because he has only 1 video

It is very true, although I notice that perhaps since he is not a person of young ages and young people consume more internet, this means that this valuable content is seen by few people because the project is quite innovative.

380

@@HakkiOgrt 603

Correct that , .... it's 605 subs ?

Check out marcoreps instead.

Not just any fiber though. Losses through conventional 62um or even 1.5um fiber would be too high and coupling light into those requires either precision optics or lasers. For UV, you need to use hollow core fiber that starts at 200um and about $40/ft, which is only about twice the resolution of this pinhole implementation. If you want to focus down to less than 200um, more optics will be needed at the other end too, at which point you may as well go with 400um hollow core to more easily pump more UV in.

Theoretically: yes Practical: no The spot size is highly dependent on the lens-object distance. The diaphragm of the lens greatly reduces the amount of light. It requires an expensive lens system. And a greater distance from the light source results in wavy lines when the machine vibrates.

​@@wilstraver962 Do you have the 3d specs for your apparatus? Do you sell the jigg?

I agree, many high quality sony UV laser diodes are cheap and simply use glass lens. Should be easy to build and provides a fine laser focus point. I have them sitting around, brand new many kinds of sony UV laser diodes and ender 3 V2. I never bothered to build the ender 3 v2 because I didn't expect it to provide decent PCB pitch lines. This youtube video shows it is very promising.

​@@wilstraver962 almost the laser marking or cutting machine work in same principles ..

Correct, the milling process vibrates alot and makes a really bad mess. I may convert the milling machine into laser/uv diode.

only upside of milling is that you get holes with your pcb but otherwise, too much pain in the ass.

makes me wonder if you could shove a bit of scrap fiber optic cable to get more resolution. I get why you don't want to use a lens in this setup, but maybe the you can push enough light in the fiber for it.

​@@warlockd I think you meant to write a message under the video and not under my comment, but I'll try to figure it out anyway. In the article "Transmission of UV Light with Optical Fiber" on the lasercomponents website there is a statement that a regular optical cable will not work, I quote: "In wavelengths shorter than 260 nm, UV light caused degradation in conventional fibers - a direct result of solarization effects." There is also a chart where you can see degradation of regular optical fiber cables. So basically you will need special optical fiber cable by Polymicro or other vendor in order to work for a long time or if you would use regular one then change it after half an hour of usage.

@warlockd this actually a brilliant idea. The only constraint would be how much overlap you could get when creating a fill type shape and of course the resolution of each step.... keen to see what that looks like numbers wise

hotend-nozzle is reflective and you will loose occuracy

That sounds excellent!!

Good point. I've myself previously removed noise from the microphone using Audacity (used in post production), then I've used VoiceMeeter and Cantabile with Reafir plugin, and now I'm switched to Nvidia Broadcast (requires Nvidia GPU). Second and third method can be used for removing noise after and during the voiceover recording. But I found Nvidia Broadcast as the winner.

rnnoise is a similar concept as nvidia broadcast (very good but not as good as nvidia's), but open source and it's included now in OBS studio (add filter to microphone -> add -> noise suppression)

other people already do with resin printers... they already have a very uniform and powerful UV source, but i like this methot too ...

If some through holes needed, would you cnc drill them after the laser pass?

@@lpjunction It is rare that there are no holes at all at least for the connectors. I drill with a hand-held drill press. I avoid holes as much as possible by favoring SMD components.

@@danielroibert5631 Simple connector hole could be hand drilled. But modern IC, even a simple one, has more legs than a centipede. One quickly come to a point when 2d is not enough, we need 2 planes with some through holes. Then of course, for prototyping, the through holes could be avoided with jump-wires. Next, you'll be playing with micro-surgery soldering. Have fun. For normal drilling, the drill action is usually downward. I once have a DC motor drill for PCB, the drill is installed under a small 'table', the drill action is upward. To locate where to drill, there is a transparent acrylic glass secured about 7mm above the table, so the pcb can slide freely on the table. For the first time, the drill poke upward to make a small hole on acrylic. Then you can use this hole to guide the pcb at the spot right above the drill bit and press the motor upward to drill. Add a magnifying glass to help, to me it is easier than using a drill press to drill downward.

@@lpjunction Thank-you. I don't know what you are getting at. I don't make PCBs for mobile phones. I just use 2.54 pitch connectors. I mainly use 0.8mm or 1mm drillings and that works fine for me. Your bottom-up thing sounds interesting, but I’m not sure how it looks. For soldering too thin to be done with an iron, I use my reflow oven. And again, I only do prototyping. If I need better quality or small runs, I have them done at JLCPCB.

​@@lpjunction That's a brilliant drilling method! Might be even better if you put the acrylic on thumb screw/spring mounts like a 3D printer bed so you can raise and lower it to be just more than the thickness of the PCB. That would eliminate parallax error when positioning, and serve as support for the drill pressure so you don't have to hold the PCB down to the table by hand.

good idea, the dot of ligth must be better and smaller

Thank you. Bob Ross built up his paintings from base to detail. In an almost scientific way, the way I worked with MSc and PhD students.

PWM and voltage control are about the same thing, since the result of PWM is the average value. The main difference is heat dissipation in the controler.

yes, but it can't make large pcb

Most makers don't have resin printer.

You make me very curious.

Try obsidian 😯😃

@@wilstraver962 0.1mm doesn't pass enough UV, it's not intense enough. this makes it a slow process.

I'm thinking putting the uv led in the nozzle with a drop of distilled water at the tip acting like a lense. You could change the focus by rising or lowering the led

direct contact with the film is simply the best method to do this. no need for any sort of lenses, cheap simple, and perfect results. the final result will be the quality of the 3d printer.

I do not understand your question. What does this method have to do with ABS, PETG etc?

@@wilstraver962 These materials are fairly common in 3D printing, and I don't know how much they like to stick to copper. If you print the circuit with them, and they stick well, then you don't need to customise a 3D printer with an UV LED, just print the circuit to the PCB, and etch it.

@@wilstraver962 he means this and yes it is possible kzfaq.info/get/bejne/bLplrcl1zODde2Q.html

Thank you for the positive response. The pcb is exposed in 1 run. The amount of light on the film is determined by the intensity of the light source, the speed that it moves over the film and the size of the pinhole (0.1 0.2 0.3mm .. etc). So 3 parameters to control the exposure.

Filling the pinhole is a good idea. I tried that with UV glue and also with a two component glue. Unfortunately a lower light output. I have not investigated this further. But any solution from anyone is interesting.

In the 1990s, at work, we used an LPKF protomat 91s. This was a reasonable way to make pcb boards. Later we had a LPKF protomat with automatic tool changer, I don't remember the type but it was the top of what was for sale. Horrible, many broken tool tips. Very expensive to repair, a faulty micro switch in the tool head would cost 3000 Euro. Anyway, I had a lot of trouble with it. Milling is a thing of the past for me.

Many such reactions have already been given, very childish to do this again.

Thanks for the update, I hadn't read any comments and thought I'd share my experience. You had better go reply to all of them, then I'll reply to tell you you're being childish.

@@Teh-Stig sad

The resin approach works pretty well for single-sided boards, but a potential problem with that approach comes up when a dual-sided board is desired - registering both sides of the board with one another. This approach allows one to use the fiducials to easily define an accurate offset to expose the second side and be assured any through-holes will line up, no matter what size/shape the PCB blank is. Guaranteeing the registration between sides would be kinda tough to do with a resin printer without substantial modification. Good work, Wil!

Watch out, you might end up with a diffraction pattern.

Uv light for pcb usually need a circa 400 nm wavelenght. Red laser is 630/670 nm, green is about 520. Blue is about 445 nm, so maybe with blue ...