Page 5 of 8

Re: DSP Commercial Controller Support

PostPosted: Tue Feb 07, 2012 7:01 am
by irfanulla
I have the X swing activated - but still fires one way ! hmmm, looks so interesting :)

Re: DSP Commercial Controller Support

PostPosted: Sun May 13, 2012 7:43 pm
by gavztheouch
Is it possible to compensate for backlash when cutting in vector mode?

Cheers

Gav

Re: DSP Commercial Controller Support

PostPosted: Mon May 14, 2012 1:45 pm
by twehr
gavztheouch wrote:Is it possible to compensate for backlash when cutting in vector mode?

Cheers

Gav


Not that I know of. If you have enough backlash that it is affecting your vectors, you really need to find and eliminate it.

Show us what your issue is.

Confused by Pulse unit settings

PostPosted: Sun May 20, 2012 6:08 pm
by gavztheouch
Hi,

Has anyone got some knowledge of the pulse unit settings, particulary when using a number above 100. After 100 there seems to be no change in the distance my motor moves. 100, 1000, 4000 all move the motor the same distance? Is there an upper limit, do I need to increase my micro stepping to bring down my distance per step?

Cheers
Gav

Re: DSP Commercial Controller Support

PostPosted: Mon May 21, 2012 3:37 am
by mattrsch
I would increase the microstepping. 100 um/step is only about 250 steps/inch, which isn't all that great for a laser. From what I gathered most people run their lasers in the 1000-2000 steps/inch range (mine is 2000).

Re: DSP Commercial Controller Support

PostPosted: Mon May 21, 2012 8:47 am
by gavztheouch
Thanks Mattrsch,

My system is set to move 28.8mm per rev of the motor. My servos have an 2500 pulse encoder and *4 multiplication on the encoder making 10,000 in total per rev. so I should need a value of (28.8/10,000)*1000 = 2.88 um. I think I need to check out my setting inside my servos, the owner before me may have changed the encoder scaling values, so it is not running at its highest resolution. At the moment a value of 2.88um on the yaxis causing about 20 metres of movement on what should be 80mm. When I try and compensate inside the dsp for the low res its seems to max out at 30um. Anything above that and it goes crazy.

Re: DSP Commercial Controller Support

PostPosted: Tue Jul 17, 2012 10:37 pm
by gavztheouch
Not that I know of. If you have enough backlash that it is affecting your vectors, you really need to find and eliminate it.


Twehr, I have fixed my backlash issue yet I am still having problems with backlash style errors, is it possible that having the acceleration set too low would cause this.

Thanks Gav

Re: DSP Commercial Controller Support

PostPosted: Wed Jul 18, 2012 2:05 am
by twehr
gavztheouch wrote:
Not that I know of. If you have enough backlash that it is affecting your vectors, you really need to find and eliminate it.


Twehr, I have fixed my backlash issue yet I am still having problems with backlash style errors, is it possible that having the acceleration set too low would cause this.

Thanks Gav


How much backlash are you talking about?

Does it show up in vector as well as engraving?

Is it one axes only or both?

Re: DSP Commercial Controller Support

PostPosted: Wed Jul 18, 2012 5:39 pm
by gavztheouch
Thanks Tim, here is a bit more info,

Is it one axes only or both? Y-axis only

Does it show up in vector as well as engraving? Yes and it appears to be dynamic as the speed is increased the backlash increases.

How much backlash are you talking about? I have <0.05mm in the y axis this is mechanical backlash, to confirm this I have started using mach3 as the controller and at much higher speed than I have been using the dsp at, it maintains less than 0.05mm backlash. In contrast the DSP is giving me up to and over 1mm backlash, in both vector and raster modes.

This is a link to my buildlog http://www.buildlog.net/forum/viewtopic.php?f=16&t=1316&p=12774#p12774

Cheers

Gav

Re: DSP Commercial Controller Support

PostPosted: Thu Jul 19, 2012 12:56 pm
by twehr
gavztheouch wrote:Thanks Tim, here is a bit more info,

Is it one axes only or both? Y-axis only

Does it show up in vector as well as engraving? Yes and it appears to be dynamic as the speed is increased the backlash increases.

How much backlash are you talking about? I have <0.05mm in the y axis this is mechanical backlash, to confirm this I have started using mach3 as the controller and at much higher speed than I have been using the dsp at, it maintains less than 0.05mm backlash. In contrast the DSP is giving me up to and over 1mm backlash, in both vector and raster modes.

This is a link to my buildlog http://www.buildlog.net/forum/viewtopic.php?f=16&t=1316&p=12774#p12774

Cheers

Gav


I doubt that it is anything inherent in the DSP. If it were, then I would not expect to see it's effect to be speed-dependent. There may be a difference between how the DSP does acceleration compared to Mach3. You mentioned slow acceleration as a possibility. I would lean the other way. The Y axis has a lot more mass to move around than does the X axis. Overcoming inertia (acceleration or deceleration) could possibly produce different speed-dependent results. Since it is less noticeable with Mach3 and more so with DSP, the algorithm for implementing acceleration may be different.

One other person reported a y-only backlash. Turns out it was the motor couplings he used. When he switched to solid couplings, the issue went away. The springiness in the others was causing the problem on change of Y direction.

A test that may help to determine if it is mechanical or not, would be to make a small test and change the x/y axis (x axis to y drive, etc). If the error stays in the Y axis, it is likely mechanical. If it moves to the x axis, it is more likely the controller.