The accent, tone of voice, plus lines like that one are exactly what I look for in this kind of youtuber. They always know what they're talking about, and the videos are always that one perfect style of video. Even getting some minor I Did A Thing vibes here, I love it!
You need a fan pointed at the work piece. The laser heats the air around the traces and that causes local refraction, so your laser accuracy is wobbly. If you keep a fan on, then that moves the hot air out of the way quickly, so you'll get more consistent lines.
I somewhat doubt that a few °C hotter air will have _such_ a drastically different refractive index that it bends "blue" light so much in such a short distance.
Between this kind of calm, deliberate narration of technical videos like these and clips of TNG where Picard admonishes people, I'm in a very happy place right now.
This is a really great presentation of laser PCB etching! I was looking at the Cetus3D because of the linear rails but had reasoned that it was of no practical use for my needs printing Ultem 9085, until now I see what can be done using it as a laser machine and only $299 plus parts that’s awesome! The download links, intelligent straight forward content that I can watch all the way through while enjoying the ever so subtle wit of the presenter is quite refreshing! Subscribed!
Hey man, I know its a bit late, but if you use a raspberry pi with octoprint on it, and connect a camera, it has an option that will tell the hotend to move out of the way after every later and it'll take a picture and store it
easily does 0.1mm features (65mA laser current, 405nm - 750mm feedspeed, careful with accel times) - ping me if u need some data (i burned a lot of bungard to get to right values ;)
Hey :) you can try whit a lower Acceleration, i believe it will fix the wave /shaking/ vibration problem you can also Test your 3D Printer vibration whit that laser, draw square in each corner, then you can see if it shake more in some of the corner ( no it was not mean as a shaking joke ;) "hmm??? what this tiny transistor for" Haha! i love all your small lovely detail in your videos : ) thanks
For a really good timelapse try attaching your camera to the build plate. I did it to my printer and you can see the part grow and it is really cool. I just attached some spare aluminum cut into a strip and screwed the aluminum to an unused hole using a nut and bolt. Make sure it is pretty stiff or you will get some wobble in the timelapse. Good luck!
A software dividing the plate in chess squares and 2 actuators moving some laser mirrors will increase the resolution a lot more, as there will be no vibrations in the frame. You could use multiple lasers as well for higher resolution - 9 smaller squares being burned at the same time, with 9 lasers, for increased productivity.
Wouldn't lowering the printhead lessen all the angular deviation you are getting? As in: "less amplitude of the wobbles". Also maybe try using the laser with a tiny masking dot right above the board?
Nice mod, but I wonder if there is a more direct way to do this... Given that the photo-resist is a layer that polymerizes when exposed to light, to protect the copper layer from the etchant, and you have a 3D printer which is designed to lay down layers of polymer anyway, perhaps there is a suitable polymer that could be directly printed in a single layer on a plain copper board, avoiding the necessity for the laser, photo-resist and developer?
good touches of humor, like! because of the nature of step motor(steps generates mechanical vibration), you will not be able to get ultra-precise lines unless you use high gear ratios on each axis with some renunciation of speed, As you know, you can use bldc motors driven by sinusoidal or FOC control (like odrive) to reduce vibrations, it may be more costly but prototyping machine are also high priced
If you manage to tune the jerk and acceleration and decrease them from the firmware side, you could stop that axis dependent vibrations propagated because the small scale. that adjustment just eliminate the resonances.
Hi Marco, thank you so much for this brilliant video! Could you try using the laser to make a soldering stencil in this set up. Also, you mentioned improvements that you made, could you go into more detail about that?
There are some nice timelapses to find on youtube that use the method you discribed. They look very statisfying. Way more than the ones with jumping print heads
Change the marlin curve divider in firmware and set jerk to a low number like 5 it will clean up the tiny round pads also belts don't use the little spring like belt tensioner it adds springy-ness to the belts make the g#-tight
Call any pring shop or silk printing shop and ask what their imagesetter provider is. They will happily print you a film with 4800dpi positive. Nothing can yet top that quality. 90's technology.
Do you really need all that extra mass on the moving head? Maybe the heat from the laser could be dissipated into the head itself. A more extreme weight saving technique is to only move a mirror, while the laser doesn't move, at least in the X direction. Laserprinters and professional PCB factories do it that way.
Couple of things to check. Looks like some of the wobbles come from change of direction, check the machines axis for backlash, huge huge quality killer is backlash.
Also, check the linear rails for adjustment, might need tweaking a bit tighter to minimise the rattle wobble effect from the balls rolling around. Looks great though!
I know this video's a bit old, but have you tried using Octoprint and Octolapse for your timelapses? That's what I use for my timelapses, and it works out alright.
i waana ask a very basic question, lets say i want to have a cnc to drill pcb holes, how can get a correct orientation/alignment in order to make sure that the drill will go to the exact location where the drilling points actually are, without deviation ?
I switched driver boards on my 3D printer to one of the 'quiet' ones - I presume this shapes the drive current or does micro-stepping. In any case, the vibration level went way down. I was wondering whether this would improve the quality of the traces on your setup.
Wouldn't the circuit get exposed with the surrounding UV? What chemical did you use for the etching? Can u pleaaaase give out more info on your our diy PCB process pleaaaaase I wanna make them with my own printer!
Hi Marco! I have a couple of questions to ask you, I'm building a multifunction 3D printer + laser/miller for PCB work, I found your video very clear and helped me already a lot with flatCam and other tricks, my setup is based on Marlin firmware, that doesn't (completely) support G0 rapid movement (G0 works but at the same federate set for G1, not rapid), how did you manage to move the laser faster in ''laser-off'' movements? other issue with Marlin is that M42 to control the laser output is not buffered, thus I need to put an M400 command every time a call the M42, are you using a different firmware or trick to sync the laser switch with the movements? thanks in advance and my best compliments for your works!
me every time I plan to do something "im going work on the heated bed and whats this random transistor for?" 12 hours later... "Ill just work on the heated bed another day" lmao
Excellent. I expect this can be done with other 3d printers. I have an anet a8 just in the mail. I hope its possible with it. Ill see once its assembled.
How did you make the onscreen text, "large" follow your movements? Did you do it manually or did you use some clever software? And if you used some clever software, did you write it yourself?
Would it be fine to use 405nm with 500mW that has TTL/PWM control? I should be able to bring down the wattage due to controlling the duty cycle, right?
An x & y motion could possibly be more accurate with leads that small. Who knows what the motion control is doing, following contours and the path. If you see in the video, there is some variance when he does a close -up. Burning it just like a laser printer would print, going with the axis in the motions would save the headache of figuring out any bounce vibrations, abrupt turns, and/or backlash. It's a laser-fine beam, not easy to calibrate the mechanics down to that precision.
Not sure what you mean by "who knows what the motion control is doing". It would be generating GCODE you can even check beforehand. With such a lightweight machine getting rid of backlash shouldn't be such a big deal - and it's something that should've already been taken care of for the 3d printing. Despite the rails I would rather be sceptical of the cantilever design.
By motion control - I was talking about the the g-code interpreter and step driving hardware. Backlash has nothing to do with the weight of a machine. Take a look at 5:09 on the video, and study the etched board. The pattern printed on the board, I am certain - does not match his layout. I have to assume the g-code export handle's arc's, and if not by now - Mr. Sweedish Accent Guy certainly knows how to process out his own g-code by this time. At 7:55, the board shows a great improvement, but still with many blemishes. There are actually crisp arc's and lines - and then some portions the pattern is distorted.
I had the same version from Eleks as yours only in the 2500mw version and I finally had enough as it just could not engrave all that well and I guess it was that electrolytic you showed is probably in it as well. I would have needed some way to be able to toggle the laser on and off much faster though it wasn't very good at greyscale either. I have an eBay Chinese laser 50w (really 40w) CO2 that will be here on Thursday thanks to that blue laser from Elek's being so crappy.
Very nice video sir.. Right now I am working on convert 3d printer to a pcb laser engraving machine.. This tutorial is very helpful to me.. After pcb engraving you have dipped the board in water right.. No chemical process.. May know the software for laser engraving pcbs
The exposed boards are developed in a sodium hydroxide solution as recommended by the manufacturer. The software that creates the machine movements from PCB gerber files is called FlatCAM and the software that sends it to the machine is called LaserWeb
Thanks for your reply sir.. I didn't understand that you have exposed the board in chemical solution before laser engraving right.. After the chemical process you put the copper clad pcb under the laser engraving machine..
The PCB is made with a photosensitive layer. That layer reacts to UV light. If I point the laser beam at it I can expose areas that I want to remove later. After exposing it like that I have to remove the now exposed photosensitive layer. That is usually done in a sodium hydroxide solution which will remove only the exposed material. Then I have a board with remaining unexposed photosensitive layer and bare copper areas which can be etched away in the next step.
@@reps Thanks! I have since found a PDF one of the guy on the Cetus3D group posted showing all the missing base resistors ! Appalling electronic design !
Does anyone know what the laser wavelength range necessary to to etch the photosensitive copper clad boards this way? I am confused by the fact this laser has a wavelength of 405nm, but near to far UV light has a wavelength range of 100-400nm.
Hi. Do you think only having 2 screws to fasten the surface of that whole module that carries the laser module is too weak? I saw construction of the cetus 3d - and I was curious why that bracket was so small.
Is a 500mw 405nm laser powerful enough to do this? As that is what if offered for printers from crwality like the ender 3. Which is a 24v system and am having trouble finding a solution for this. Help please...
The board is coated with a film that, when exposed to light, will wash away. You can buy boards with the film already applied. Search for "presensitized PCB".
I cannot understand why are you relating laser output with physical measurement. There are more factors to consider to achieve finer details, eg. microstepping, vibration, driver, optics, and the shape of the laser diode itself (which I assume is a rectangle for this kind of laser).
"warning labels are indicators of good tools"
I love this guy
Its kinda true though lol
The accent, tone of voice, plus lines like that one are exactly what I look for in this kind of youtuber. They always know what they're talking about, and the videos are always that one perfect style of video. Even getting some minor I Did A Thing vibes here, I love it!
You need a fan pointed at the work piece. The laser heats the air around the traces and that causes local refraction, so your laser accuracy is wobbly. If you keep a fan on, then that moves the hot air out of the way quickly, so you'll get more consistent lines.
I somewhat doubt that a few °C hotter air will have _such_ a drastically different refractive index that it bends "blue" light so much in such a short distance.
Not to mention it could damage the lens on the laser
Maybe is the plume created from ablation cause scattering? not sure just guessing
Leo Curious You say drastic, but we are only talking about sub .1mm vibration here.
@Blondie SL I know, how did we survive the stoneage 70's???
7:29 - Thank you for letting that sound ring out completely; that was very pleasant.
Between this kind of calm, deliberate narration of technical videos like these and clips of TNG where Picard admonishes people, I'm in a very happy place right now.
This is a really great presentation of laser PCB etching! I was looking at the Cetus3D because of the linear rails but had reasoned that it was of no practical use for my needs printing Ultem 9085, until now I see what can be done using it as a laser machine and only $299 plus parts that’s awesome! The download links, intelligent straight forward content that I can watch all the way through while enjoying the ever so subtle wit of the presenter is quite refreshing! Subscribed!
Hey man, I know its a bit late, but if you use a raspberry pi with octoprint on it, and connect a camera, it has an option that will tell the hotend to move out of the way after every later and it'll take a picture and store it
easily does 0.1mm features (65mA laser current, 405nm - 750mm feedspeed, careful with accel times) - ping me if u need some data (i burned a lot of bungard to get to right values ;)
maultrommel ping :P would you mind making your findings public? While i don't own this printer, i am sure others would appreciate it. :)
Hey :) you can try whit a lower Acceleration, i believe it will fix the wave /shaking/ vibration problem
you can also Test your 3D Printer vibration whit that laser, draw square in each corner, then you can see if it shake more in some of the corner ( no it was not mean as a shaking joke ;)
"hmm??? what this tiny transistor for" Haha! i love all your small lovely detail in your videos : ) thanks
Some of the squiggles will always remain due to resolution aliasing.
7:04 -- "It probably just uses the Google Translate feature." LOL! Looks like a great project. Thanks for the laugh!
For a really good timelapse try attaching your camera to the build plate. I did it to my printer and you can see the part grow and it is really cool. I just attached some spare aluminum cut into a strip and screwed the aluminum to an unused hole using a nut and bolt. Make sure it is pretty stiff or you will get some wobble in the timelapse. Good luck!
The hard part is not making small etched features - it is the drilling and vias that make a PCB useful for anything more than a blinking LED project.
A software dividing the plate in chess squares and 2 actuators moving some laser mirrors will increase the resolution a lot more, as there will be no vibrations in the frame. You could use multiple lasers as well for higher resolution - 9 smaller squares being burned at the same time, with 9 lasers, for increased productivity.
I really like your humor!
Wouldn't lowering the printhead lessen all the angular deviation you are getting? As in: "less amplitude of the wobbles".
Also maybe try using the laser with a tiny masking dot right above the board?
Nice mod, but I wonder if there is a more direct way to do this...
Given that the photo-resist is a layer that polymerizes when exposed to light, to protect the copper layer from the etchant, and you have a 3D printer which is designed to lay down layers of polymer anyway, perhaps there is a suitable polymer that could be directly printed in a single layer on a plain copper board, avoiding the necessity for the laser, photo-resist and developer?
good touches of humor, like!
because of the nature of step motor(steps generates mechanical vibration), you will not be able to get ultra-precise lines unless you use high gear ratios on each axis with some renunciation of speed,
As you know, you can use bldc motors driven by sinusoidal or FOC control (like odrive) to reduce vibrations, it may be more costly but prototyping machine are also high priced
Get yourself some JST XH connectors and a crimping tool then, if you plan on doing mod stuff for electric from China it worth the money :)
If you manage to tune the jerk and acceleration and decrease them from the firmware side, you could stop that axis dependent vibrations propagated because the small scale. that adjustment just eliminate the resonances.
Hi Marco, thank you so much for this brilliant video!
Could you try using the laser to make a soldering stencil in this set up.
Also, you mentioned improvements that you made, could you go into more detail about that?
I meant improvements for the other machine (EleksMaker), but yes, I'll definitely show them! Stencils, maybe by the end of the year or early 2018
and at 2:50 he just lets that laser drop and hang by the wire :D freaking comedy club, love it
Excellent results. Perhaps bringing the laser diode closer to the bed would negate some of the wiggles. At least on the print head axis.
I've seen a ton of your videos. finally subscribed! great videos!
Warning labels are a sign of good tools..
I'm going to use this line often :)
0:24 the pause and the "huh?!" won me over 🤣
Congratulations on the beautiful videos, I wanted to know if I can machine PCBs with a 20 watt diode laser
There are some nice timelapses to find on youtube that use the method you discribed. They look very statisfying. Way more than the ones with jumping print heads
Idea: Use the bed to "shake" and heat the board etchant etching etcher thing that etches..! :)
2300 mw/2,3 watt 450nm laser with TTL. I dont know if wavelength matter to cut somethink, but the Power does.
Looking forward for the next video on this topic. Great work ! Impressive workshop/equipment you have.
Change the marlin curve divider in firmware and set jerk to a low number like 5 it will clean up the tiny round pads also belts don't use the little spring like belt tensioner it adds springy-ness to the belts make the g#-tight
Call any pring shop or silk printing shop and ask what their imagesetter provider is. They will happily print you a film with 4800dpi positive. Nothing can yet top that quality. 90's technology.
Try placing the Cetus on the ground. My results where MUCH better compared to the printing done on a table!
Do you really need all that extra mass on the moving head? Maybe the heat from the laser could be dissipated into the head itself. A more extreme weight saving technique is to only move a mirror, while the laser doesn't move, at least in the X direction. Laserprinters and professional PCB factories do it that way.
Do you still have the test file you used in this? I want to run it on my new laser to see how it performs.
Really nice, man. Looks promising.
Couple of things to check. Looks like some of the wobbles come from change of direction, check the machines axis for backlash, huge huge quality killer is backlash.
Also, check the linear rails for adjustment, might need tweaking a bit tighter to minimise the rattle wobble effect from the balls rolling around. Looks great though!
I can’t wait to try this. How did you get the hot end bit to synchronise with the camera for the stop motion?
23bit absolute servo motors could vastly imporve the print. Also spring-loaded grounded ballscrews and of course pre-loaded B class linear guideways.
I know this video's a bit old, but have you tried using Octoprint and Octolapse for your timelapses? That's what I use for my timelapses, and it works out alright.
you should use optical stabilizer attached to a laser diode to get rid of vibrations.
Always a good watch! Thanks for uploading
i waana ask a very basic question, lets say i want to have a cnc to drill pcb holes, how can get a correct orientation/alignment in order to make sure that the drill will go to the exact location where the drilling points actually are, without deviation ?
I switched driver boards on my 3D printer to one of the 'quiet' ones - I presume this shapes the drive current or does micro-stepping. In any case, the vibration level went way down. I was wondering whether this would improve the quality of the traces on your setup.
Also maybe look into cheap LCD based SLA printer, the resolution is super high for something like this.
Wouldn't the circuit get exposed with the surrounding UV? What chemical did you use for the etching? Can u pleaaaase give out more info on your our diy PCB process pleaaaaase
I wanna make them with my own printer!
Hi Marco! I have a couple of questions to ask you, I'm building a multifunction 3D printer + laser/miller for PCB work, I found your video very clear and helped me already a lot with flatCam and other tricks, my setup is based on Marlin firmware, that doesn't (completely) support G0 rapid movement (G0 works but at the same federate set for G1, not rapid), how did you manage to move the laser faster in ''laser-off'' movements? other issue with Marlin is that M42 to control the laser output is not buffered, thus I need to put an M400 command every time a call the M42, are you using a different firmware or trick to sync the laser switch with the movements?
thanks in advance and my best compliments for your works!
Vibration by X- axis, Y-axis is perfect, so diagonal lines are middle quality.
You mentioned creating 3D printing time lapses with the based on nozzle position, Octoprint will actually do that for you!
Nice! What photoresist do you use?
me every time I plan to do something "im going work on the heated bed and whats this random transistor for?" 12 hours later... "Ill just work on the heated bed another day" lmao
Maybe the lines are wobbly because of the stepper motors microstepping? (I also like the fumes theory as well)
Excellent. I expect this can be done with other 3d printers. I have an anet a8 just in the mail. I hope its possible with it. Ill see once its assembled.
How did you make the onscreen text, "large" follow your movements? Did you do it manually or did you use some clever software? And if you used some clever software, did you write it yourself?
Where have you been bro we missed you...
Great video, is your Cetus printer MkII or MkIII?
How did you determine the "Tool dia" value in FlatCam for a Laser?
Would it be fine to use 405nm with 500mW that has TTL/PWM control? I should be able to bring down the wattage due to controlling the duty cycle, right?
Good morning Marco. Another great video, thank you sir. Quick question: On your bench, what is the device with the red, green, and yellow knobs?
Oltronix Power Supply B703 th-cam.com/video/Tr0FfZ-A3P4/w-d-xo.html
Thanks for the quick reply Marco and have a great weekend :-)
I need to see what you will use with that other LED driver board!
One of many LT demo boards lying around. I don't have an application yet, and because it is from 2005, it's not that great by todays standards.
you could change the resistor to 15 or 20 ohms and decrease the speed.
Why do you raster the PCB? Following the traces will likely give you even better results.
Torsten Curdt You can use the fact that Gerber files were originally for plotters, though newer extensions may assume a rasterizer.
An x & y motion could possibly be more accurate with leads that small. Who knows what the motion control is doing, following contours and the path. If you see in the video, there is some variance when he does a close -up.
Burning it just like a laser printer would print, going with the axis in the motions would save the headache of figuring out any bounce vibrations, abrupt turns, and/or backlash. It's a laser-fine beam, not easy to calibrate the mechanics down to that precision.
Not sure what you mean by "who knows what the motion control is doing". It would be generating GCODE you can even check beforehand. With such a lightweight machine getting rid of backlash shouldn't be such a big deal - and it's something that should've already been taken care of for the 3d printing. Despite the rails I would rather be sceptical of the cantilever design.
But in the end we are just speculating here - I would just give it a try and see if it makes a difference.
By motion control - I was talking about the the g-code interpreter and step driving hardware. Backlash has nothing to do with the weight of a machine. Take a look at 5:09 on the video, and study the etched board. The pattern printed on the board, I am certain - does not match his layout. I have to assume the g-code export handle's arc's, and if not by now - Mr. Sweedish Accent Guy certainly knows how to process out his own g-code by this time.
At 7:55, the board shows a great improvement, but still with many blemishes. There are actually crisp arc's and lines - and then some portions the pattern is distorted.
Is the second board, the board with the zebra, with etching? Looks good!
What kind of PCB are you using? Positive acting photosensitive ones?
I had the same version from Eleks as yours only in the 2500mw version and I finally had enough as it just could not engrave all that well and I guess it was that electrolytic you showed is probably in it as well. I would have needed some way to be able to toggle the laser on and off much faster though it wasn't very good at greyscale either. I have an eBay Chinese laser 50w (really 40w) CO2 that will be here on Thursday thanks to that blue laser from Elek's being so crappy.
How would I find out what all would be involved in upgrading the laser to something more high powered? 3000mW, for example?
Did you remove the inside PCB laser and connect diode directly to constant power source?
Very nice video sir.. Right now I am working on convert 3d printer to a pcb laser engraving machine.. This tutorial is very helpful to me.. After pcb engraving you have dipped the board in water right.. No chemical process.. May know the software for laser engraving pcbs
The exposed boards are developed in a sodium hydroxide solution as recommended by the manufacturer. The software that creates the machine movements from PCB gerber files is called FlatCAM and the software that sends it to the machine is called LaserWeb
3:25 very nice Raft as well :)
hello, what power does the laser have?
what type of pbc is ?? does he put any paint on the pbc ?, engraver directly on the copper?
Can't te laser be focused so you could place it closer to the board to reduce angular vibration?
I found that the lens on my laser jiggled a little bit during movements, causing wiggles like that. perhaps its not the frame vibrations after all
Another nice and interesting video, good work once again.
You aren't too bad yourself :) Hope to see the Fluke working soon
How is the PCB coated? I'd like to do the same, but no idea how to make these pcb sheets...
From a computer to a projector for light the pcb, or a diy laserprinter
Friend good night! wanted to know what paint you used to make a pcb
Synchronized capture is easily accomplished using the Octolapse plugin in Octoprint.
You got any details about your 3D printer ? Looks very nice. Can you tell me the make and model so I can look it up please?.
nope, sorry. got no details about that at all
Greetings from BlackSea! Good job!
know its been a while but mind posting up some example codes what power that socket on in Gcode for us cetus users?
Thanks for your reply sir.. I didn't understand that you have exposed the board in chemical solution before laser engraving right.. After the chemical process you put the copper clad pcb under the laser engraving machine..
The PCB is made with a photosensitive layer. That layer reacts to UV light. If I point the laser beam at it I can expose areas that I want to remove later. After exposing it like that I have to remove the now exposed photosensitive layer. That is usually done in a sodium hydroxide solution which will remove only the exposed material. Then I have a board with remaining unexposed photosensitive layer and bare copper areas which can be etched away in the next step.
What is the PCB type you used please?
Did you have any issue with the lack on base resistor on the NPN transistor on the Cetus main board you used to run the laser diode ?
I've only ran the laser for a single PCB and didn't see any problems in that short period
@@reps Thanks! I have since found a PDF one of the guy on the Cetus3D group posted showing all the missing base resistors ! Appalling electronic design !
Does anyone know what the laser wavelength range necessary to to etch the photosensitive copper clad boards this way?
I am confused by the fact this laser has a wavelength of 405nm, but near to far UV light has a wavelength range of 100-400nm.
Octoprint has a time lapse feature that does exactly what you are after.
I, personally, would appreciate an updated video on FlatCAM. 2015 is aeons ago in software land.
Edit:I guess so is 2017 :/
Maan. I love your humor! Keep it up! 😉
Repetier server pro has timelapse that you can sync with the layers
Bet the double sided tape woulda dampened it enough for nice straight lines 😃 love yer format awesome vid 👏
Why you didnt show after etching?
Why do you have a pic of micrometers on the front?
Super
Hi. Do you think only having 2 screws to fasten the surface of that whole module that carries the laser module is too weak? I saw construction of the cetus 3d - and I was curious why that bracket was so small.
Erich Stein No, those two screws are overkill, by several orders of magnitude. You could have used double sided tape.
Hallo Marco, hast du die Idee mit der Fräse schon umgesetzt?
Weiter so ich schaue deine Videos sehr sehr gerne.
What's the linear rails brand on the Cetus 3D?
Is a 500mw 405nm laser powerful enough to do this? As that is what if offered for printers from crwality like the ender 3. Which is a 24v system and am having trouble finding a solution for this. Help please...
Your accent is very pleasing compared to some other German speakers.
Can you plz explain how laser could engrave(wash out) the copper layer on pcb. You cover the surface of pcb with a special coat right?
The board is coated with a film that, when exposed to light, will wash away. You can buy boards with the film already applied. Search for "presensitized PCB".
Very nice, great tip to convert the gcode
haha I thought so too...
A 300mW laser does NOT dissipate 300mw of heat. More like 1-1.5W of heat. Looks like the heat sink is beefy enough regardless.
I cannot understand why are you relating laser output with physical measurement. There are more factors to consider to achieve finer details, eg. microstepping, vibration, driver, optics, and the shape of the laser diode itself (which I assume is a rectangle for this kind of laser).