Machsupport Forum

Mach Discussion => General Mach Discussion => Topic started by: octavio_bs on May 11, 2009, 08:07:43 AM

Title: extrange behavior in Z axis
Post by: octavio_bs on May 11, 2009, 08:07:43 AM
Dear friends, something weird is happening me from time to time.

I will explain

I create a piece in cambam and export gcode, i verify the g code by hand ( you know what i mean ) and also with a couple of simulators, also in mach the gcode seems to be correct...

Well the problem is that sometimes, can not reproduce the problem consistently, the z axis gos lower than expected, lets say that the progarm says to go z-10 and ig goes to z-13 and from then on everything is done with this -3 offset, so when it is instructed to go to z1.5 it is really on z-1.5 and... piece damaged, mill broken, etc...

The explanation given is just an example, it can have -3 added or maybe -1.0935 or whathever, the problem is that z goes down more than expected, but if the gcode is reexecuted it will work correctly and maybe tomorrow it will fail again but in a different line of code.

I have only seen this problem with the Z axis

Any clues?

Sorry about my english :-)

Title: Re: extrange behavior in Z axis
Post by: Overloaded on May 11, 2009, 08:19:23 AM
Does the Z axis DRO reflect the offset ?
If not, you are likely loosing steps during a +Z rapid move then the remainder of z moves will be offse that amount.
Just a hunch,
RC
Title: Re: extrange behavior in Z axis
Post by: octavio_bs on May 11, 2009, 08:30:55 AM
The DRO does not reflect the offset and it is not lost steps, it is a lot more steps sent to the controller, also the z movement is made very slow and the machine has a lot more torque than really needed.
Title: Re: extrange behavior in Z axis
Post by: Overloaded on May 11, 2009, 09:48:25 AM
  I think the DRO accurately represents the signals sent from Mach. So if your DRO shows the position correctly as coded there probably is not "a lot more steps sent to the controller". Maybe a mechanical slippage....pulleys, coupling or ? ?.
RC
Title: Re: extrange behavior in Z axis
Post by: octavio_bs on May 11, 2009, 09:57:17 AM
I have seen the machine moving up to 3mm more than expected, my machine is a proxxon mf70 ( a toy I know ) and it has 1mm movevent per turn so it is up to three complete turns ( approx ) at 4 microsteps that makes about 2400 steps.

Also there is no way of mechanical slip, the stepper has a flat in its axis and the "long screw" ( cant rememnber the english word ) has also a mechanical blocking, in short , no way for a mechanical slip
Title: Re: extrange behavior in Z axis
Post by: Hood on May 11, 2009, 10:37:29 AM
What makes you say that there are more steps sent to the controller? As RC has said and as far as I am aware Mach counts what it actually puts out of the port so if the DRO reads correctly then it is what Mach has pout out.
Could it be noise affecting the outputs to the drive?

Hood
Title: Re: extrange behavior in Z axis
Post by: octavio_bs on May 11, 2009, 10:48:04 AM
I thought about noise, but that will trigger an step here and there, but when the system "gets crazy" it keeps moving at the same speed ( it doesn't matters it it is F30 F100 or F55 ) and noise from steppers is as smooth as in any other move.

When the system doesn't get crazy, most of the times it works perfectly, the result is as expected and have never seen lost steps. Have put a micrometer and repeated movements up to two hundred times and repeatability is correct.

So is that I am lost, thinking about incompatibility with any other soft, antiviral or so.

Thanks for your efforts
Title: Re: extrange behavior in Z axis
Post by: Hood on May 11, 2009, 10:52:08 AM
Can you swap the drive on the Z for the one on the X and see if the problem stays with the Z or moves to the X.
Hood
Title: Re: extrange behavior in Z axis
Post by: octavio_bs on May 11, 2009, 11:02:41 AM
have already changer the driver, but will try to change the stepper as have an spare one

will get back with results, but as the fail only happens from time to time will take some days to have an asnwer
Title: Re: extrange behavior in Z axis
Post by: Hood on May 11, 2009, 11:09:31 AM
Is this a demo version? If licenced is it your name it is licenced to?
Hood
Title: Re: extrange behavior in Z axis
Post by: octavio_bs on May 11, 2009, 11:16:48 AM
it is licenced, will send you a privete message

Octavio
Title: Re: extrange behavior in Z axis
Post by: octavio_bs on May 23, 2009, 04:10:41 AM
Well...

I have changed the stepper and the driver and the problem persists.

Have tested with the oscilloscope and the driver continues to receive pulses when mach3 goes nuts, but the DRO does not updates correctly but haven't been able to reproduce the problem consistently, so I can“'t give a base to investigate the problem.

Now I am going to install a second hard disk with a fresh english windows XP installation and will install only mach3 on this disk just to see if this solves the problem.

will give some news in a few days.

Regards: Octavio
Title: Re: extrange behavior in Z axis
Post by: octavio_bs on June 02, 2009, 11:26:11 AM
Well... it's me again....

since I put a fresh English windows xp everything seems to be correct, so the moral is do not put anything you don't need in your mach partition

thanks