Page 1 of 1

speed specs of laser 2.x

PostPosted: Wed Jan 16, 2013 4:20 pm
by MasonAtom
I've been looking all around but I can't seem to find any specs on the laser 2.x. In particular, I am looking for engraving speed. I'd like to see how it compares to my current cheapo chinese laser that basically tops out at 400 mm/s in any useable sense.

Mason

Re: speed specs of laser 2.x

PostPosted: Wed Jan 16, 2013 4:42 pm
by bdring
The 2.x can go a lot faster than that, but both are using the same types of tubes so you are limited by the power and physics. If you use the DSP controller, that is better than the cheapo Chinese controllers.

Re: speed specs of laser 2.x

PostPosted: Wed Jan 16, 2013 6:56 pm
by BenJackson
My laser's top speed has been 470mm/s (due to PC latency -- but I could have reduced microstepping to go faster) but I just got a Mesa Electronics card to drive it and I've played around with 1000mm/s with no problems. At some point acceleration becomes the bigger factor if you want to overscan your engraving enough to get up to speed. I have not been focused on speed lately so I don't really know what the top end is.

Re: speed specs of laser 2.x

PostPosted: Wed Jan 16, 2013 7:39 pm
by MasonAtom
BenJackson wrote:My laser's top speed has been 470mm/s (due to PC latency -- but I could have reduced microstepping to go faster) but I just got a Mesa Electronics card to drive it and I've played around with 1000mm/s with no problems. At some point acceleration becomes the bigger factor if you want to overscan your engraving enough to get up to speed. I have not been focused on speed lately so I don't really know what the top end is.


Thanks for the input guys.

Yes, in my system I find that if I use speeds over 400 mm/s it actually slows things down a bit because most of the time I'm engraving across 2" or less increments. So in this sort of configuration, so in this sort of setting, if you increase the speed, the overshoots gets larger and the cumulative time increases due to that.

The reason I asked is that for laser 2.x, the laser head's movement is rotated 90 w/r/t to my current system, so the engraving path would now be very long, allowing me to make better use of a higher speed. Still working things out for my build, hopefully I'll catch batch #7 and get an order in before they sell out.

Mason

Re: speed specs of laser 2.x

PostPosted: Wed Jan 16, 2013 11:43 pm
by BenJackson
MasonAtom wrote:Yes, in my system I find that if I use speeds over 400 mm/s it actually slows things down a bit because most of the time I'm engraving across 2" or less increments. So in this sort of configuration, so in this sort of setting, if you increase the speed, the overshoots gets larger and the cumulative time increases due to that.

Here's the part of the README for my LinuxCNC config for the 2.x Laser:
The overscan is to give the laser carriage time to accelerate to full speed.
If the laser is still accelerating within the field of the image then the
left/right edges will be engraved slightly darker. You can compute the
exact distance needed as 0.5 * F^2 / A where F is feed speed (in mm/s, not
mm/min) and A is [AXIS_0]MAX_ACCELERATION. Assuming you always engrave
at top speed ([AXIS_0]MAX_VELOCITY = 470mm/s) and modulate power to suit,
the value of overscan should be about 15mm. If you tune the accel faster
or engrave at a lower speed you can use smaller overscan. If you are willing
to tolerate some uneven engraving power at the edges you can turn overscan
way down to get near the edge of your work area. I have not tested with
overscan of 0, that might break the algorithm.


(You can see my whole config at https://github.com/bjj/2x_laser ) I've not determined the max X acceleration, I just quit increasing it when it was pretty snappy. I'll probably go higher for a higher top speed. My Y accel is actually less than the machine can accomplish: I increased it until rapids were capable of knocking the open lid over center (!) and backed it off from there. My machine is on a table over carpet which makes it wobble more than I'd like.

Re: speed specs of laser 2.x

PostPosted: Thu Jan 17, 2013 1:18 pm
by MasonAtom
BenJackson wrote:
MasonAtom wrote:Yes, in my system I find that if I use speeds over 400 mm/s it actually slows things down a bit because most of the time I'm engraving across 2" or less increments. So in this sort of configuration, so in this sort of setting, if you increase the speed, the overshoots gets larger and the cumulative time increases due to that.

Here's the part of the README for my LinuxCNC config for the 2.x Laser:
The overscan is to give the laser carriage time to accelerate to full speed.
If the laser is still accelerating within the field of the image then the
left/right edges will be engraved slightly darker. You can compute the
exact distance needed as 0.5 * F^2 / A where F is feed speed (in mm/s, not
mm/min) and A is [AXIS_0]MAX_ACCELERATION. Assuming you always engrave
at top speed ([AXIS_0]MAX_VELOCITY = 470mm/s) and modulate power to suit,
the value of overscan should be about 15mm. If you tune the accel faster
or engrave at a lower speed you can use smaller overscan. If you are willing
to tolerate some uneven engraving power at the edges you can turn overscan
way down to get near the edge of your work area. I have not tested with
overscan of 0, that might break the algorithm.


(You can see my whole config at https://github.com/bjj/2x_laser ) I've not determined the max X acceleration, I just quit increasing it when it was pretty snappy. I'll probably go higher for a higher top speed. My Y accel is actually less than the machine can accomplish: I increased it until rapids were capable of knocking the open lid over center (!) and backed it off from there. My machine is on a table over carpet which makes it wobble more than I'd like.


thanks for posting this Ben. I have had some issues with the outer edges of my engraving area coming out slightly darker, just as is described here. Particularly when I'm working on anodized aluminum. Looks like I'll have to fiddle around with the acceleration config files to see if can improve this. thanks!