วีดีโอ

The free image viewer: www.faststone.org/

That's a very interesting idea... My small circuit is primarily aimed at really small projects, which should also find space in the smallest possible housing. Because I am also not a professional in programming, this circuit was obvious and easy to realize for me. But of course the MCP23S17 opens up much greater potential. I would be happy if you would also share your project with me and other users.

3mm

@@stromlinge2669 so 2.5W laser can cut 3mm acrylic & plywood

@@user-vm1xj2bb8p black acrylic is tested with 3mm. but playwood has not been tested

@@user-vm1xj2bb8p Yes, i tried 3mm Black And Red Acrylic, and it cut both, Red took 2 more passes thou.. but haven't tried wood

I don't see any problem with the Ender 3. However, I would protect the printing surface with something fireproof, because in my experience the lasers get stronger quickly :-). I also did it beforehand with my 3D printer and used a wafer-thin sheet of copper to protect the printing surface. The advantage is that copper oxidizes quickly and this coating hardly causes unwanted reflections from the laser (also suitable for focusing).

@@stromlinge2669 Did you get good results from the 3D printer as compared to your home built CNC? Which one will provide better laser trace accuracy, Ender 3 or home built CNC?

I only know the Ender 3 from pictures. But it should be stable enough. If the results of the 3d printing are good, the laser should work too. Otherwise I was happy with the result with my 3D printer, but had to reduce the speed a bit. At that time, however, I was unable to configure the mechanical acceleration with my 3d printer. But with grbl it works very well. So you can reduce the mechanical forces a little and work more precisely. Nevertheless, for PCB I also have to work with my small cnc laser in unidirectional mode, because otherwise the image will not be completely clean (saw teeth on the edges). This can have mechanical causes or it can be due to latencies in the control electronics.

​@@stromlinge2669 Is it possible to control the Ender 3 CNC bed? I specifically changed the original motherboard of the Ender 3 to a much better upgraded motherboard which uses the Trinamic's TMC2225 stepper drivers. I was hoping it should be able to provide precise mechanical acceleration and be very silent to prevent any jitters or shakes produced by the stepper motors which would render to the photoresist. The firmware uses Marlin 2.0.1 - How would I be able to use this 3D printer mother board to accept laser software for PCB sensitizing?

@@ShopperPlug It's not that easy. That's why i ultimately decided to build a simple device myself. In general, Marlin also uses G-Codes, just like LaserGRBL. The Marlin support in LaserGRBL is available, but still experimental. The control line of the laser is usually connected to a PWM fan connection or to another PWM-compatible PIN on the MCU. However, you then have to make sure that the signal levels match each other (5V). The actual work is then the configuration. I managed to cut plastic, but then it seemed to me that the effort was too great to try everything out and reconfigure it again and again. Hence the solution with the GRBL Board and LaserGRBL software. For Marlin you can find information on the homepage: marlinfw.org/docs/configuration/laser_spindle.html Unfortunately, I can't help you much more .... But the thing with the Hexastorm is just too interesting :-) .... When I find the time, I'll read something there too ....

I've never tried pvc. So far I have been able to test with 3 plastics: Plexiglas (acrylic) black: works fine ;-) Plexiglas white & transparent: nothing ... no cutting and engraving POM black: just like white plexiglass ... nothing ... Everything tested with this 2.5 watt laser ...

@@stromlinge2669 if you do try PVC do it outside!!! Extremely toxic fumes! Chlorine gas kills :D

Hello Max Dax, unfortunately I don't know this laser. I am not using this Sony 50mW laser diode for no reason. On the one hand, it is exactly the right wavelength. On the other hand, it is also the specialty of lasers with low power. The laser crystal has a very small exit area. This is why you can focus the laser beam on an extremely small point, which then results in the relatively high resolution. The control of the laser is also crucial. My laser driver is pure analog technology, controlled by the GRBL board. That's why he immediately reacts as he should.

@@stromlinge2669 I have sony blu ray laser diode, 3.8 and 5.6 package, they are about 400-500mw. Few days ago I also bought your Sony 50mw diode on eBay. What is your laser beam spot size? What kind of lens specification do you use? is it a 3 element glass lens? how can I make the CNC bed? Interesting you are using analog technology.

@@ShopperPlug The laser driver is just a constant current source (implemented with an LM358 OPA and PowerMOSFET), which can be controlled by the Arduino via PWM. This is a very reliable and proven analog technology. This supplies the necessary 50mA current for the small laser diode. I've tried different lenses. But in terms of the size of the spot, I achieved the best results with the standard 3-element glass lens, which is also sold with many lasers. This should also have the right coating for 405 - 450nm. The distance from the PCB to the lens is about 7 to 8 cm. I cannot accurately measure the point size. However, with a resolution of 10 lines / mm, the area was not completely covered. The stripes only disappeared at 16 lines / mm. Accordingly, the point should be a little smaller than 0.1mm. The use of laser safety goggles is absolutely important for focusing. This is the only way to see the focus point exactly. The green glasses are best for these low-energy lasers.

@@stromlinge2669 There are many different types of 3-element lens. Can you please send me the link of the one you are using? I know that 3-element lens might have a beam spot of around 0.07mm (70 microns). I really do not understand what you mean by stripes of 10 lines/mm. Are you referring it to filling the traces by lines? The resolution of dots or lines is 16 for every mm? That is a good point, I do have green goggles and will try it out. I saw someone on the internet who was able to accurately determine the laser's beam spot size using a webcam's CCD sensor I believe, an interesting topic. With the results you produced I would never thought such basic components can be used to build a laser driver. I have some Ichaus iC-HG laser driver chips, they can pulse at 200MHz... I planned on building a high resolution polygon laser scanner using prism rather than a mirror (opensource project called "hexastorm"), but your build seems much easier and cheaper to get some amazing PCB trace results.

@@ShopperPlug this are the lens i used: www.ebay.de/itm/313498675744?hash=item48fdfa2a20:g:mc4AAOSwJK5gd-al The iC-HG chips look good. If you have these, you would recommend using them. My claim was different. I wanted to develop everything myself and deliberately not use high tech. So go back to simple and proven technology. So I was able to use my optimization potential with my simple means. When I exposed areas with 10 lines / mm, there were still thin lines of the photoresist left. Only at 16 lines / mm did everything disappear. That would support the theory that the point is smaller than 0.1mm. But as I said, I would advise against using this technology for PCBs. You can do that if you have such a device anyway (I mainly use mine to laser cut acrylic front panels). With 16 lines / mm in unidirectional mode, it just takes too long to expose a larger PCB. For small boards <5x5 cm it is quite ok. The Hexastorm would indeed be better suited for this and would be worth considering for me as well. If I were to build more circuit boards than a single piece myself, that would be my first choice.

Hello Max Dax, I developed this PCB with KiCAD and then ordered it from JLCPCB. I have already tried to apply a solder mask myself for individual projects, but unfortunately my results were not satisfactory. That's why I decided to make my PCB as universal as possible so that I can use it for many things. Then it also makes sense to have these made by a manufacturer.

Hello Max Dax, thank you very much for your interest. This project is completely self-made in almost all respects. For the replica you need some tools, aluminum profiles (20x20mm and 30x30mm) with slot nuts with spring and brackets for mounting the 30 mm profile at right angles. Unfortunately, there are no proper instructions on how to rebuild it, because it is typical for such projects that some experiments and modifications are necessary. Maybe it will help you if you can print out the parts yourself with a 3D printer. You can find everything you need on my account at www.thingiverse.com/thing:4873062. However, I have to admit that it takes a long time to expose circuit boards with the laser. With a resolution of 16 lines / mm, it only makes sense for very small boards (max 5x5 cm). In the meantime, I prefer to manufacture large circuit boards in the traditional way (print on foil with a laser printer and then expose conventionally).

:-) Unfortunately I don't have any instructions. Please have a look here: www.thingiverse.com/thing:4873062

Thank you for your interest. Unfortunately, I have no good news. At the moment I have no more boards left. I have already exchanged the ones that I have built with my friends :-)

For the ATmega644 the MightyCore is needed. github.com/Optiboot/optiboot