at the 12 and 6 o'clock positions, where the Y axis changes direction, they are almost perfect. Now compare them to the 3 and 9 o'clock positions where the X changes direction.
If this is a servo machine you could be accumulating error pulses in the X drive before movement. Increasing the I gain should help if this is the case.
Darek
-
Your're right Darek, thanks for the work.
Servo machine? I'm not sure, is this the type of motor Darek, how could I tell?
Jim
-
Yup, that pictures better, does look like backlash...
-
Your're right Darek, thanks for the work.
Servo machine? I'm not sure, is this the type of motor Darek, how could I tell?
Jim
Ther are two types of motors you can have, servos or steppers, servos will have encoders on them and steppers usually dont.
What kind of machine do you have and do you have a link to a web page for it?
Hood
-
It's a home built beauty, not by me, by two others, one who made the structure, and put some controls on, and one who took the wiring out again and reworked it. I'll post some pics tomorrow if I can, and see what you think.
The guy I got it off said this forum was great, and very helpful, and he was right.
Thanks guys..
Jim
-
Are the motors square in section? if so they will be steppers. If they are round then they could still be steppers but may also be servos.
Hood
-
Not always ;) I've got some square servo's!
Steppers will generally have 4, 6 or 8 wires. Servo's 2 for dc and 3 for brushless + the encoder cables.
Wayne....
-
Not always ;) I've got some square servo's!
Might of known you would be different ;)
Hood
-
It's great to have you guys on my case. They're definitely round, but I'll have to count the cables in the morning.
'night all.
Jim
-
I think if you can post a pic of the inside of the control box we can tell you what you are running :) The guys here have installed and run about everything ;)
-
Brian
True.. The drives will tell the whole story. But as others have mentioned this looks like classic mechanical backlash on the x axis screw.
Are they ballscrews or standard acme I wonder.
-
Hello guys.
Some answers here.. 8 wires to the round Servo/stepper, wired in 4 pairs I notice.
Some pics attached
-
Ooop that was to big for my screen
-
You have stepper motors. Can't spot your drives though.
-
Thanks fdos, that discounts the accumulation of error pulses causing it, and leaves me with good old mechanical backlash on the x axis, and it doen't matter what the drives are.
Photo attached of screws at base of z axis
Cheers
-
Looks like you have standard Acme screws. If there are two nuts on each you might be able to take out some of the backlash. There are ways of doing it with one nut too.
Sometimes the drive does matter, some are sensitive to direction change timing.
Where in the uk are you. I assume you are over here!
You could check to see how much backlash you have either with a dial gauge or even a fine pen mounted on you z axis. you appear to have a fair bit of slop there, so that methos would show it.
Wayne...
-
Sunny Buxton in Derbyshire Wayne.
Art and Hood have suggested ways of measuring the backlash, which I have yet to do.
Considering, like you say, the amount of backlash, I expected to be able to feel it in the machine, but rocking the thread from the end does seem to move the zaxis instantly. this leads me to want to investigate the stepper/thread interface. I seem to remembr reading years ago that things like a short length of tubing can be used to form a crude clutch in case of a crash, and I am wondering if there is something like this that has malfunctioned. Either this or there is some direction change timing problem like you say. Also there is the problem of apparently losing steps quite badly,...I'm starting to think mechanical rathr than electronic.
thanks for your help.
Where are you Wayne?
-
James,
I'm down on the Notts Derby border M1 J25.
Do you want me to pop up and take a look one night?
John S.
-
I far enough away in Poole dorset ;)
Also check the bearings at the end on the screw, backlash can be created there. Normally these bearings are back to back angular contact bearings with some preload of them.
Wayne..
-
John S Thanks, thats a fantastic offer.
I'll have a go myself before I ask you to put on your snowsuit and head up into the hills. -The suggestions from the guys on the forum have given me loads of leads, so I'll follow them up as far as I can.
thanks again.
Jim
-
John I need my lathe finished if you are looking for work ;)
If you guys can't find the problem try sherline mode. this will give a very long step pulse and could help the drive (Whatever they are)
-
Send air fare....................
John S.
-
I'm on Mach2 Brian, so no Sherline mode..
Jim
-
okay ...put in a pulse width of 15 on the motor tuning page and see if that helps
-
Will do, Brian thanks.
I think I have tracked the problem down. Turning the handwheel on the xaxis produces instant movement of the Zaxis. Turning it back again gives instant movement the other way, as it should.
But...there is a small arc of this movement where I cannot feel the smooth gentle ratchetting of the stepper being turned.
Put the other way round, theres a small arc of stepper motor turning where there is no Xaxis movement.
A slack motor/screw coupling?
I'll get me toolbox out and have a look.
Jim
-
Hmmm Loose coupling will do it!
Personally I always use Oldham couplings, with the nitrile insert. Not the harder nylon one or whatever it is.
Wayne...
-
Thanks Wayne.
I had a look at them in RS and they look really good. I'm not sure what is on it at the minute, because it is all boxed in, with only 2 holes, (I bet for an allen key) which is partly why I didn't spot it earlier.
Cheers.
Jim
-
So did you get it sorted Jim ?
John S.
-
It feels like I have, - the extra movement has gone, but I just havn't had time to spark it up and test it yet. - I'm out of the workshop all week. I can't wait to try it, and I will post the results,...watch this space.
Thanks for your continued interest
Jim
-
Here is the photo of the new look circles.
Thanks Art, Hood, Brian, Ynneb, Hillbilly, Fdos, John S.
It took me a while because as soon as the circles were sweet, another problem became more obvious, and I've been eliminating various causes. I'll upload some photos for your analysys soon.
Thanks again you guys.
Jim
-
That's no improvement :-\ I'll try this...
-
Hang on, ..this-