Another laser adventure

[dlgabry]

First, a thankyou to JJW for starting me on this path.

This is a 1.75 watt blue laser with an analog input for modulation, and I'm using PVC to produce the gcode. Rather than using an analog encoder driven by the z-axis, I wanted to develop an electronic means of producing an analog voltage to modulate the laser using the step and direction lines from Mach3. The modulator circuit consists of two 4 bit cmos binary up/down counters, an R-2R resistor ladder to produce the analog voltage, and an op-amp to provide enough current to drive the laser's analog input. The op-amp I chose only gives an output of 0 to 3.9 volts, but I'll be replacing it with a 'rail-to-rail' op amp that will come with in a few hundredth's of the supply voltage of 5 volts.

MAch3 is set up with a phantom 'C' axis, as I wanted to be able to use the X, Y, and Z axes to jog to the correct location and set the laser's focus. The make-believe C axis's step and direction lines drive the counters directly, with up to 256 possible shades.
The test photo of a tiger was done with a max depth of 0.035, on my machine that works out to 140 shades.

Dave

[JJWMACHINECO]

Great job Dave!

That engraving came out very, very good and the detail is amazing. It's a bummer that I can not open that schematic file to check it out.

Jeff

[dlgabry]

Hi Jeff,
Here's an image of the circuit diagram. I also have a version that includes preset switches to load the counter with a predetermined value (in hexadecimal) but haven't built that one.
Dave

[JJWMACHINECO]

Hi Dave,

Just a suggestion, would it be easier to make your Z axis move up and down to focus the laser with the phantom C axis? That way you still use the Z axis code generated by PVC to vary the analog voltage so you do not have to edit the G-code. Like I said before, I'm not a Mach expert and maybe I'm not fully understanding how you are able to get this setup to work using PVC and Mach together. Can you explain in more detail how it works with the g-code that PVC generates? Did you make a special PP for doing this that uses C instead of Z?

Thanks. Jeff

[dlgabry]

Hi Jeff,
Yes, I'm using a modified post processor. For PVC, I changed:
VAR Z_POSITION = [Z|C|Z|1.4]
VAR Z_HOME_POSITION = [ZH|A|Z|1.4]
to
VAR Z_POSITION = [Z|C|C|1.4]
VAR Z_HOME_POSITION = [ZH|A|C|1.4]

The third parameter is what gets put into the gcode file when the PVC is outputting what it thinks are the Z values. VCP has these and some additional ones that also need to change. Mach3 has no knowledge of the change, so its jog keys still move the X, Y, and Z steppers as they normally do.

In addition, I also have a mach3 profile like yours, with the numeric pad up and down arrows defined as the C axis jog keys.

To set the machine up to laser a picture, the procedure I'm using is:
- Run a test pattern that ramps the c axis down 256 steps while moving in Y, note the depth where a light burn occurs.
- MDI the C axis above zero by the same amount.
- Reset the counters
- Jog to the position the laser on the target, load the production g code and go.

Given the step size of 0.00025 on my machine, 256 steps from zero works out to a 'depth' of 0.06375. The counters wrap at that point, as with the analog encoder, so the max depth should be just a few steps less. With the present op-amp, the max voltage is about 3.9 and increasing the count beyond that doesn't raise the voltage. The new op-amp (the tlv272in the diagram) should fix that, assuming it works as described in the data sheet.

Some woods burn lots easier than others, so I MDI to c-0.06375 and set the potentiometer to the darkest burn voltage for the wood being used so that I can use the full 'depth' to maximize the number of shades.

Dave

[JJWMACHINECO]

Dave, all that is way over my head. I'm still using DOS CNC controlled machines in my shop and I get somewhat intimidated with all the extra things that Mach is capable of. I have allot of learning to do and you are definitely a good teacher.

My setup is up to 168 hours now with no failures, so until it does, I don't think I will be changing from the encoder until then. I have downloaded your schematic anyway for future reference.

I have another question out of curiosity. What diameter beam were you able to focus your lens down to and what step over are you using? Like I posted before, .007" is the smallest I'm getting with the 3 element glass lens. I also found .008" step over works the best where you can not see the burn lines and I get the most detail, but it takes longer to engrave. Is that approximately where yours is at?

Your setup is better in the aspect that your laser does not move up and down to be able to effect the focus at all. I have not noticed any issues concerning the focus, but theoretically it must be somewhat.

Thanks again. Jeff

[dlgabry]

Hi Jeff,
The ball bearing flavor of the encoder looks like it has a really long life, depending on how tight the the drive belt is.
I was running a dos mode stepper driver setup until a few months ago when I changed to Mach3. I'm not going back to dos in this lifetime And yes, there is a boatload of settings to play around with, most of which I have no idea what they do. It would be nice if Mach's default screens had a DRO and a zero button for the C axis, but I'm not up to messing around with trying to add one. I'm using the MDI, jog settings display and the voltmeter to set up the C axis anyway, so I can get by without a DRO for it.

I have a really hard time seeing the laser's focal point, even when it's not over the lasing threshold. I resorted to blowing some smoke into the beam when I was trying to find the focus distance. The beam's cone shape is pretty narrow with a focal distance of 4" or so, just a few degrees, so I'm guessing that a few thousandth's in movement up or down won't make much of a difference. I'm using a stepover of .010", and looking really closely at the picture, I can see the individual raster lines. From a couple of feet away, they're not visible.

Dave

[JJWMACHINECO]

Don't get me wrong Dave, I'm a firm believer in simple and stupid setups. Like using an encoder for Laser engraving. My DOS AH-HA CNC stepper controled machines are very reliable and will do very complex 3D machining operations with no problem. The hard drives in them are solid state Compact Flash cards. No moving, complex parts to fail. They have Win 98 loaded on them only for using a USB thumb drive to transfer large G-code files. I built 4 of them out of manual machines, 2 lathes and 2 mills about 10 years ago and still use them on a daily basis. Even the Hurco in my Avatar runs in DOS. Why?? Because it's allot more stable operating system then windows and not as memory hungry.

DDDD No, I didn't stutter (Don't Down DOS Dave) I only use Windows and Mach for engraving and CAD/CAM. Oh yea, and typing in this response. Just kidding with ya Dave.

Jeff

[dlgabry]

No Problem, Jeff! I like being able to jump from mach to vcp or pvc without going though a boot.
I have a question for you re your retro fit of the laser to a larger case. Did you find that the existing wiring was long enough, or did you have to re-wire it? I haven't opened mine up yet to see what I'm dealing with....
Dave

[JJWMACHINECO]

Hey Dave,

Yea, there are advantages to be able to jump from one program to another without reboot. I just have this phobia about a Window crash right in the middle of machining a very complex, expensive part. Only my hobby machines run in Windows XP and Mach which I have scrapped allot of inexpensive wood on. It took me a long time to work out all the hickups on my Servo CNC router, but I have not had any issues with the little one. I use KeyCreator for all my solids CAD/CAM work.

There should be enough wire in there to do what you need to. You seen the pictures I showed of the inside of mine when I first cracked it open. With all the extra stuff I added, there is a bigger rats nest of wires in there now. I had to cram all the wires to each side so the fan could blow directly on the heat sink.

When you get it all together, post some pictures of your machine with the laser on it.

Jeff

[Tweakie]

Nice work Dave, that Tiger certainly has more detail than your previous picture.

Can you reproduce an 8bit colour photograph similar to the ones Jeff has done ??

Tweakie.

[dlgabry]

Hi Tweakie,
The two counter chips give me 8 bit resolution or 256 shades with a "depth" of 0.06375". I say depth, but that's only in the gcode - there's no real stepper motor for this so there's no actual up or down movement. Output to the laser modulation input works out to about 0.02 volts per step.

The new op amps arrived today, should have one installed over the weekend. The laser's apart right now while I migrate it into a new enclosure with better cooling fans as it runs hot in the original enclosure.

The tiger actually looks lots better than the photograph taken with my cheap webcam. I'll be doing some more playing after I get it back together.

Dave

[dlgabry]

Migrated the laser into a new larger case with dual fans, and it's running much much cooler now, about 5 degrees F above the ambient temperature even at full power.

Dave

[jseiler]

I'm looking at doing something similar. What kind of laser goggles are you guys using? I'm having trouble finding information on how to spec out laser goggles. I know they need to match wavelength, but is there more to it than that?

-- John

[dlgabry]

Hi John,
The glasses I have are rated for a 445nm laser. The laser's a really bright blue in color, the glasses are red plastic, got them on ebay. They say it's not enough protection for looking into the laser beam itself though, but I can't imagine someone trying to do that.
Dave