Commercially available laser controllers

Electronics related to CNC

Re: Commercially available laser controllers

Postby lasersafe1 » Thu Jan 14, 2010 3:06 am

Still waiting, should be any day now. That control board is the one being used in the Triumph laser video carving the Aztec calendar. If I can get my laser to do that I think it will be worth every penny. I am also now awaiting the arrival of my craigslist Universal Laser cutter, so I'm going to be busy pretty soon. I thought I might take you up on the offer to do my own "build" in your new buildlog section of this forum to show the arrival condition, the breakdown and the rebuild of the new system. I'm still not sure what I'm going to do with the M40 engraver. I have a spare tube, so I shouldn't need to tear it down to build up the new system. I am short one HV supply, because the Universal Laser had an RF laser tube. The tube was returned to the factory for trade-in value by the previous owner.

Everything appears to be on the up & up with this craigslist deal, but I didn't want to brag about my awesome find until it actually arrives on the truck. Apparently there is a problem with things falling off of trucks...... ;)
lasersafe1
 
Posts: 599
Joined: Fri Nov 27, 2009 8:23 pm

Re: Commercially available laser controllers

Postby bdring » Thu Jan 14, 2010 1:14 pm

That sounds good. I added you to the group that can start new topics in the "Build Logs" forum.
Bart
"If you didn't build it, you will never own it."
bdring
Site Admin
 
Posts: 2966
Joined: Sun Nov 22, 2009 7:33 pm
Location: Chicago, IL, USA

Re: Commercially available laser controllers

Postby willyinaus » Mon Jan 18, 2010 8:44 am

I am frothing at the mouth over this looking forward to seeing it up and running.
willyinaus
 
Posts: 93
Joined: Fri Dec 04, 2009 9:08 pm

Re: Commercially available laser controllers

Postby lasersafe1 » Sun Jan 31, 2010 7:08 pm

The replacement controller has arrived and it has no communication issues. It is super easy to use, but I still have one small problem bugging me. I'm waiting to hear back from the vendor, who can speak Chinese to communicate the problem to the manufacturer in China.

I did a test raster of the Mayan Calendar on a piece of wood. First I told it to cut a square around the perimeter using a vector cut. No problem. Next I told it to raster the image. As you can see, the raster is squashed in the Y plane. It is also a little "muddy" because I didn't get the focus and power quite right. The circular image is now an oval in the final output. Something is wrong, and it is most likely something I have setup wrong. It has something to do with losing a step in Y because it cannot move less than one microstep. After the raster is completed I can see that it no longer returns the proper origin. Another hint that I have lost Y steps along the way.

The solution will probably be to go with higher precision microstepping.

I will take a video showing all the features after I work out this last bug.
Attachments
mayan.jpg
mayan.jpg (203.92 KiB) Viewed 19972 times
lasersafe1
 
Posts: 599
Joined: Fri Nov 27, 2009 8:23 pm

Re: Commercially available laser controllers

Postby lasersafe1 » Thu Feb 11, 2010 4:52 pm

I finally have the controller working perfectly. I have been super busy, so I haven't been able to produce a video yet. This thing is awesome!

After a few experiments it is changing my outlook on PWM. If I am wishing to do a vector cut, then by all means PWM is used to set the power level. The rapid on/off during a vector cut is done through the second control input to the power supply, the TTL that one might associate with an interlock.

This control method is also used during a raster. This changes how we have been thinking of grey scale. We can set the "power level" to the maximum burn depth we would want for a raster at a given speed. This power level setting sets the PWM frequency. The grey scale occurs because it is rapidly turning the laser on and off with the other TTL input.

Confused yet? This should clear things up.

First lets look at a typical grey scale image:
snap.jpeg




Next we will zoom in on the grey scale and we can clearly see the different levels of grey:
snap0.jpeg



Now we do a little trick so the laser can create a "pseudo-grey" image. Yes, it might look like crap in this photo, but the output on the laser blends it all together:
snap1.jpeg



Zooming in on this image we can see that the processed image is actually pure black and white with bit density creating the appearence of grey scale:
snap2.jpeg



This image processing is done with normal image processing software. The CorelDraw software can do it immediately before launching the laser plugin. PhotoShop and other packages can also do it. The conversion is called "Dither".

If starting from a color image, we first convert to greyscale. In photoshop, select IMAGE, MODE, GREYSCALE.
Once we have a greyscale we select IMAGE, MODE, BITMAP.
Now we have some choices to make.

1. The resolution will be important because it will select the ultimate "density" of the black and white pixels. We shouldn't select a pixel resolution that is higher than our machine can produce. This will require experimentation on your machine because it is a function of step size and focused laser spot size.

2. The method of diffusion will also make a difference depending on the type of image and the material being processed. In my example above I selected a "diffusion dither" method that creates a more randomized placement pattern of the black and white pixels.

This is going to need much more experimentation on my part to get it right, but my first run was pretty nice. I will show a picture later since I'm away from my machine.

Another VERY interesting surprise in the ultimate outcome was the fact that I can indeed acheive an anolog depth control into the wood. It can make true 3D depths varying from 0 to about .2 inches with nice rounded transitions. I wasn't expecting this. I thought that this could only be done with regular 3D CNC milling. HOW, WHY? I don't know for sure but it must be some slight overlaps in the burned holes allow the edges to be knocked down in the hole next door. Lame explaination, but it just works.

When you first finish the burn it looks ....well... burnt. After cleaning the burn with the Orange hand cleaner and toothbrush for a minute, the black burns disappear and the original wood shows through with different cut depths. Perhaps this is the part that helps knock down the hole edges to blend it all together. This is way too cool! :D
lasersafe1
 
Posts: 599
Joined: Fri Nov 27, 2009 8:23 pm

Re: Commercially available laser controllers

Postby lasersafe1 » Thu Feb 11, 2010 5:25 pm

Now can anyone out there spot my error in the image processing?

Let's look closely at this:
snap2.jpeg


Yes it is obvious that we will burn deeper as the shadows progress from white to dark, but look at the area around the star... It would have no burn, so it would jump back out at us.

If we want a true 3D of this item into the wood we would want to put the whole image on a black background first so the background is also deep. After the correction it would look like this:
after.jpg


Think.. Darker=Deeper

One thing should be obvious. An image in the wood can never look better than the original. If the shading looks wrong in the original it will carry through to the burn. The "Dodge" and "Burn" tools in the image processing software are your friends. Repeat after me... The "Dodge" and "Burn" tools are my friends. You will use the Dodge tool to make the laser dodge the area, which results in a lift toward the surface. Conversely you will use the Burn tool to burn deeper. Isn't it interesting that the old photography darkroom nomenclature fits so perfectly into the laser engraving process? ;)

Perhaps laser "engraving" is the wrong term to use for this. When going for 3D I think it should be called laser "machining".
lasersafe1
 
Posts: 599
Joined: Fri Nov 27, 2009 8:23 pm

Re: Commercially available laser controllers

Postby bdring » Thu Feb 11, 2010 8:00 pm

You need to post a picture of an engraved piece.

I understand how this works. From a mathematics/theoretical point of view, both an "analog PWM" method and a "bit bang dither" method should yield similar results. When dealing with real world laser tubes and power supplies, though, one method is obviously going to win out. I am interested to see how this plays out.

From a controller point of view the "bit bang dither" method is way, way simpler to implement. You never have to do on the fly power calculations due to pixel color.

I have another question. Does the laser head travel past the sides of the engraving to compensate for acceleration or does the head stop exactly at the edges?

Note: My Vectric Aspire program creates perfect grayscale "height maps" of any 3D object.
Bart
"If you didn't build it, you will never own it."
bdring
Site Admin
 
Posts: 2966
Joined: Sun Nov 22, 2009 7:33 pm
Location: Chicago, IL, USA

Re: Commercially available laser controllers

Postby pixpop » Thu Feb 11, 2010 9:50 pm

How are these two different? (I mean 'bit-bang dither' and 'analog PWM').

I have a Synrad laser, which has only a PWM input. Would that be bit-bang dither, or analog PWM? This laser is different from the Chinese HV supplies, in that the PWM input is digital, whereas I think that with the Chinese supplies it is analog. The analog input also accepts a digital signal, which can be PWM'd if necessary. I guess this means that it's basically an analog control, with a low pass filter to deal with the PWM if it's used as a digital input.

I was planning to implement an engraving controller that would synchronize the PWM pulses to the step pulses (or encoder pulses for a servo) of the raster axis. The pulse length would encode the intensity at that pixel. So, if there were 1000 step pulses per inch, and the PWM frequency was 5 kHz, you would raster at 5 inches per second. Maybe that's too fast to produce an image though, I don't know.

Neil
pixpop
 
Posts: 24
Joined: Fri Dec 18, 2009 7:47 pm

Re: Commercially available laser controllers

Postby bdring » Thu Feb 11, 2010 10:30 pm

Those are my own invented terms.

On the Chinese CO2 tube power supplies, the power is controlled via an analog voltage or a high speed PWM on the same pin, so I assume there is just a low pass filter that is turning the PWM into an analog voltage. So, while PWM is digital, I am calling it analog power control.

Bit Bang Dither - Describes using full on/full off on a 2 color dithered image. The PWM just controls the actual power of full on.
Bart
"If you didn't build it, you will never own it."
bdring
Site Admin
 
Posts: 2966
Joined: Sun Nov 22, 2009 7:33 pm
Location: Chicago, IL, USA

Re: Commercially available laser controllers

Postby lasersafe1 » Fri Feb 12, 2010 12:00 am

Perhaps I'm all wrong. Perhaps it is all going in the PWM and the TTL on my other pin is simply toggling as a backup. This was mentioned in the manual, so I hooked it up. There is a different hookup requirement depending on if you are using a tube or an RF laser. They have the tickle pulse built-in. I guess one way I can test it for sure would be to disconnect the TTL portion and see if the output is the same.

In my mind I simply couldn't imagine that the PWM could alter the power so quickly since the circuit has the low pass filter. I recalled seeing in one of the power supply manuals a turn-off time in milliseconds. This seems way too slow for the speed of the raster I was performing. Maybe it is only my mind working too slow. The controller and the laser are both quite fast. I'm going to try to squeeze in the video tonight with some still pix of the final engraving.

To answer the other questions: Yes, it looks like it overshoots the image a little to give the stage time to accelerate and decelerate on either side.
lasersafe1
 
Posts: 599
Joined: Fri Nov 27, 2009 8:23 pm

PreviousNext

Return to Electronics

Who is online

Users browsing this forum: No registered users and 5 guests

cron