Commercially available laser controllers

Electronics related to CNC

Re: Commercially available laser controllers

Postby lasersafe1 » Mon Sep 20, 2010 12:34 pm

I'm sure the wavy lines were not due to mechanical errors. My stepover was .005mm. It did show a funny behavior during the raster. It would scan back and fourth about 4 times at normal speed and then every fifth time it would do a quick return. This might have something to do with the wavy line. Perhaps the .005mm stepover is actually smaller than the bit resolution (I haven't calculated it). Perhaps it would repeat the .005 several times before incrementing the Y for the next known row of dots. I think this is where FSE has the biggest advantage. He knows exactly what is required for an image because he has designed both the software and hardware. Where you and I struggle to determine an appropriate Y stepover, his software has already optimized it and will run it every time.

For as much complaining as we do, it's funny that our lasers still behave as good as lasers that cost upwards of $12000. :mrgreen: My heart sinks for the poor fella's that I see on Craigslist trying to sell their engraving business for $15000 because they can't make it in todays economy. I feel that my laser is ready to do some business if I can think of something cool to make. I wish someone would help me out with the Blender software question, because I think that could be a fine program for producing 3D carving heightmaps. (Note: it might also be useful for makerbots if they can take heightmaps... actually, I think Blender can export as STL too.)

http://www.buildlog.net/forum/viewtopic.php?f=5&t=288
lasersafe1
 
Posts: 599
Joined: Fri Nov 27, 2009 8:23 pm

Re: Commercially available laser controllers

Postby lasersafe1 » Mon Sep 20, 2010 12:50 pm

FSE, If Blender can export an STL, and STL information is stored in planes from the bottom up, then would this be of any advantage in processing a laser file? I guess it really couldn't be better than a heightmap in 256 gray.
lasersafe1
 
Posts: 599
Joined: Fri Nov 27, 2009 8:23 pm

Re: Commercially available laser controllers

Postby fullspeceng » Tue Sep 21, 2010 12:33 am

Tweakie is pretty demanding as a non-paying customer ;)

The Mach3 engravings he has done are very impressive and much better than we expected possible with Mach3. Congrats at finding those settings.

We will post more pics when we get a better macro camera. In the meantime, please feel free to order a cut sample from one of our partners: http://www.fullspectrumengineering.com/ ... hp?f=3&t=6

There are a lot of issues involved in getting a good engraving but we've already said our controller is bit perfect.

There is nothing better than perfect :)

Maybe your laser power supply can turn on/off faster or your beam spot size is smaller with a short focus lens but those are unrelated to the controller.

When the 1-bit bitmap image is black, our laser fires exactly at the right time for the right duration and that's pretty much all you can hope for in a controller.

We can do it perfectly at 1000dpi at 2000ipm or at 4000dpi at 6000ipm. It's all switching time limited by then. Our controller runs at 400MHz.

There also seems to be some confusion about image processing.

Perhaps we should revisit what are digital images. Images have pixels. If you want perfect representation you need to have your system EXACTLY matched to the resolution of the image.

How you went from 1000 dpi -> 600 dpi -> whatever the step resolution on your laser is a mystery.

1000dpi->2000dpi or 3000dpi or 4000 dpi is ok. Integer multiples will just repeat the data of the corresponding pixel.

But if your 1-bit image says point (100,100) = 0 at 1000dpi and you scale that image to 600dpi, what will that pixel say now? When you output that image on your laser, does it fire or not fire on that spot?

For optimal output you must understand the basic principals involved in outputting the image.

Outputting pretty images is fun but true science requires a more careful analysis. We have specific test patterns that are generated using math software that we use to test specific conditions.

One of the best tests is analysis on a high speed digital logical analyzer then reconstruction of the image from the captured data. This takes other factors out of the equation.

Finally, let us remind you Lightobject doesn't manufacture anything. They just import some controller from China and resell it. John is a really clever guy and we are very impressed he was able to get that controller working but you can see from the beginning of this thread it was a several week if not month effort.

When Bart needed to get his system working, we were able to recompile the firmware and adjust the step pulse time from 1us to 4us to accommodate his Gecko drivers. Try getting something like that done with any other manufacturer :)

We are not saying we will attend to all random requests but if it affects many users then we will do our best.
fullspeceng
 
Posts: 38
Joined: Thu Dec 17, 2009 10:13 pm

Re: Commercially available laser controllers

Postby lasersafe1 » Tue Sep 21, 2010 1:53 am

Take it easy. Tweakie is simply playing devils advocate and trying to keep everyone honest. If you can scan your engraving with a flatbed scanner, I think it would satisfy his curiosity. I think the scanner can show better resolution than a camera with good macro setting. If you still want to ship me your engraving, I can do my engraving on the back side and then scan both for a comparison.
lasersafe1
 
Posts: 599
Joined: Fri Nov 27, 2009 8:23 pm

Re: Commercially available laser controllers

Postby macona » Tue Sep 21, 2010 4:35 am

I have a Zeiss stereomicroscope that I can take reasonable microphotos with:


Image
SITe ccd by macona, on Flickr
macona
 
Posts: 363
Joined: Tue Aug 17, 2010 10:30 pm

Re: Commercially available laser controllers

Postby Tweakie » Tue Sep 21, 2010 7:05 am

Fullspec,

Thank you for the explanation regarding DPI scaling - to be totally honest this is something I had not even considered (but I will now).
I plan on obtaining more knowledge in this, and other areas when I start using my RF tube. I will be a lot more comfortable measuring the control signals and timing pulses when there is not an 18,000 volt transformer within 70mm of my scope probes. :D

Thank you for being up front and helpful.

Tweakie.
Tweakie
 
Posts: 120
Joined: Tue Apr 27, 2010 5:17 pm
Location: UK.

Previous

Return to Electronics

Who is online

Users browsing this forum: No registered users and 10 guests

cron