[kong]

Just a thought, since it is only drifting in one direction - are you getting a full 5v from the parallel port, since if it is too much under 5v, it might not get the direction signal to change direction.
You can tell I am clutching at straws now, but might pay to check.

[extreme]

Will check that. Is there a difference if I run dos that win98 has, or instal that freedos thing ? I mean runing it as "reastart in MS-DOS mode" . What are the limitations for pulsewidth ? I read on yahoo groups that people have pulsewidth 30-60. What is better, shorter or longer?

[kong]

No difference in the dos, as long as you restart in dos mode it should be ok. Can't answer other question!

[extreme]

There is something bad happening in my garage I took my home pc (this one) to garage to try out mach2. I like this controller, it goes smoother on the curves. But, like I said, there is something "bermuda" in there. This pc works well here, in my room, but when i took it to the garage it was awfull. When win loads, the pc stops responding for about 15 secs, then goes alive for 2 or seconds, then again stops for 15 and so on ... Once a while I do get a good boot to do some aircuts, but after some time it hangs again. Brought it back, and it runs ok.
Any comments ?

[randyf1965]

Power? grounding?

[ViperTX]

I suspect that you could be plugging it into a GFI outlet....depending on how old the house is....and you probably have some leakage on the power supply...just a guess...

[extreme]

House is 1.5 years old.
we have grounding .

[Rance]

It could be like a dog. Some dogs are 'outside dogs', some are 'inside dogs'. Try bringing your CNC into the house rather than taking the PC outside to the garage. I'll bet your CNC is just an 'inside CNC'.

Rance

[extreme]

, it will be kind'a hard to take it upstairs to my room

[sp00k1e]

its not looking for a wireless network that it cant access in the garage is it?
Paul

[Techteacher]

Just recieved my new Red/Cam CNC, I am using Mach2 as a controller program. I too am have the same problem with drifting of my X Y and Z axis. They always drift in the same direction X and Y negative and Y in the positive direction (Y is reversed in program). I live in northern Ontario and Red/Cam is in Cambridge about 2 hours away. They are sending me a new computer but don't know if this is the problem. Someone mentioned voltage, how do I check if I'm getting 5 volts?

Thanks for you help
Don

[esmiller]

EXTREME ---

Question.. I went to the website that displayed your cuts, and noticed a photo of a DIS-ASSEMBLED stepper motor.. you haven't taken apart the steppers you are using on your machine, have you? If you have, they will have lost a great deal of their torque as a result.. or atleast, people CLAIM that they lose their torque when you dis-assemble them...

Can anybody agree or dis-agree with me on this one? .. heard arguments both ways, but proof SEEMS to indicate that dis-assembly reduces torque greatly..

ANYWAYS.. reason I ask is that loss of torque results in loss of steps..

As for the incorrect setup.. I think the person referred to the distance per step.. i.e..

5/16-18 threaded rod with a motor running full step at 200 steps per revolution, the distance per step would be:
1/(18*200)=
1/3600=
.00027777777

Hmmmmm.. I sure hope i figured that right.. would be embarassing if i messed it up ...

Eddie

[esmiller]

OH YEAH.. one more thing.. have you checked for BACKLASH? That may account for small discrepanancies, and if i understand correctly, it would be cumulative, the longer the program was....

Eddie

[Khalid]

Is the backlash commulative?..I don't agreed.. I am having same drift problem along positive X... I have compensateed the backlash in Mach3...

http://www.cnczone.com/forums/showth...142#post798142
In multipass DOC, i get Step-ladder effect along +X-axis always.. I don't think the backlash is commulative....

This may be electrical interference and un proper grounding of control panel...
Any thought

[coldpenguin]

In multipass DOC, i get Step-ladder effect along +X-axis always.. I don't think the backlash is commulative....

I wouldn't have thought so [mechanically], however don't some programs try to backlash compensate, by turning an extra X percent after changing direction. If this is happening, and you have unequal backlash compensation, that could be a cause.

[Khalid]

I wouldn't have thought so [mechanically], however don't some programs try to backlash compensate, by turning an extra X percent after changing direction. If this is happening, and you have unequal backlash compensation, that could be a cause.

Hmmm... You are right.... My step ladders have equal steps... and i will check the backlash again.. My Question is: If the backlash is 0.2mm and if i wrongly set it 0.1mm in the Mach3, should i get the step ladder effect???... I think No....

[cyclestart]

and you have unequal backlash compensation, that could be a cause.

Unequal backlash comp ? Steppers controls have no idea what the real backlash is and will apply compensation equally at every change of direction. No accumulation.
2-2+2-2+2-2 = 0

Just the same as mechanical backlash will not accumulate because the control cares only for rotation of the stepper motor.

At least that's my theory. Shoot holes in it and maybe I learn something new

btw, why are 5 year old threads being dragged from the crypt ?

Edit/ Random thought while weeding the garden: What happens if the changes of direction are .001" and the backlash comp is .002" ? And then there's the internal inaccuracies of the steppers themselves to consider (although these may even out ?). Old threads can be interesting LOL.