Dynomotion

Group: DynoMotion Message: 11112 From: fouijar Date: 2/18/2015
Subject: Lag or vibration on small moves

Hello Tom,


I have found a strange issue on my system behaviour.

Mainly on small moves and short transitions there's a "lag" under transient motion state which causes a vibration. To illustrate my word, I joined a couple pictures/video of the test part and associated moves.

The excel file contains vibrations measurements and graphs related to the problem.

This link is a Dropbox folder where videos and all files are available for more detailed understanding.

I captured a video of the machine executing the code and in this video you can hear the noise from the broken moves.


I first inspected all parameters: mechanical, drive, regulation etc. I found some possible issues and fixed them.

Under step moves and/or bode plot, I can't simulate or more exactly reproduce this lag/vibration.


I also changed the acceleration values and lowered them to check that option and this doesn't change anything.


Well, If you have a fresh point of view on this particular issue it would be great.


Regards,


Jérôme



   @@attachment@@
Group: DynoMotion Message: 11114 From: Tom Kerekes Date: 2/18/2015
Subject: Re: Lag or vibration on small moves [4 Attachments]
Hi Jerome,

I assume you are running Mach3?  Have you tried all the CV (Constant Velocity) related settings like Lookahead=200 ?

BTW There is an example KFLOP C Program called CaptureXYZMotionToFile that should allow you to capture actual Trajectory Data to a disk file that you can then plot in Excel.  That wwould probably be a lot more useful than your acceleration measurements.

Regards
TK

 
From: "fouijar@... [DynoMotion]" <DynoMotion@yahoogroups.com>
To: DynoMotion@yahoogroups.com
Sent: Wednesday, February 18, 2015 3:43 PM
Subject: [DynoMotion] Lag or vibration on small moves [4 Attachments]

 
[Attachment(s) from fouijar@... [DynoMotion] included below]
Hello Tom,

I have found a strange issue on my system behaviour.
Mainly on small moves and short transitions there's a "lag" under transient motion state which causes a vibration. To illustrate my word, I joined a couple pictures/video of the test part and associated moves.
The excel file contains vibrations measurements and graphs related to the problem.
This link is a Dropbox folder where videos and all files are available for more detailed understanding.
I captured a video of the machine executing the code and in this video you can hear the noise from the broken moves.

I first inspected all parameters: mechanical, drive, regulation etc. I found some possible issues and fixed them.
Under step moves and/or bode plot, I can't simulate or more exactly reproduce this lag/vibration.

I also changed the acceleration values and lowered them to check that option and this doesn't change anything.

Well, If you have a fresh point of view on this particular issue it would be great.

Regards,

Jérôme
Group: DynoMotion Message: 11116 From: fouijar Date: 2/19/2015
Subject: Re: Lag or vibration on small moves [4 Attachments]
Hello Tom,

About the parameters:
CV ok, not changed
Lookahead from 20 to 200, no changes

Attached the C file used to capture motion data, the output file and the excel file with graphs included. It's obvious that the error is somewhat different from X axis compared to Y axis, would it be a sufficient reason to produce such a lag? It's seems there's a correlation between the broken motion and the error. But it does'nt explain where the error comes from.
Under the generic tests I can't reproduce that amount of error but for sure if you generate motion like a square wave you'll find it.

The previously given acceleration data has the advantage to clearly show the vibration issue, it was to illustrate my word only.

What do you think?

Jerome
   @@attachment@@
Group: DynoMotion Message: 11123 From: Tom Kerekes Date: 2/21/2015
Subject: Re: Lag or vibration on small moves [4 Attachments]
Hi Jerome,

I plotted your Captured data and there are a lot of sudden stops and starts I suspect related to the Mach3 CV Trajectory Planning.  There are a number of confusing Mach3 settings regarding this.  Were you aware of these?  Have you tried adjusting them?  See:


You don't want to switch to KMotionCNC?

Regards
TK



Group: DynoMotion Message: 11125 From: fouijar Date: 2/22/2015
Subject: Re: Lag or vibration on small moves [4 Attachments]
Hi Tom,

Well I'm on the same line as you're. I also think it's time for me to change and use kmotion instead, the cv options weren't successful.
I already have configured the basics and still need to tweak a little the settings to get the machine running smoothly. But from the first trial it seems to be much better than lately under mach3.
I'll need to rewrite my m6 file from vb to c and have a simple set of questions:
G28, where is it possible to define it?
Does my probe program needs to be implemented within the init file?

I expect to have some questions coming within a near future. ..

Thanks for your attention,

Jérôme
Group: DynoMotion Message: 11126 From: Tom Kerekes Date: 2/23/2015
Subject: Re: Lag or vibration on small moves
Hi Jerome,

Regarding G28 in KMotionCNC:  The 6 axis reference position is defined in specific GCode Variables as:

 5161, 5162, 5163,   /* G28 home */
 5164, 5165, 5166,
 5181, 5182, 5183,   /* G30 home */
 5184, 5185, 5186,
Change those in the *.var file you are using.


Regarding Probing in KMotionCNC:There isn't any specific G Code supported.  But any M Code might be used to run a C Program Probe Sequence.  This would normlly be separate from your Init C Program and only executed when called from GCode.   Here is an example of passing parameters to a C Program.  See also the ProbDirection.c example.

Regards
TK