Welcome, Guest. Please login or register.
Did you miss your activation email?
May 26, 2012, 04:39:01 AM

Login with username, password and session length
Search:     Advanced search
* Home Help Search Calendar Links Login Register
+  Machsupport Forum
|-+  Mach Discussion
| |-+  General Mach Discussion
| | |-+  Feed Hold and Stop
Pages: 1 2 »   Go Down
Print
Author Topic: Feed Hold and Stop  (Read 731 times)
0 Members and 1 Guest are viewing this topic.
archcomp
Active Member

Offline Offline

Posts: 8


View Profile
« on: June 06, 2010, 06:25:41 AM »

Frieds,
While using Mach3 Mill, while I use stop button, axis stops abruptly and axis positioning is also showing variation. This does not happen with Feed Hold. I understand that Mach stops the movement and other outputs at the earliest when Stop command given. Should this cause loss of steps with servos? Because, Mach is able to maintain the count of the steps given to the output and so is servo. I fail to understand why should I lose steps in this case.
Thanks.
Mahesh Vyas.
Logged
manmeran
Active Member

Offline Offline

Posts: 241


View Profile
« Reply #1 on: June 06, 2010, 06:40:20 AM »

if you use Stop ,Axes are stopped immediately and cause LOSS STEP
but when you use FEEDHOLD dont loss step and it is safe

Amir
Logged

**Even a clock that does not work is right twice a day**
ger21
Global Moderator
*
Offline Offline

Posts: 2,612



View Profile WWW
« Reply #2 on: June 06, 2010, 07:21:42 AM »

Stop is an instant stop with no deceleration. If you want to stop in the middle of a program, use Feedhold and wait until all motion stops, then use Stop
Logged

archcomp
Active Member

Offline Offline

Posts: 8


View Profile
« Reply #3 on: June 06, 2010, 08:19:37 AM »

Garry, and Man,
Thanks for the insight. I agree, it is a stop without deceleration. But I feel thMach will know how many pulses it has applied to the axis and servo having received that much must be in sync with Mach even though stop is issued. It is possible to lose steps for Stppper Motors as there is a sudden stop and having no feed back, it has no reason to know that the the motor has over travelled due to inertia. But servo will take care of the sudden stop, and will not loose sync with Mach. Please correct me if wrong. Stop is required, when the deceleration travel undertaken by Mach and obeyed by servo may damange some thing on machine to warrant Stop action.
Mahesh Vyas.
Logged
HimyKabibble
V4 Screen Contributor

Offline Offline

Posts: 1,347



View Profile
« Reply #4 on: June 06, 2010, 08:52:39 AM »

Garry, and Man,
Thanks for the insight. I agree, it is a stop without deceleration. But I feel thMach will know how many pulses it has applied to the axis and servo having received that much must be in sync with Mach even though stop is issued. It is possible to lose steps for Stppper Motors as there is a sudden stop and having no feed back, it has no reason to know that the the motor has over travelled due to inertia. But servo will take care of the sudden stop, and will not loose sync with Mach. Please correct me if wrong. Stop is required, when the deceleration travel undertaken by Mach and obeyed by servo may damange some thing on machine to warrant Stop action.
Mahesh Vyas.

Your assumption is correct only if the axis is moving slowly enough that the machine does not "coast", due to inertia, far enough to cause the servo to exceed its allowable following error and fault.  If you're using Geckos, the maximum following error is only 128 counts.  On my machine, that's only 0.0064", which is not much when the 400# table is moving at 300 IPM.

Regards,
Ray L.
Logged

Regards,
Ray L.
manmeran
Active Member

Offline Offline

Posts: 241


View Profile
« Reply #5 on: June 06, 2010, 08:55:53 AM »

Quote
But servo will take care of the sudden stop
i think is wrong
not different between step motor & servo motor In this particular issue.
i think if we use "STOP" In all circumstances there is likely lost steps.Although this amount is small

Amir
« Last Edit: June 06, 2010, 08:58:17 AM by manmardam » Logged

**Even a clock that does not work is right twice a day**
archcomp
Active Member

Offline Offline

Posts: 8


View Profile
« Reply #6 on: June 06, 2010, 11:09:47 PM »

Ray, Amir,
I think this discussion may help sort out the confusion about Stop and Feed Hold. When Stop is used, it stops all the outputs like Spindle, coolant and pulse as soon as possible. But Mach is able to keep track of issued pulses to the axis. Servos are normally behind the pulse train being issued by Mach and is following it by the counts what we call "Following Error." So, when pulse train is suddenly stops, the servo which is slightly behind the actual issued pulses, will be able to take care of sudden stop. If the inertia is heavy and speed is high at the time of Stop, servo may overshoot a little and then will return back at the last commanded pulse count. Hence Stop should not result in lost steps if servos are used and not induced to the Fault mode. Definitely, stop being sudden, it exerts  heavy thud  on the machine components. Comments are invited.
Mahesh Vyas.
Logged
HimyKabibble
V4 Screen Contributor

Offline Offline

Posts: 1,347



View Profile
« Reply #7 on: June 07, 2010, 08:47:00 AM »

Ray, Amir,
I think this discussion may help sort out the confusion about Stop and Feed Hold. When Stop is used, it stops all the outputs like Spindle, coolant and pulse as soon as possible. But Mach is able to keep track of issued pulses to the axis. Servos are normally behind the pulse train being issued by Mach and is following it by the counts what we call "Following Error." So, when pulse train is suddenly stops, the servo which is slightly behind the actual issued pulses, will be able to take care of sudden stop. If the inertia is heavy and speed is high at the time of Stop, servo may overshoot a little and then will return back at the last commanded pulse count. Hence Stop should not result in lost steps if servos are used and not induced to the Fault mode. Definitely, stop being sudden, it exerts  heavy thud  on the machine components. Comments are invited.
Mahesh Vyas.

No, you're not understanding what I said - Every servo tracks a value, called following error, which is the difference between where its encoder says it is is, and where it is actually supposed to be.  Every servo controller also has a limit to how larage a following error it will allow.  If this limit is ever exceeeded, the servo controller will fauilt.  For Geckos, the maximum following error allowed is only 128 steps, which typically maps to only a few thousandths of an inch of axis motion.  If the axis position is ever more than 128 steps from where it's supposed to be at any given time, the Gecko will reset itself, which will cause its position to no longer match where Mach3 thinks it is.

And when Mach3 does a Stop, it does not stop "as soon as possible".  It simply stops issuing step pulese RIGHT NOW.  There is NO deceleration whatsoever.  It ceases to output step pulses the instant it recognizes that a Stop has been requested.

Regards,
Ray L.
Logged

Regards,
Ray L.
BluePinnacle
Active Member

Offline Offline

Posts: 229



View Profile
« Reply #8 on: June 07, 2010, 05:44:02 PM »

A point worth noting; using feedhold stops the program at the end of the line it's on within the main program. So if that line is M98 P5 L100, and p5 takes 5 minutes, you will have a long wait for it to actually stop.
Logged
ger21
Global Moderator
*
Offline Offline

Posts: 2,612



View Profile WWW
« Reply #9 on: June 07, 2010, 06:33:31 PM »

There is the option in General Config, something like  "no FRO on que" which will feedhold instantly, at the expense of FRO response.
Logged

Pages: 1 2 »   Go Up
Print
Jump to:  

Powered by MySQL Powered by PHP Powered by SMF 1.1.16 | SMF © 2011, Simple Machines Valid XHTML 1.0! Valid CSS!