Ben's 2.x Laser Build

Post your build logs here

Re: Ben's 2.x Laser Build

Postby jv4779 » Fri Aug 31, 2012 3:15 pm

I implemented my raster engraving as distinct dots. Really only because of speculation that low DPI engraving would look smeared by the much more bold horizontal lines vs the vertical spacing. At .005in laser spot size you start to get overlapping dots at greater than 200 DPI, so that was the DPI I got my best results with. The code uses the same type of logic as PPI where a dot is just firing the beam for 3ms.

It would be interesting to see the difference between rasters done using distinct dots vs. connecting adjacent pixels.
jv4779
 
Posts: 40
Joined: Fri Dec 09, 2011 10:54 pm

Re: Ben's 2.x Laser Build

Postby educa » Fri Aug 31, 2012 3:28 pm

Rastering with dots makes -> dots

Rastering the other way makes stripes.


I can personally also not imagine that stripes would look better then dots.

On the other hand, I have the impression that your ON time of 3ms per dot is quite high. That means you can at max shoot around 300 dots per second.

If you shoot at around 200 DPI, then 300 dots means a raster engraving speed of max 1.5 Inches per second. Thats really slow.

My goal is to create the controler on dedicated hardware. I know I could do it with emc2, but honestly I tried about 20 different pc combinations and almost no combination was able to give me a decent latency.
And even the PC's that had quite low latency tended to give errors now and then.

So the primary choice was the arduino platform.
This would MAYBE be possible, but you would have very little cpuspace left on a 16MHz 8 bits controller.

On the other hand, since a few months there is the very nice 32 bits MicroChip PIC32 and this chip is certainly capable.



In other posts where dirktheeng discusses PPI, There is somewhere a graph which shows the intensity of a laser pulse over time.
In that graph you can see that the pulse is allready at maximum after only 0.3 to 0.5 ms. Dirk choose to have 3ms pulses for cutting purposes, but I guess you should be able to burn dots at 0.3 to 0.5 ms and that would speedup a lot.

Time will tell... its a lot of software writing, but I have to, because not a single emc configuration was able to provide me with good results.
Cutting with PPI was not a problem, but the rastering isn't what I expected. Its even strange, but when I run a cutting job with 2 or 3 small raster subjobs in it, then mostly the first raster engraving works, but stuff after that fails at random places. I don't know why, but its certainly not helping me out.
educa
 
Posts: 239
Joined: Thu Dec 22, 2011 9:13 pm

Re: Ben's 2.x Laser Build

Postby kbob » Fri Aug 31, 2012 3:48 pm

In other posts where dirktheeng discusses PPI, There is somewhere a graph which shows the intensity of a laser pulse over time.
In that graph you can see that the pulse is allready at maximum after only 0.3 to 0.5 ms. Dirk choose to have 3ms pulses for cutting purposes, but I guess you should be able to burn dots at 0.3 to 0.5 ms and that would speedup a lot.


Is this the post you mean?

http://www.buildlog.net/blog/2011/12/ge ... er-system/

There's a lot of chewy information in that article, and I am still masticating it.
Bob
"If you didn't code it, it will never own you." (-:
kbob
 
Posts: 151
Joined: Mon Jul 09, 2012 6:39 am
Location: Eugene, OR, US

Re: Ben's 2.x Laser Build

Postby educa » Fri Aug 31, 2012 4:15 pm

It is ALSO that article, but I have seen that graph somewhere in a bigger resolution too.

As you can read there by dirk, a 3ms was a nice cutting timing. 2 ms did cut paper very thin and 1ms didn't cut through (but it did cut)

The stuff dirk made was/is working, but as far as I know he uses mach and only created some hardware to do the PPI pulsing.

My purpose it to not need a pc at all.

I'll write PC software which is able to layout gcode and raster engravings on a workpiece, but the effective work is then saved onto a SDHC card. You walk with the card to the lasercutter, put it in, set your zero-point and press START.

Thats all.

To get it speedy enough I will hardcode some stuff in my software. For example 1 step of a stepper in X or Y is 0.02mm, so I have 1270 steps per inch on my machine.

In cutting mode my speed aim is around 27 inches per second max speed and in raster engraving that'll go to 39 inches per second.

PPI setting is dynamic (very heavily based on the system Dirk used) and pulse length will be possible to set from 0ms to +- 30ms in 0.001ms increments

Laser power will also be settable in gcode of course (pwm channel) and I hope that the system is fast enough to try to raster 8-bits greyscale. Most of the buildlog projects only do 1 bit (black or white) engraving. The only thing you need to do greyscale is SPEED. But... I'm not sure about that one since I believe that is one of the key differences between our cheap (yeah, they tend to call them cheap) co2 tubes with the pro rf lasers.

It is a challenge :)

I did it allready with a 0.7Watt 445nm Bluray diode laser at much lower speed. http://www.youtube.com/watch?v=xfhDgUTKdb4 (smurfette at the end of that movie is in greyscale)
educa
 
Posts: 239
Joined: Thu Dec 22, 2011 9:13 pm

Re: Ben's 2.x Laser Build

Postby jv4779 » Fri Aug 31, 2012 4:44 pm

educa wrote:On the other hand, I have the impression that your ON time of 3ms per dot is quite high. That means you can at max shoot around 300 dots per second.

If you shoot at around 200 DPI, then 300 dots means a raster engraving speed of max 1.5 Inches per second. Thats really slow.


You are correct, I rechecked my raster settings and the on-time for each dot is .5ms. It is configurable per raster and could depend on the DPI and speed being used to give best results. .5ms on time, 600in/min, 200 dpi. The machine could go faster but I never pushed it. I then verified the dot placement under a microscope and it was all dead on.

If the raster speed was fast enough and the DPI was high enough that consecutive dots were running together faster than .5ms you would end up with a stripe for them, ending the stripe .5ms after the start of the last dot.
jv4779
 
Posts: 40
Joined: Fri Dec 09, 2011 10:54 pm

Re: Ben's 2.x Laser Build

Postby educa » Fri Aug 31, 2012 4:49 pm

JV4779 do you have any idea how long the laser is then OFF between your 0.5ms pulses ?
educa
 
Posts: 239
Joined: Thu Dec 22, 2011 9:13 pm

Re: Ben's 2.x Laser Build

Postby jv4779 » Fri Aug 31, 2012 4:52 pm

educa wrote:JV4779 do you have any idea how long the laser is then OFF between your 0.5ms pulses ?


Unless I messed up the math, 200 DPI is .005in per dot and at 600in/min = 50ms per dot, so 49.5ms of off time.
jv4779
 
Posts: 40
Joined: Fri Dec 09, 2011 10:54 pm

Re: Ben's 2.x Laser Build

Postby educa » Fri Aug 31, 2012 4:58 pm

600in/min = 10inch/sec

200DPI @ 10inch/sec = 2000 dots per second

2000 x 0.5ms on time leaves NO time for off time ???
educa
 
Posts: 239
Joined: Thu Dec 22, 2011 9:13 pm

Re: Ben's 2.x Laser Build

Postby MattyZee » Thu Oct 18, 2012 10:31 am

Hey Ben,

I'm getting a error when trying to Raster Engrave. I'm not sure i'm running it correct. The error says
Code: Select all
Near line 5 of /home/laser/raster-test.ngc:
File ended with no percent sign or program end


My "raster-test.ngc" file contains only 4 lines of code
Code: Select all
M3 S1
M68 E0 Q0.2
F14000
O145 call [123] [100] [100] [100] [100] [0.04] [0.04] [15]


It doesn't seem to matter if i have an image file in the [RASTER]IMAGE_PATH or not. I just get that error.
I've customised your EMC2 configuration to match my machine (pinout changes and different base period/speeds etc) but everything else is the same.

I'm am opening the "raster-test.ngc" using file>open on the linuxcnc window.

Any ideas what i could have done wrong?
MattyZee
 
Posts: 15
Joined: Thu Mar 03, 2011 8:29 am

Re: Ben's 2.x Laser Build

Postby educa » Thu Oct 18, 2012 10:45 am

Have a read about the conventions of .nc files please.

It's not that I don't want to explain it to you, but I had this problem too but can't remember the 100% solution.

Somehow you have to put 1 last line of gcode in your file to tell linuxcnc that the job is over

This is a percent sign or something else, but it is documented so you should be able to find it.

Regards,

Bart
educa
 
Posts: 239
Joined: Thu Dec 22, 2011 9:13 pm

PreviousNext

Return to Build Logs

Who is online

Users browsing this forum: Google [Bot] and 1 guest

cron