What sort of photodiode are you using for timing the laser pulses? How are you biasing it? Is it getting triggered by stray light scatter, perhaps? You could mount it at the end of an opaque tube with a diameter no larger than the beam width, with a light-trapping inner surface (like vantablack), so that only a well-placed beam can hit the sensor. Wrap the photodiode assembly in copper foil and use a shielded cable to connect it to the amplifier. You could also use two photodiodes and derive a differential signal for better noise rejection. Overall mighty impressive!
Since you have the dual galvos and the extra driver channel, you could use a DC bias on the vertical galvo signal to 2D pan the output around, would add a lot to function to be able to move a lower resolution/ higher refresh image around in 2D space than having to use the raster space for movement. Would be awesome for some signage, yard art, even theatrical applications where a smaller image needs translated around a larger surface.
That's right, I remember now. I was screaming at you because you only used a single mirror array. My old lab setups from the very early days used two spinning arrays. The first one basically looked like a table top cut of a diamond. Very close to what you have here. The second one was more of a Decagonal rod. You'll get much more control if you use motors with resolvers rather than trying to match up two free spinning elements. However you are going to require some more computing power. These setups are a lot of fun to play with in free Spin but you have to calibrate on every power up and you need a very stable power supply. I made one for the sole purpose of projecting Flames onto smoke for halloween. Projecting through film with a 0 point light source is a wonderful effect. I couldn't afford resolver Motors so I tried to cheat with steppers. It made synchronizing a little easier but there was still drift. Didn't last very long so I ended up using a DMD chip.
Great improvement! Would replacing the horizonal galvo with spinning mirror help with the refresh? Also, would a higher resolution encoder disk help with the timing?
@Space_Reptile While reading your comment I instantly harkened back to one of my childhood's favorite characters of the 80's....🥁 Max Headroom! I have probably thought about building a Max Headroom of my very own since I was a kid. I even remember using my first computer, a Tandy128K I think, to program a talking avatar head in Basic. I wish that I hadn't given it to my friend 😕.
As the mirror has a significant mass, it can't change speed quickly, so you don't need to hard-sync on every pulse - you could filter/phase-lock the sync signal in software to eliminate the jitter, though starting with a cleaner signal would be better. Maybe use a reflective opto sensor looking at a mirror face to avoid the need for extra mechanical parts. Or perhaps tap into the brushless motor driver to avoid the need for sensing?
Exactly what I was thinking, these polygon mirrors are already phase locked, just finding one with a workable driver IC is a pain, I went through this already and found a really good one from a printer. If both axes are phase locked then the scan will be super stable.
I do think that using a beam sensor would be superior to the method I used, and in fact, I did experiment a little. At 8:10 in the video, you can see a phototransistor in line with the beam. The problem with that approach was that I kept getting false positives due to secondary reflections inside the enclosure.
It could be even simpler: since you know approximately your motor speed, you can use timer interrupt to start with something approximately close to the index pulse, and then correct the timer duration with every cycle
When modulating the lasers in analog mode, do not let the video level drop below the cut out point of the laser diode. In other words, do not let the diode switch all the way off. You will find that at a low level, after projection at speed, the line will be black, though the laser is technically still on. In this mode, you can modulate into the MHz region with ease. I used this technique to modulate a fiber optic driving laser at 330MHz back in the 1990s.
^ Yes!. For gas lasers this is "pre-ionization". In laser printers using IR diodes, just prior to the start of each page, a calibration routine ramps up the laser current until it barely lases. Then this value is held in a "bias" DAC that establishes a baseline for the driver. A second DAC establishes the proper "drive" level. Now the circuitry easily modulated between these two levels.
@@AcuteChronicplease excuse this "grammar nazi" post, but something you wrote got me thinking... Do lasers lase, or laser.. I think it's the latter. While adding "r" suffix to a verb creates a noun for the person or object that performs the verb, the word laser wasn't coined that way. The "r" isn't a suffix, so the verb isn't "lase" I guess a laser should really be a laserer... I'll go away now...
@@dougaltolan3017 LASER is an acronym for an action. LASER's "laser", as in they perform "light amplification (by) stimulated emission of radiation", their name is the "verb" itself.
@@dougaltolan3017 lase (lays), lasing (lay-sing) is a verb found in major dictionaries, "(of a substance, especially a gas or crystal) undergo the physical processes employed in a laser; function as or in a laser:"
@@benmakeseverything Dithering would work, but I am not so sure that PWM is ruled out here, sure the lines would be dotted, but as long as the frequency is high enough, I would think it would have the apparent effect multiple levels of intensity, kind of like a one dimensional half tone. After all, with a few exceptions, every printing and display technology is based, more or less, on the phenomena that different dots close together is perceived as one of the average intensity of them all.
Agree, but I would suggest not just dither static image, but convert it to video or gif with dithering by using ffmpeg. It has multiple dithering methods, I would probably suggest atkison or sierra3. You can try other as well, but with B/w screens atkison works best for me even with resolutions like 100x50. High framerate is important though.
I have worked on some "low resolution pixel magic" before, and yes, i believe it can be quite some upgrade (given that there are still only those 8 colors available) BUT: i think it only really has a nice effect with pictures and not with cartoon like images. It will get more blurry but if u see it from the far u might have a better chance at recognizing what is shown. ... but other idea: can u draw one line two times? Or rather set the vertical speed to 1/2 to then have 28(hope my math is mathing xD) diffrent colors? (U just take every possible mix of your 8 colors) ... it might add flicker or halfes ur y resolution, but adds colors. Also maybe u could make a algorithm to share the pixel colors a lil more (maybe add 28 colors roughly but only reduce y resolution by 1/3 🤔)
You also don't **need** a galvo to do the secondary scan, you can use another polygon mirror just spinning at a different rate. This is how many older thermal cameras work with a single detector pixel. Most of those polygon mirrors have pretty bad controls and won't like low speed reference signals, but there's one that does it really well, it's the MP6500/MP7000/MP8000 mirrors (edit: Ricoh MP6500 photocopiers, give the polygon mirror motor a square wave, and it locks to that square wave). These will require no accelerating or decelerating like the galvo, and has phase locked loops to provide a very stable spin speed.
I considered that, but thought it would be very hard to synchronize them. There may be purpose built motors for that kind of thing, but I'm unfamiliar with them.
How many sides are these polygon mirrors? I have checked the listing online but the casing hides all the good stuff Do you have any yourself? Can you describe the polygon mirror dimensions and number of sides / faces please? Would really appreciate
@@benmakeseverything the motors I'm referring to are very similar to the one you used in the last video, but ones from Ricoh MP6500 photocopiers have much higher ranges of locking frequencies, give it a square wave and it phase locks the motor to it. to synchronize multiple you just need to get a clock divider driven by the same clock source and they will all sync up perfectly, stably.
@@alexwang007 This is a good solution. You would want to have two polygons with differing number of mirrors if the non-square aspect ratio is desirable. A 8 and 12 mirror pair would give you the traditional 3:4 ratio and so on keeping the final scan angle in mind. The further you intend to project the more mirrors you would probably want to maintain higher light output for the same area. This is reminiscent of the sensors in military night sights from the olden days and some first generation mechanical TV systems. A few TV prototypes had around 100 tilted. Some of the night sights had 16 mirrors and 8 microbolometers to give 128 lines of resolution. With the flexibility of modern micro and galvo he does not need to combine the lasers, just have then vertical and then use software to alight the coloured images and save the optical complexity and losses.
Holy crap this video is a masterpiece of improvements being made. Awesome how something like a CRT can be replicated into a projector. Usually projectors are just boring lcds but this one has LASERS!!! Nice job.
@@videolabguy Yeah, the early pocket projectors before DLP used lasers and a single mirror that was electromagnetically scanned in two directions with resonant frequencies using just one coil, genius.
@@benmakeseverything you can film it in "slow motion" by matching the refresh rate of the projector to the framerate of the camera. For example if you're recording at 29.97 fps make the laser go at 30 hz so there will be a slight missmatch in speed resulting in the laserdot seeming to complete one full cycle at just 0,03Hz or once every 33s. Very interesting project btw
I have exactly the same project in my boxes but for a volumetric display on a semi-transparent plastic surface rotating in a glass ball. This video helps me a lot to understand the challenges that await me in this project. Thank you! Very good video and project!
I started collecting pieces parts from laser printers and super market checkout scanners years ago to build out this very concept. Cool to watch the evolution of your project. Nice job so far.
I really love the jitter, it gives the exact amount of retro/future vibes. And I love your project especially this, IT IS SO AWESOME!! I thought to rebuilt the first version but second version blew my mind!. Thanks for your awesome videos!!
Commenting for the algo. I love the start up sound and the scanline jitters. Projects like these, that marry digital and analogue, are the best kind of learning opportunities - great work!
Love the project and the video. I wrote a paper about building one of these about 20 years ago. Instead of using spinning mirrors and motors, I went with Opto-Acoustic modulators.
For timing the spinning mirrors you could probably use a system similar to what's used on some closed loop control BLDC motors. A radially polarized disk magnet on the axis of rotation and a hall effect sensor
I worked for canon in 1990 I was a CLC specialist it had a large Michaelson scanner, the laser had to be focused- surprisingly often, it was a large version of your laser printer scanner, the rotation of the drum being the rasta scan. Great stuff took me back 😊
The laser printer scanning module has a sync output. You could use that for the timing. Instead of a galvo, you could use another spinning mirror that's slower and synced to the horizontal scanner, but I suspect they're not readily available. If you could get the distance between mirrors small enough, you might be able to do an X/Y configuration, but that's a lot of precision machining.
One of Baird's later mechanical televisions used a helical mirror, made of a series of flat, angled sections. This produced a raster pattern directly from the mirror geometry. Then all that's required is the laser timing and mirror position to be sync'd. It does require a line output from the laser rather than a point, but that's fixed optics.
THIS is what I need!!!!!!!! For the sake of it, I made an open source light show sequencing software 2 years ago, that works for embedded systems like this, though I was only able to do and make rgb strobe lights or flood lights, and somehow come into my mind of someday making a laser show fixture out of using embedded systems as well, then I saw the first video, it got me hyped and sub on the channel, I just can't get enough leverage to outsource the materials needed, but this video is really one of the thing I cant miss, someday I might be back here when I decide to build one myself, for this crazy hobby.
Amazing project, I absolutely love the static glitched lines that occasionally happens. I could see this being used in a movie as practical effects rather than CGI.
Amazing project! I work with lasers and one of my laser marking machines use a galvo head. This machine have two lasers: a 50W fiber laser which is the laser that engraves and cuts metals, and the other laser is a low-power red laser diode whose beam is aligned with the fiber laser beam. Before engraving, the machine can project the design onto the part that will be engraved using only the red laser to visualize the design.
I feel like dithering could help the colors a bit. I also feel like color should be secondary to size and resolution. Makes more sense to add 3 green lasers to help with the resolution vs. RGB. That's just me though. Nice work. I want to build one!
Your video is great! I knew very little about how these projectors worked before your video. Despite being more complex than I imagined, you have made this device and the technology that controls it much easier to understand. Thank you! I look forward to your next video.
Why not use another spinning wheel mirror for the vertical axis ? THen the only limitations is how the 2nd wheel can go and how fast the modulation of the laser can go as well THe potential of this new invention are amazing!!!! Thank you for sharing! I am halfway through the video and already plan to reaplicate your system with RGB maybe ? Really I think you found something Thinking out of the box !
Oh my god there's a sequel! Fantastic work on this project, and thanks for showing your process in so much detail - it's very inspiring for stuff I'd like to build myself!
How about using a slower rotating polygon mirror also for the vertical scanning, with its rotation axis perpendidular to the horizontal scan rotatation axis? The scanlines would get slightly slanted due to the constant rotation of the vertical scan mirrors, but that should not be a major issue. Very cool project like the previous ones! I had this kind of idea myself many years ago, but only on paper.
I thought about that - just wasn't sure how to sync them up without using gears. Given this motor is spinning at about 15,000 RPM, I thought that would generate an insane amount of friction. Maybe they could somehow be synchronized electronically.
magnetic worm gear? or let the software figure out what color to emit based on the current position of horizontal and vertical mirrors, ie. let the hardware move the (x, y) indexes in the software. This way the better the synchronization, the better image quality, but no longer will they need to be synchronized perfectly.
If living near an airport the authorities will get their knickers in a knot if you do not apply for and receive suitable laser show licence/permission, impossible in some places.
He would just have to call his local air traffic authorities. I'm sure as long as he doesn't hit a plane he's fine. It's not illegal to shine lazers in the sky, only at airplanes
Any unterminated lazer use is oh, so frowned upon. Is it worth going to jail? I don't think so as this has the potential of being charged with assault.
@@NibNa5ty That would be logical but there are a lot of people who have been brainwashed into believing that a glint of a laser will cause a pilot to malfunction and some like to throw their weight around. Some places no one will care, other places you will have a visit from unfriendly folk. Wisest choice is to make sure who you have for neighbours if you live near air traffic. Obviously there may be busybodies that will object even without nearby air traffic but this is less likely to cause you big problems if you are not building a space weapon. In some places use of a more than 5mW laser that is handheld/portable is considered a no-no unless used for astronomy education as a pointer. Many places will require laser show permits if the beam could strike someone in the eye and is above some threshold. If the projector was used to display constellations in the sky that tried to align with the real stars it might not break laws.
I had almost this same exact idea back in 1993. Using spinning mirrors to project a laser beam which recreated how a CRT television worked. There were no flat screen or LED monitors or televisions back then. I figured it be good for outdoor advertising on billboards or with the right reflective surface allow drive in movie theaters to show movies in the daytime. Standard televisions had 525 horizontal lines. So I was going to use a flat disk the size of a vinyl record with 525 mirrors each set at a slightly different angle. Then I thought if I had a second set of spinning mirrors which angle only changed so one mirror would be angled for the even horizontal lines and the second field of the odd horizontal lines you only need 4 mirrors on this second spinning set of mirrors with the opposite side mirrors having the same angle. This would cut down the number of mirrors in the first set of mirrors from 525 mirrors to 263 mirrors. If all 4 mirrors were set to slightly different progressive angles you could cut down to 132 mirrors. I never figured out the exact geometry though. I never had the skills to actually build the thing. So happy to see someone actually do it. Wonderful job! You got a new subscriber here. :D
I have been researching the laser world for a while now, and while I haven't built anything yet, I think I could probably make it happen. What if you used a board from a DVD ROM? This would give very precise control of the lasers, as well as resolving any clocking issues. Also, your battery is a bunch of 18650's in series. Easy to remove the housing and directly solder to. Instead of 2 res LED's, use a beam splitter. Almost all laser diodes produce a rectangular spot. A convex lens at the aperture may solve resolutions. Thanks a million for these videos, as well as your entire channel!
Is there a known maximum switching frequency that can be obtained on a breadboard given its stray capacitance? Is 100 nanoseconds even attainable in those conditions?
My highest frequency I've got working on breadboards is 27 MHz (~37 ns). 32 MHz should be doable, but haven't managed it myself. I'd almost be willing to say that anything under 100 MHz is doable, tho have no backing for it or any references - just a wild guess.
Are you familiar with BeamBrush? That is how the laser show industry is moving forward with “filled in vector imagery,” and is the reason we don’t try to use raster imagery. It may be an interesting next project for you to tackle. BeamBrush controls the divergence of the beam on a per-point basis. It has two lenses that are sat parallel to each other and next to each other, and on one side of these lenses is a 90° mirror pair on a voice-coil actuator that moves toward and away from the lens pair. When it moves away, it increases distance of the light’s path between the lenses. When it moves toward the lenses, it decreases the distance of the light’s path. Very patented, very Pangolin, but also perfectly awesome territory for a TH-cam video. ~Your friendly neighborhood Laser Safety Officer for Laser Light Shows
I think you should consider using esp32 for the next version. It has even higher clock speed, built-in DACs and quite a lot of processing power. Built-in wi-fi can also be convenient for wirelessly updating image in real time.
Great project! 🤩 I am impressed with your experience. You can add "free" cooling by just drilling some nice holes in the bottom and in the top cover. Holes to be covered with some dustproof net from inside (like in good laptops), and foots to be added to have airflow from the bottom. Convection will do the job possibly without any additional fan.
2:06 - Soooo… only a few more iterations then we can play melee on it right? Jokes aside, there apparently is a display company, Prysm, that makes laser phosphor displays which are basically CRTs but using lasers to excite the phosphor grid instead. Apparently they have high resolutions and refresh rates available, CRT like motion clarity, and in theory no “native” resolution just like CRTs but there’s very little information about them online and they have no intention of making a consumer version. (I would pay a concerning amount of money for a modern day version of a CRT 💀). It seems sorta similar to what you are doing, at least with trying to accurately steer a laser beam that fast
Fantastic progress you made with this. One thing I know about the scanner PCB is that it is equipped with feedback circuitry to prevent overshoot. Projecting a square with a galvo setup without this circuitry, you get a pincushion effect. The feedback circuitry slows down the mirror before getting to the setpoint and can give the galvo a kick to overcome inertia when starting to move, so utilising the driver board may allow you to get a cleaner output but to be fair you managed pretty good. Also, use lasers with the best TEM00 beam profile. Easier said than done but back in the day, gas lasers were better in that respect compared to cheaper solid state types.
Great video. I am currently designing and building a high speed FDM printer but have since your first video strongly considered continuing your work on a affordable DIY laser projector as my next project. Keep up the good work.
I have been looking at something very similar and had the same idea of a spinning disc and galvo mirror for the vertical. You can use a set of optics to get the spot more circular and help focus the beams to a tighter single point. I've been toying with getting a disc CNC machined, but as it would be a one off, runs expensive. Good to see what you've done so far, and encourages me to keep at mine. Good luck moving forward.
I had the same idea years ago, so glad you thought of it and made it! :) What I had in mind was a bit different in exactly the RGB laser drive area: use analog signal like VGA and use actual analog video amplifiers! ;) Basically you can take the RGB output amplifier from a VGA monitor and modify them.
Me and my brother conceived this idea before blue LEDs even existed a long long time ago I'm so glad that somebody did it I know it'll be a commercially available product
Fantastic ideas and builds and love your attention to detail with design and aesthetics, keep up the great work looking forward to see what you come up with next. Love from Canada (East) 🦊
This is really cool. It more or less works like the lasers in photo minilabs do, wild how you arrived at that setup on your own. Great minds and all that. Really, most of the remaining work is optimization honestly, you could probably fix the EM noise and any tracking issues, optically correct the diodes and all that. I do kinda like the cyberpunky feel to it though.
As u said you kept the video length reasonable, but people like me would like it more detailed which would help us understand more and make such stuff too. btw creator/makers/engineers like you guys motivate me to do good stuff and not waste time, thank you ben😊😊.
This would make a pretty cool visual element for live productions. Most of your indoor tests look like they'd be right at home at an EDM show, or even a nightclub.
The "normal" laser projectors keep the lasers on all the time, then have an array of mirrors that either project or dump the pixel for each color (to put it simply) - this is a pretty cool idea !
Really cool project! If refresh rate allows, temporal dithering might help being able to expand the color palette. It'd essentially be displaying the same frame at least twice with subtle differences between each to simulate different colors. POV effects might combine both frames into one, although doubling the colors would halve the frame rate.
I had tried to do a similar thing 40 years ago during my university days, but we did not have the laser diode technology back then (no blue either), nor fast enough microcontrollers. But one thing that I thought could work fast enough was to use KDP crystals with the electro-optic effect to modulate the laser beams. These days I would consider the relatively new DLP technology. Enjoyed your video, keep em coming.
This is super cool. Shortly after your first video I got a portable MEMS raster laser projector from China which scans at 33khz, unfortunately the beam profile and convergence is quite poor, and interestingly enough it uses 5 lasers. Yours is actually sharper, haha. That was the only one i found for sale then, now they seem non-existent outside ebay.
Ah someone came by at our makerspace Hack Manhattan making a very similar Laser Image with a switch and similar device. Not sure if it was you or not as this was a few months ago but it is rad that this is what it has the potential to do!! A lot of possibilities with this! Awesome work!
It wasn't me, but that's interesting. This is my second version of the device, and I'm sure others have built similar things in the past. Thanks for watching!
For extra resolution, i wonder if you could: A. Wiggle the laser beam with a piezo or voice coil piece attached to a mirror or lens B. Use a beam splitter and fast attenuators to draw two lines at once C. Use a grating beam splitter and thin, off center grid filter to draw multiple lines at once (the moving dots would "blink" in a regular pattern as they hit the diagonals of the grid filter). Idk. Great job with this!
I'm really digging this device. Very cool project. I built a Lifesize 1:1 R2-D2 and would love to build a smaller version that could be used to project the Leia Hologram onto a glass block or other surface to simulate the famous scene from the first Star Wars film where Artoo projects Leia's plea for help message to Obi Wan Kenobi. Really cool project man!!
This one took a while... Let me know how you would improve this device!
What sort of photodiode are you using for timing the laser pulses? How are you biasing it? Is it getting triggered by stray light scatter, perhaps? You could mount it at the end of an opaque tube with a diameter no larger than the beam width, with a light-trapping inner surface (like vantablack), so that only a well-placed beam can hit the sensor. Wrap the photodiode assembly in copper foil and use a shielded cable to connect it to the amplifier. You could also use two photodiodes and derive a differential signal for better noise rejection. Overall mighty impressive!
Too cool!
Since you have the dual galvos and the extra driver channel, you could use a DC bias on the vertical galvo signal to 2D pan the output around, would add a lot to function to be able to move a lower resolution/ higher refresh image around in 2D space than having to use the raster space for movement. Would be awesome for some signage, yard art, even theatrical applications where a smaller image needs translated around a larger surface.
That's right, I remember now.
I was screaming at you because you only used a single mirror array.
My old lab setups from the very early days used two spinning arrays.
The first one basically looked like a table top cut of a diamond.
Very close to what you have here.
The second one was more of a Decagonal rod.
You'll get much more control if you use motors with resolvers rather than trying to match up two free spinning elements.
However you are going to require some more computing power.
These setups are a lot of fun to play with in free Spin but you have to calibrate on every power up and you need a very stable power supply.
I made one for the sole purpose of projecting Flames onto smoke for halloween.
Projecting through film with a 0 point light source is a wonderful effect.
I couldn't afford resolver Motors so I tried to cheat with steppers.
It made synchronizing a little easier but there was still drift.
Didn't last very long so I ended up using a DMD chip.
Great improvement! Would replacing the horizonal galvo with spinning mirror help with the refresh?
Also, would a higher resolution encoder disk help with the timing?
that jitter just feels nice, unless it goes out of control
feels retrofuturistic
Futuristic but also vintage! Looks like vhs artifacts, which i loove
Yes, Blade Runner feel.
@Space_Reptile While reading your comment I instantly harkened back to one of my childhood's favorite characters of the 80's....🥁
Max Headroom! I have probably thought about building a Max Headroom of my very own since I was a kid. I even remember using my first computer, a Tandy128K I think, to program a talking avatar head in Basic. I wish that I hadn't given it to my friend 😕.
wonder if instead of an EMF Artifact you dont have a stray optical "glint" causing mis-trigers of the horiz-scan.
As the mirror has a significant mass, it can't change speed quickly, so you don't need to hard-sync on every pulse - you could filter/phase-lock the sync signal in software to eliminate the jitter, though starting with a cleaner signal would be better. Maybe use a reflective opto sensor looking at a mirror face to avoid the need for extra mechanical parts. Or perhaps tap into the brushless motor driver to avoid the need for sensing?
10/10 excellent suggestions.
Exactly what I was thinking, these polygon mirrors are already phase locked, just finding one with a workable driver IC is a pain, I went through this already and found a really good one from a printer. If both axes are phase locked then the scan will be super stable.
I do think that using a beam sensor would be superior to the method I used, and in fact, I did experiment a little. At 8:10 in the video, you can see a phototransistor in line with the beam. The problem with that approach was that I kept getting false positives due to secondary reflections inside the enclosure.
@@benmakeseverything Just use one of those ir laser led's on the other side of the mirror
It could be even simpler: since you know approximately your motor speed, you can use timer interrupt to start with something approximately close to the index pulse, and then correct the timer duration with every cycle
The jittering scan lines is an aesthetic in of it's own. Very reminiscent of Star Wars holograms and Blade Runner advertisements. 💕
Rent your projector out to film companies wanting that kind of aesthetic.
@@jimmyzhao2673way way cheaper and easier to do as vfx in post unfortunately
Seriously… looks great
When modulating the lasers in analog mode, do not let the video level drop below the cut out point of the laser diode. In other words, do not let the diode switch all the way off. You will find that at a low level, after projection at speed, the line will be black, though the laser is technically still on. In this mode, you can modulate into the MHz region with ease. I used this technique to modulate a fiber optic driving laser at 330MHz back in the 1990s.
^ Yes!.
For gas lasers this is "pre-ionization".
In laser printers using IR diodes, just prior to the start of each page, a calibration routine ramps up the laser current until it barely lases. Then this value is held in a "bias" DAC that establishes a baseline for the driver. A second DAC establishes the proper "drive" level. Now the circuitry easily modulated between these two levels.
@@AcuteChronicplease excuse this "grammar nazi" post, but something you wrote got me thinking...
Do lasers lase, or laser.. I think it's the latter.
While adding "r" suffix to a verb creates a noun for the person or object that performs the verb, the word laser wasn't coined that way. The "r" isn't a suffix, so the verb isn't "lase"
I guess a laser should really be a laserer...
I'll go away now...
@@dougaltolan3017 LASER is an acronym for an action. LASER's "laser", as in they perform "light amplification (by) stimulated emission of radiation", their name is the "verb" itself.
@@dougaltolan3017 lase (lays), lasing (lay-sing) is a verb found in major dictionaries, "(of a substance, especially a gas or crystal) undergo the physical processes employed in a laser; function as or in a laser:"
@@dougaltolan3017 It's called "back formation". English if full of them.
dither your images! can get much more true-to color images even with your limited palette
I'll have to look into that, hadn't considered it. Thanks for the input!
@@benmakeseverything Dithering would work, but I am not so sure that PWM is ruled out here, sure the lines would be dotted, but as long as the frequency is high enough, I would think it would have the apparent effect multiple levels of intensity, kind of like a one dimensional half tone.
After all, with a few exceptions, every printing and display technology is based, more or less, on the phenomena that different dots close together is perceived as one of the average intensity of them all.
Agree, but I would suggest not just dither static image, but convert it to video or gif with dithering by using ffmpeg. It has multiple dithering methods, I would probably suggest atkison or sierra3. You can try other as well, but with B/w screens atkison works best for me even with resolutions like 100x50. High framerate is important though.
Dithering requires more resolution to spare, as does PWM
I have worked on some "low resolution pixel magic" before, and yes, i believe it can be quite some upgrade (given that there are still only those 8 colors available)
BUT: i think it only really has a nice effect with pictures and not with cartoon like images. It will get more blurry but if u see it from the far u might have a better chance at recognizing what is shown.
... but other idea: can u draw one line two times? Or rather set the vertical speed to 1/2 to then have 28(hope my math is mathing xD) diffrent colors? (U just take every possible mix of your 8 colors) ... it might add flicker or halfes ur y resolution, but adds colors. Also maybe u could make a algorithm to share the pixel colors a lil more (maybe add 28 colors roughly but only reduce y resolution by 1/3 🤔)
Man the "shimmer and wave" of the projection is sick. Gives it a nice 80s Future vibe.
Even down to the happy incidental pink edge tint. The perfect retrowave music visualizer maybe. 😻
You also don't **need** a galvo to do the secondary scan, you can use another polygon mirror just spinning at a different rate. This is how many older thermal cameras work with a single detector pixel. Most of those polygon mirrors have pretty bad controls and won't like low speed reference signals, but there's one that does it really well, it's the MP6500/MP7000/MP8000 mirrors (edit: Ricoh MP6500 photocopiers, give the polygon mirror motor a square wave, and it locks to that square wave).
These will require no accelerating or decelerating like the galvo, and has phase locked loops to provide a very stable spin speed.
I considered that, but thought it would be very hard to synchronize them. There may be purpose built motors for that kind of thing, but I'm unfamiliar with them.
How many sides are these polygon mirrors? I have checked the listing online but the casing hides all the good stuff
Do you have any yourself? Can you describe the polygon mirror dimensions and number of sides / faces please?
Would really appreciate
@benmakeseverything if they need to be linked maybe a belt system?
@@benmakeseverything the motors I'm referring to are very similar to the one you used in the last video, but ones from Ricoh MP6500 photocopiers have much higher ranges of locking frequencies, give it a square wave and it phase locks the motor to it. to synchronize multiple you just need to get a clock divider driven by the same clock source and they will all sync up perfectly, stably.
@@alexwang007 This is a good solution. You would want to have two polygons with differing number of mirrors if the non-square aspect ratio is desirable. A 8 and 12 mirror pair would give you the traditional 3:4 ratio and so on keeping the final scan angle in mind. The further you intend to project the more mirrors you would probably want to maintain higher light output for the same area.
This is reminiscent of the sensors in military night sights from the olden days and some first generation mechanical TV systems. A few TV prototypes had around 100 tilted. Some of the night sights had 16 mirrors and 8 microbolometers to give 128 lines of resolution.
With the flexibility of modern micro and galvo he does not need to combine the lasers, just have then vertical and then use software to alight the coloured images and save the optical complexity and losses.
Holy crap this video is a masterpiece of improvements being made. Awesome how something like a CRT can be replicated into a projector. Usually projectors are just boring lcds but this one has LASERS!!! Nice job.
Thank you so much! It was a lot of work but a fun project for sure.
This project is quite retro. The first television systems were mechanical in nature and worked precisely like this.
@@videolabguy Yeah, the early pocket projectors before DLP used lasers and a single mirror that was electromagnetically scanned in two directions with resonant frequencies using just one coil, genius.
The end result is super cool. I'm one who enjoys the jitter, it's got a real Blade Runner 2049 aesthetic vibe going with the color hue and jitter.
Those flickers from side to side make the projected image/gif look soooo cool!
I wanna see this captured in slo mo soo badly xD
Me too! wish I had a slow-mo camera
just run the laser clock slower and record at regular speed. free dad joke.
@@benmakeseverything You just shout out to the Slowmo Guys! :D
@@benmakeseverything you can film it in "slow motion" by matching the refresh rate of the projector to the framerate of the camera. For example if you're recording at 29.97 fps make the laser go at 30 hz so there will be a slight missmatch in speed resulting in the laserdot seeming to complete one full cycle at just 0,03Hz or once every 33s. Very interesting project btw
@slomoguys @theslowmoguys
This is by far the one of the coolest projects I've seen.
I have exactly the same project in my boxes but for a volumetric display on a semi-transparent plastic surface rotating in a glass ball. This video helps me a lot to understand the challenges that await me in this project.
Thank you! Very good video and project!
I started collecting pieces parts from laser printers and super market checkout scanners years ago to build out this very concept. Cool to watch the evolution of your project. Nice job so far.
I really love the jitter, it gives the exact amount of retro/future vibes. And I love your project especially this, IT IS SO AWESOME!! I thought to rebuilt the first version but second version blew my mind!. Thanks for your awesome videos!!
Commenting for the algo. I love the start up sound and the scanline jitters. Projects like these, that marry digital and analogue, are the best kind of learning opportunities - great work!
Whoa, how does your comment link to a search?
Love the project and the video. I wrote a paper about building one of these about 20 years ago. Instead of using spinning mirrors and motors, I went with Opto-Acoustic modulators.
this is golden for digital glitch effect reference, my goodness
For timing the spinning mirrors you could probably use a system similar to what's used on some closed loop control BLDC motors. A radially polarized disk magnet on the axis of rotation and a hall effect sensor
Imagine advertisment on clouds with a beefer set up!!! It could be seen from far far away!!! U sure have a commercial potential with this! Love it.
Would also blind anyone that looked at the emitter, there's a reason this hasn't been done yet
I need a playlist of every video sponsored by pcbway. These are always the best videos
I worked for canon in 1990 I was a CLC specialist it had a large Michaelson scanner, the laser had to be focused- surprisingly often, it was a large version of your laser printer scanner, the rotation of the drum being the rasta scan. Great stuff took me back 😊
17:52 "Now I know what you're probably waiting for ..."
YES, BAD APPLE ANIMATION CLIP
The laser printer scanning module has a sync output. You could use that for the timing.
Instead of a galvo, you could use another spinning mirror that's slower and synced to the horizontal scanner, but I suspect they're not readily available. If you could get the distance between mirrors small enough, you might be able to do an X/Y configuration, but that's a lot of precision machining.
One of Baird's later mechanical televisions used a helical mirror, made of a series of flat, angled sections. This produced a raster pattern directly from the mirror geometry. Then all that's required is the laser timing and mirror position to be sync'd. It does require a line output from the laser rather than a point, but that's fixed optics.
THIS is what I need!!!!!!!!
For the sake of it, I made an open source light show sequencing software 2 years ago, that works for embedded systems like this, though I was only able to do and make rgb strobe lights or flood lights, and somehow come into my mind of someday making a laser show fixture out of using embedded systems as well, then I saw the first video, it got me hyped and sub on the channel, I just can't get enough leverage to outsource the materials needed, but this video is really one of the thing I cant miss, someday I might be back here when I decide to build one myself, for this crazy hobby.
The wobbling effects and glitches are SO F*CKING COOL !!!
Except you aren't seeing what he is seeing, once downmixed into 60fps video by the camera
Amazing project, I absolutely love the static glitched lines that occasionally happens. I could see this being used in a movie as practical effects rather than CGI.
the algorithm nailed this recommendation. What a wonderful build. liked and subscribed.
The slight glitching makes this look insanely cyberpunk
Amazing! This went really fast and beyond the results the German TV producer Schneider made in the 80ies with its LaserTV project.
Looks like a fun project. Good job sticking with it and finishing it!
I understood just half of the things you've said in this video, but the world needs more people like you!
Thats beautiful! I love the interference, feels like watching a VHS 😍
Amazing project! I work with lasers and one of my laser marking machines use a galvo head. This machine have two lasers: a 50W fiber laser which is the laser that engraves and cuts metals, and the other laser is a low-power red laser diode whose beam is aligned with the fiber laser beam. Before engraving, the machine can project the design onto the part that will be engraved using only the red laser to visualize the design.
This is awesome.
The only worthwhile improvement is making two of these, and using them like an interferometer for touch applications.
Amazing! Back in maybe 2005 I got similar idea, never attempted to make it. It's wonderful to see such invention in reality!
I feel like dithering could help the colors a bit.
I also feel like color should be secondary to size and resolution. Makes more sense to add 3 green lasers to help with the resolution vs. RGB. That's just me though.
Nice work. I want to build one!
That's a fair assessment. I could have gone that route, just got so many people asking for RGB in the first design, I figured I'd give it a shot.
Why would I use multiple lasers when you can buy commercial products with HD output? How do you calibrate those lasers to be even?
@@ArneChristianRosenfeldtWhy do anything?
I loved the first video on this project, and this was even better! Tremendous work, very impressive.
I can't wait to see you play DOOM on that thing!
Man, you're starting to make a habit of beating me to the punchline on projects I've had on my bucket list for ages.
Neat implementation Ben, it is terrific that you pushed for colour and animation. I love the repurposing of existing mirrored parts
17:54
Ben: "I know what you probably waiting for..."
Me: "HDMI support, yaay!"
Ben: "Longer range tests!"
Me: "Oh"
Well, that would be cool. Maybe if I ever build another one...
@@benmakeseverything is it possible to bend the laser light with magnet like crt?
@@fdgdfgdfgdfg3811 Photons don't have a charge unlike electrons
Your video is great! I knew very little about how these projectors worked before your video. Despite being more complex than I imagined, you have made this device and the technology that controls it much easier to understand. Thank you! I look forward to your next video.
I love the glitchiness and wobble, awesome retrofuture effect. very cool project
The images have a real Cyberpunk feel to them. Awesome work!
Why not use another spinning wheel mirror for the vertical axis ?
THen the only limitations is how the 2nd wheel can go and how fast the modulation of the laser can go as well
THe potential of this new invention are amazing!!!! Thank you for sharing!
I am halfway through the video and already plan to reaplicate your system with RGB maybe ? Really I think you found something
Thinking out of the box !
It would be very difficult to synchronize two spinning mirror wheels.
@@KNfLrPn What if he used a mechanical link between the two mirror wheels ? Like gears or a small belt or chain
@@MonoIit could be viable, but the high speeds are a concern.
@@KNfLrPn someone else mentionned using magnetic gears transmission, could solve the issue of friction
Mag gears have no torque tho it would take a long time for it to get up to speed
This is serene and a unique type of image screen.
18:37 Nothing like being a recently sleeping bird in the tree that is now being blinded by a laser.
Oh my god there's a sequel! Fantastic work on this project, and thanks for showing your process in so much detail - it's very inspiring for stuff I'd like to build myself!
How about using a slower rotating polygon mirror also for the vertical scanning, with its rotation axis perpendidular to the horizontal scan rotatation axis? The scanlines would get slightly slanted due to the constant rotation of the vertical scan mirrors, but that should not be a major issue. Very cool project like the previous ones! I had this kind of idea myself many years ago, but only on paper.
YES YES YES
I thought about that - just wasn't sure how to sync them up without using gears. Given this motor is spinning at about 15,000 RPM, I thought that would generate an insane amount of friction. Maybe they could somehow be synchronized electronically.
@@benmakeseverything timing belt
@@benmakeseverythingMaybe possible with contactless magnetic transfer?
magnetic worm gear? or let the software figure out what color to emit based on the current position of horizontal and vertical mirrors, ie. let the hardware move the (x, y) indexes in the software. This way the better the synchronization, the better image quality, but no longer will they need to be synchronized perfectly.
this is way beyond my skillset so 90% of this video is way over my head but the end result is awesome! has that cyberpunk aesthetic got a sub from me!
On a dark overcast night, please go try to write on some clouds!
If living near an airport the authorities will get their knickers in a knot if you do not apply for and receive suitable laser show licence/permission, impossible in some places.
He would just have to call his local air traffic authorities. I'm sure as long as he doesn't hit a plane he's fine. It's not illegal to shine lazers in the sky, only at airplanes
Any unterminated lazer use is oh, so frowned upon. Is it worth going to jail? I don't think so as this has the potential of being charged with assault.
@@NibNa5ty That would be logical but there are a lot of people who have been brainwashed into believing that a glint of a laser will cause a pilot to malfunction and some like to throw their weight around. Some places no one will care, other places you will have a visit from unfriendly folk. Wisest choice is to make sure who you have for neighbours if you live near air traffic. Obviously there may be busybodies that will object even without nearby air traffic but this is less likely to cause you big problems if you are not building a space weapon.
In some places use of a more than 5mW laser that is handheld/portable is considered a no-no unless used for astronomy education as a pointer. Many places will require laser show permits if the beam could strike someone in the eye and is above some threshold.
If the projector was used to display constellations in the sky that tried to align with the real stars it might not break laws.
Aw man, I’m surprised you couldn’t get a 2 bit R2R working. Great improvement! I especially love the use of Mitxela’s PCB melting script!
I had almost this same exact idea back in 1993. Using spinning mirrors to project a laser beam which recreated how a CRT television worked. There were no flat screen or LED monitors or televisions back then. I figured it be good for outdoor advertising on billboards or with the right reflective surface allow drive in movie theaters to show movies in the daytime. Standard televisions had 525 horizontal lines. So I was going to use a flat disk the size of a vinyl record with 525 mirrors each set at a slightly different angle. Then I thought if I had a second set of spinning mirrors which angle only changed so one mirror would be angled for the even horizontal lines and the second field of the odd horizontal lines you only need 4 mirrors on this second spinning set of mirrors with the opposite side mirrors having the same angle. This would cut down the number of mirrors in the first set of mirrors from 525 mirrors to 263 mirrors. If all 4 mirrors were set to slightly different progressive angles you could cut down to 132 mirrors. I never figured out the exact geometry though. I never had the skills to actually build the thing. So happy to see someone actually do it. Wonderful job! You got a new subscriber here. :D
I have been researching the laser world for a while now, and while I haven't built anything yet, I think I could probably make it happen. What if you used a board from a DVD ROM? This would give very precise control of the lasers, as well as resolving any clocking issues. Also, your battery is a bunch of 18650's in series. Easy to remove the housing and directly solder to. Instead of 2 res LED's, use a beam splitter. Almost all laser diodes produce a rectangular spot. A convex lens at the aperture may solve resolutions.
Thanks a million for these videos, as well as your entire channel!
Is there a known maximum switching frequency that can be obtained on a breadboard given its stray capacitance? Is 100 nanoseconds even attainable in those conditions?
Agreed. And indictance. Looked like damped oscillations.
People build CPUs and Ben eaters world worst video card on breadboards. 20 MHz.
My highest frequency I've got working on breadboards is 27 MHz (~37 ns). 32 MHz should be doable, but haven't managed it myself. I'd almost be willing to say that anything under 100 MHz is doable, tho have no backing for it or any references - just a wild guess.
If the waveform at 11:16 is real, thats most likely inductive ringing, not stray capacitance.
those lines at 8:08 actually make a pretty cool effect
Are you familiar with BeamBrush? That is how the laser show industry is moving forward with “filled in vector imagery,” and is the reason we don’t try to use raster imagery. It may be an interesting next project for you to tackle.
BeamBrush controls the divergence of the beam on a per-point basis. It has two lenses that are sat parallel to each other and next to each other, and on one side of these lenses is a 90° mirror pair on a voice-coil actuator that moves toward and away from the lens pair. When it moves away, it increases distance of the light’s path between the lenses. When it moves toward the lenses, it decreases the distance of the light’s path. Very patented, very Pangolin, but also perfectly awesome territory for a TH-cam video.
~Your friendly neighborhood Laser Safety Officer for Laser Light Shows
No haven't heard of it. Sounds cool! I'm pretty much learning all of these things as I make the videos, I have no industry experience in any of this.
This is a really cool machine, its like a CRT TV but if they were using lasers
I think you should consider using esp32 for the next version. It has even higher clock speed, built-in DACs and quite a lot of processing power. Built-in wi-fi can also be convenient for wirelessly updating image in real time.
ESP32 with more than 600 MHz? Which one?
Great project! 🤩 I am impressed with your experience. You can add "free" cooling by just drilling some nice holes in the bottom and in the top cover. Holes to be covered with some dustproof net from inside (like in good laptops), and foots to be added to have airflow from the bottom. Convection will do the job possibly without any additional fan.
2:06 - Soooo… only a few more iterations then we can play melee on it right?
Jokes aside, there apparently is a display company, Prysm, that makes laser phosphor displays which are basically CRTs but using lasers to excite the phosphor grid instead. Apparently they have high resolutions and refresh rates available, CRT like motion clarity, and in theory no “native” resolution just like CRTs but there’s very little information about them online and they have no intention of making a consumer version. (I would pay a concerning amount of money for a modern day version of a CRT 💀). It seems sorta similar to what you are doing, at least with trying to accurately steer a laser beam that fast
Just looked up Prysm. Very cool - wish they had more details on the mechanics of it.
So cool - love that Bladerunner aesthetic! I'd buy a kit to make one of these in a heartbeat!
3:03 why not ESP32 ? S3
Fantastic progress you made with this. One thing I know about the scanner PCB is that it is equipped with feedback circuitry to prevent overshoot. Projecting a square with a galvo setup without this circuitry, you get a pincushion effect. The feedback circuitry slows down the mirror before getting to the setpoint and can give the galvo a kick to overcome inertia when starting to move, so utilising the driver board may allow you to get a cleaner output but to be fair you managed pretty good. Also, use lasers with the best TEM00 beam profile. Easier said than done but back in the day, gas lasers were better in that respect compared to cheaper solid state types.
Great video. I am currently designing and building a high speed FDM printer but have since your first video strongly considered continuing your work on a affordable DIY laser projector as my next project. Keep up the good work.
I have been looking at something very similar and had the same idea of a spinning disc and galvo mirror for the vertical. You can use a set of optics to get the spot more circular and help focus the beams to a tighter single point. I've been toying with getting a disc CNC machined, but as it would be a one off, runs expensive.
Good to see what you've done so far, and encourages me to keep at mine. Good luck moving forward.
Wow top level content once again! Thanks for your videos they give me inspiration to do electronic projects my self
Watching a movie would be awesome! Also great device, wish I had your knowledge of how to do stuff like this!
i know you mentioned it as a point of frustration, but i really love the way it occasionally glitches and flickers
I had the same idea years ago, so glad you thought of it and made it! :)
What I had in mind was a bit different in exactly the RGB laser drive area: use analog signal like VGA and use actual analog video amplifiers! ;) Basically you can take the RGB output amplifier from a VGA monitor and modify them.
how does this channel not have a million subscribers? great video! would love to see a part 3!
This is awesome! I have always want an actual laser projector since the Graffiti Research Lab laser tag video, great work
glad you took this further. i have a pipe dream of adding it to a ROS robot to draw the robots mental model of the room onto the room itself.
pretty cool what people can make at home these days! ❤
Me and my brother conceived this idea before blue LEDs even existed a long long time ago I'm so glad that somebody did it I know it'll be a commercially available product
Insane effort. Im sure you are already thinking of version 3. Looking forward to the next episode in 6 months. Ill wait.
I hope this is the new projector technology in the coming years, I love projectors and hate how fast I go through them.
Fantastic ideas and builds and love your attention to detail with design and aesthetics, keep up the great work looking forward to see what you come up with next.
Love from Canada (East) 🦊
This is really cool.
It more or less works like the lasers in photo minilabs do, wild how you arrived at that setup on your own. Great minds and all that.
Really, most of the remaining work is optimization honestly, you could probably fix the EM noise and any tracking issues, optically correct the diodes and all that. I do kinda like the cyberpunky feel to it though.
the imperfections are so cool and 80s cyberpunk-y
i like the rgb version , it is really a nice project , that might have some real life use
Huge progress since previous one!
Perfect project. With a powerful laser, it would be interesting to try to project onto the clouds on a cloudy night.
As u said you kept the video length reasonable, but people like me would like it more detailed which would help us understand more and make such stuff too. btw creator/makers/engineers like you guys motivate me to do good stuff and not waste time, thank you ben😊😊.
Awesome mate. The design everyone talks about but nobody actually builds. Great job.
This would make a pretty cool visual element for live productions. Most of your indoor tests look like they'd be right at home at an EDM show, or even a nightclub.
The "normal" laser projectors keep the lasers on all the time, then have an array of mirrors that either project or dump the pixel for each color (to put it simply) - this is a pretty cool idea !
Really cool project! If refresh rate allows, temporal dithering might help being able to expand the color palette. It'd essentially be displaying the same frame at least twice with subtle differences between each to simulate different colors. POV effects might combine both frames into one, although doubling the colors would halve the frame rate.
Very nice devices. I'd like to see something alike on the market. Thanks for making these
Neat and nifty. Thank you for your documentation efforts.
This is so cool man, wow! Love the glitchy visuals ;D
Good job!
ppl like you are amazing, doing things I wish I had or could do. keep it up bro
I had tried to do a similar thing 40 years ago during my university days, but we did not have the laser diode technology back then (no blue either), nor fast enough microcontrollers. But one thing that I thought could work fast enough was to use KDP crystals with the electro-optic effect to modulate the laser beams. These days I would consider the relatively new DLP technology. Enjoyed your video, keep em coming.
This is super cool. Shortly after your first video I got a portable MEMS raster laser projector from China which scans at 33khz, unfortunately the beam profile and convergence is quite poor, and interestingly enough it uses 5 lasers. Yours is actually sharper, haha.
That was the only one i found for sale then, now they seem non-existent outside ebay.
Ah someone came by at our makerspace Hack Manhattan making a very similar Laser Image with a switch and similar device. Not sure if it was you or not as this was a few months ago but it is rad that this is what it has the potential to do!! A lot of possibilities with this! Awesome work!
It wasn't me, but that's interesting. This is my second version of the device, and I'm sure others have built similar things in the past. Thanks for watching!
For extra resolution, i wonder if you could:
A. Wiggle the laser beam with a piezo or voice coil piece attached to a mirror or lens
B. Use a beam splitter and fast attenuators to draw two lines at once
C. Use a grating beam splitter and thin, off center grid filter to draw multiple lines at once (the moving dots would "blink" in a regular pattern as they hit the diagonals of the grid filter).
Idk.
Great job with this!
I'm really digging this device. Very cool project. I built a Lifesize 1:1 R2-D2 and would love to build a smaller version that could be used to project the Leia Hologram onto a glass block or other surface to simulate the famous scene from the first Star Wars film where Artoo projects Leia's plea for help message to Obi Wan Kenobi. Really cool project man!!