Hi Steve,
I haven't noticed any delays before. With any Windows Application the response should be within milliseconds most of the time with an occasional much longer delay if Windows happens to be busy.
Could it just be your Jerk or Acceleration settings? KFLOP does Arrow Key Jog motion using 3rd order Jerk Limited motion using the Acceleration and Jerk Settings in KFLOP which can be different then the Mach3 Acceleration (with infinite Jerk). This can potentially allow higher speed, and smoother Jog motion than would otherwise be possible.
A low Jerk setting sometimes appears as a delay because of the time required to ramp up acceleration, in order to ramp up velocity, in order to finally get some position movement. The Jerk Setting in KFLOP
is normally set at least 10X the value of the Acceleration in order to ramp up the acceleration in less than 0.1sec.
Regards TK
Group: DynoMotion |
Message: 7522 |
From: Steve |
Date: 5/24/2013 |
Subject: Re: Delay in Jogging. Keypress-to-movement. Fixable? |
I searched for posts on this topic and it seems a setting of 100X was being proposed, so I tried that and it seems to have eliminated the long delay.
As this was the last issue to be resolved, I purchased my second Kflop today to use with another project.
I am too busy right now to dig into the Kflop's extensive capabilities, but I do have it working now with MACH which was the initial priority.
I understand that the init file has many params needed to control servos, but I use industrial drives that have all of the capabilities built into the drives (and configurable/changeable on the fly thru commands or signals), so what I want is just the unaltered, unfiltered, unmolested motion in step/dir form of course.
Is there a recommended setup that would pass thru the exact motion without regard for accel/jerk/ or any other characteristics of the motion?
Thank you for all the prompt help.
- Steve
--- In DynoMotion@yahoogroups.com, Tom Kerekes <tk@...> wrote:
>
> Hi Steve,
>
> I haven't noticed any delays before. With any Windows Application the response should be within milliseconds most of the time with an occasional much longer delay if Windows happens to be busy.
>
> Could it just be your Jerk or Acceleration settings? KFLOP does Arrow Key Jog motion using 3rd order Jerk Limited motion using the Acceleration and Jerk Settings in KFLOP which can be different then the Mach3 Acceleration (with infinite Jerk). This can potentially allow higher speed, and smoother Jog motion than would otherwise be possible.
>
>
> A low Jerk setting sometimes appears as a delay because of the time required to ramp up acceleration, in order to ramp up velocity, in order to finally get some position movement. The Jerk Setting in KFLOP is normally set at least 10X the value of the Acceleration in order to ramp up the acceleration in less than 0.1sec.
>
> Regards
> TK
>
>
>
>
>
> ________________________________
> From: Steve <steve@...>
> To: DynoMotion@yahoogroups.com
> Sent: Tuesday, May 14, 2013 6:46 AM
> Subject: [DynoMotion] Delay in Jogging. Keypress-to-movement. Fixable?
>
>
>
> Â
> Using keyboard arrow keys, there is a noticable delay between the keypress and machine movement. This delay was not present with the smoothstepper so I tend to believe the delay is coming from the Kflop.
>
> I have tried adjusting the debounce interval, just in case it was tied in somehow, but it seems to have little effect.
>
> Is Kflop putting an intentional delay in this function? If so, where can it be found and adjusted or eliminated?
>
> Thanks,
>
> - Steve
>
|
|
Group: DynoMotion |
Message: 7525 |
From: Tom Kerekes |
Date: 5/24/2013 |
Subject: Re: Delay in Jogging. Keypress-to-movement. Fixable? |
Hi Steve,
Regarding the External Step/Dir Drives unmolested motion: I don't really understand how that could work. If the drives close the servo position loop themselves then yes all the PID tuning and servo parameters will be handled by the drive and you will not need to configure any of those in KFLOP. From KFLOP's perspective the drive is open loop just like a Stepper Drive. However the motion trajectories (which generate the Step/Dir pulse sequences) still need to be
configured in KFLOP/Mach3 so that all the axes move in a coordinated manner and so forth. It sounds like you are thinking that if the desire was to move an axis 1 inch that KFLOP would just send a quick burst of 1 inch of pulses and then the drive would handle the acceleration profile to move 1 inch? I don't think that would work for cases like circular interpolation. I suppose I'm missing your point.
Regards TK
Group: DynoMotion |
Message: 7534 |
From: Steve |
Date: 5/27/2013 |
Subject: Re: Delay in Jogging. Keypress-to-movement. Fixable? |
My point has to do with using Kflop with MACH only. Since I do not know what kind of informtaiton a plug-in receives from MACH, it is difficult to be concise in what I am asking.
I understand the coorinated motion requirment, but is that done inside MACH or does the motion control device have to do those calculations? For example, MACH has settable accell which seems to be active and effects the behavior of the drives, so I make the assumption that the accell setting (in MACH) has already been taken into account before the data is passed to the plug-in.
Following on that assumption, I was looking for some example setting in Kflop that would be far faster than MACHs so that the Kflop would not 'slow down' the accell over what MACH is already doing. This seems to be what was happening with the jerk setting/jog delay issue.
There is not a problem that needs to be solved at this point, I just want to have a better understanding of how the system works and what the interactions are.
- Steve
--- In DynoMotion@yahoogroups.com, Tom Kerekes <tk@...> wrote:
>
> Hi Steve,
>
> Regarding the External Step/Dir Drives unmolested motion: I don't really understand how that could work. If the drives close the servo position loop themselves then yes all the PID tuning and servo parameters will be handled by the drive and you will not need to configure any of those in KFLOP. From KFLOP's perspective the drive is open loop just like a Stepper Drive. However the motion trajectories (which generate the Step/Dir pulse sequences) still need to be configured in KFLOP/Mach3 so that all the axes move in a coordinated manner and so forth. It sounds like you are thinking that if the desire was to move an axis 1 inch that KFLOP would just send a quick burst of 1 inch of pulses and then the drive would handle the acceleration profile to move 1 inch? I don't think that would work for cases like circular interpolation. I suppose I'm missing your point.
>
> Regards
> TK
>
>
>
>
>
> ________________________________
> From: Steve <steve@...>
> To: DynoMotion@yahoogroups.com
> Sent: Friday, May 24, 2013 2:33 PM
> Subject: [DynoMotion] Re: Delay in Jogging. Keypress-to-movement. Fixable?
>
>
>
> Â
> I searched for posts on this topic and it seems a setting of 100X was being proposed, so I tried that and it seems to have eliminated the long delay.
>
> As this was the last issue to be resolved, I purchased my second Kflop today to use with another project.
>
> I am too busy right now to dig into the Kflop's extensive capabilities, but I do have it working now with MACH which was the initial priority.
>
> I understand that the init file has many params needed to control servos, but I use industrial drives that have all of the capabilities built into the drives (and configurable/changeable on the fly thru commands or signals), so what I want is just the unaltered, unfiltered, unmolested motion in step/dir form of course.
>
> Is there a recommended setup that would pass thru the exact motion without regard for accel/jerk/ or any other characteristics of the motion?
>
> Thank you for all the prompt help.
>
> - Steve
>
> --- In DynoMotion@yahoogroups.com, Tom Kerekes <tk@> wrote:
> >
> > Hi Steve,
> >
> > I haven't noticed any delays before.ÃÂ With any Windows Application the response should be within milliseconds most of the time with an occasional much longer delay if Windows happens to be busy.
> >
> > Could it just be your Jerk or Acceleration settings?ÃÂ KFLOP does Arrow Key Jog motion using 3rd order Jerk Limited motion using the Acceleration and Jerk Settings in KFLOP which can be different then the Mach3 Acceleration (with infinite Jerk).ÃÂ This can potentially allow higher speed, and smoother Jog motion than would otherwise be possible.
> >
> >
> > A low Jerk setting sometimes appears as a delay because of the time required to ramp up acceleration, in order to ramp up velocity, in order to finally get some position movement.ÃÂ The Jerk Setting in KFLOP is normally set at least 10X the value of the Acceleration in order to ramp up the acceleration in less than 0.1sec.
> >
> > Regards
> > TK
> >
> >
> >
> >
> >
> > ________________________________
> > From: Steve <steve@>
> > To: DynoMotion@yahoogroups.com
> > Sent: Tuesday, May 14, 2013 6:46 AM
> > Subject: [DynoMotion] Delay in Jogging. Keypress-to-movement. Fixable?
> >
> >
> >
> > ÃÂ
> > Using keyboard arrow keys, there is a noticable delay between the keypress and machine movement. This delay was not present with the smoothstepper so I tend to believe the delay is coming from the Kflop.
> >
> > I have tried adjusting the debounce interval, just in case it was tied in somehow, but it seems to have little effect.
> >
> > Is Kflop putting an intentional delay in this function? If so, where can it be found and adjusted or eliminated?
> >
> > Thanks,
> >
> > - Steve
> >
>
|
|
Group: DynoMotion |
Message: 7535 |
From: Tom Kerekes |
Date: 5/27/2013 |
Subject: Re: Delay in Jogging. Keypress-to-movement. Fixable? |
Hi Steve, Mach3 has two modes of motion. One when running a job and one when jogging. You might read the Mach3 "Bible" that Art wrote. When running a job the trajectory planning is performed by Mach3. There is a "ring buffer" that continuously supplies positions where all the axes should be at fixed time increments. It is upto the Plugin, controller, and drives to make sure all the axes are at those positions at those times for proper coordinated motion.
Jogging is different. Mach3 just tells the Plugin to move one or more axes to move at a desired speed. It is up to the Plugin to make that
happen. This is how how our Plugin+KFLOP is able to make 3rd order Jog motions. And it is how we are able to ignore the Acceleration setting in Mach3 and use the Acceleration and Jerk Settings in KFLOP. For Jogs I suppose it would be possible to pass on the Mach3 request to an intelligent drive and just tell it to perform the motion profiling and blending to the speeds requested by Mach3. But a Step/Dir interface wouldn't be appropriate for this. Something like Modbus would be needed to transfer parameters and speed requests. But then Modbus wouldn't be appropriate specify the interpolated motion.
HTH Regards TK
Group: DynoMotion |
Message: 7536 |
From: Wcarrothers Yahoo |
Date: 5/27/2013 |
Subject: Re: Delay in Jogging. Keypress-to-movement. Fixable? |
I have been very happy with k and Mach. I should use dynomotion but to attached to Mach. But I've had the opposite complaint on my parallel port machine when I tried a smooth stepper. Where I that case the Jog movements with the SS were delayed when using my pendant. But in parallel mode it was not. Was relieved That the movement with the K was handled internal so did not have Any delay
Perhaps you would have less delay if you setup a pendant to enter the moves into the k io direct On May 27, 2013, at 1:51 PM, Tom Kerekes <tk@...> wrote:
Hi Steve, Mach3 has two modes of motion. One when running a job and one when jogging. You might read the Mach3 "Bible" that Art wrote. When running a job the trajectory planning is performed by Mach3. There is a "ring buffer" that continuously supplies positions where all the axes should be at fixed time increments. It is upto the Plugin, controller, and drives to make sure all the axes are at those positions at those times for proper coordinated motion.
Jogging is different. Mach3 just tells the Plugin to move one or more axes to move at a desired speed. It is up to the Plugin to make that
happen. This is how how our Plugin+KFLOP is able to make 3rd order Jog motions. And it is how we are able to ignore the Acceleration setting in Mach3 and use the Acceleration and Jerk Settings in KFLOP. For Jogs I suppose it would be possible to pass on the Mach3 request to an intelligent drive and just tell it to perform the motion profiling and blending to the speeds requested by Mach3. But a Step/Dir interface wouldn't be appropriate for this. Something like Modbus would be needed to transfer parameters and speed requests. But then Modbus wouldn't be appropriate specify the interpolated motion.
HTH Regards TK
Group: DynoMotion |
Message: 7537 |
From: Tom Kerekes |
Date: 5/27/2013 |
Subject: Re: Delay in Jogging. Keypress-to-movement. Fixable? |
Hi William,
I think the original problem was just low Acceleration and Jerk settings in KFLOP appearing like a delay.
But I agree a pendant directly connected to
KFLOP will always give the best guaranteed response.
Thanks TK
Group: DynoMotion |
Message: 7538 |
From: Steve Simpson |
Date: 5/28/2013 |
Subject: Re: Delay in Jogging. Keypress-to-movement. Fixable? |
Thanks, this explains a lot. I'm glad I asked this question so that I know the accel settings in MACH only effect jogging. My rigid tapping macro will need to be reworked because I set the Z axis accell to an extreme level for tapping and then set it back after the tapping is completed. I would have been pulling my hair out trying to figure out why the accel was not changing.
- Steve | | | |
| | | | | | | |