Laser diode self-mixing: Range-finding and sub-micron vibration measurement

[@jer_h]

I did my PhD on this exact topic. Yes, it is true that you can get the signal using just the laser diode itself, you need an extremely stable current source for the laser and then a very high gain amplifier with low noise on the terminal voltage of the laser. Even in academia, few people try to do it because using the photodiode is way easier. There is also some semiconductor noise which shows up in the terminal voltage signal which is hard to get rid of. You can determine the direction of movement in your piezo speaker example by just looking at the slope between the fringes.

[@jer_h]

Also, the quality of the signal you get back using the terminal voltage depends strongly on the structure of the laser itself (ie VCSEL, DFB, etc)

[@JackKSharp]

I did my Master's on SMI, shake hands :)

[@talbakish9479]

look at vocalzoom. they did this sensor and it's for sale. extremely robust and can also work as laser mic

[@jer_h]

@@CG-cw3ps "high power" means different things to different people, and certainly depends on the wavelength. A 10W CO2 laser (about 10um wavelength) is considered to be a baby CO2 laser, whereas a 10W 405nm laser would be an absolute monster. For me, 50mW at 850nm was about the limit, but that would easily cause permanent blindness and the beam is invisible to people. As far as unexpected behaviour, I sort of think of lasers like cats; every cat is different, and you can only really encourage them to do what you want.

[@jer_h]

C G I’m not totally sure, but my guess is that you made some sort of copper oxide by heating the copper

[@CuriousMarc]

The Japanese company Keyence makes commercial laser sensors based on a multi wavelength refinement of the basic interferometric technique you demonstrate so nicely here. They can get down to 1 nanometer resolution!

[@phillipsalisbury1385]

Sony makes laser hologram encoders with 8 picometer resolution :D

[@russellzauner]

Bolt it to a slab of metal that weighs at least 1000lbs and you'll see your jitter drop enormously. lol There is a lot one can do in their lab for cheap to approximate the big kids. When doing dimensional analysis, particularly at this small of feature geometry (or at least the hardware is capable of it with some tuning), having standards that are at least in the ballpark of your dynamic range is critical and if you can't afford a set of $1k slides/blocks then buy 3 sets of cheapos that might have been built with something moderately calibrated in some factory somewhere and reconcile them until you get your field of measurement dialed in to where you can tell the difference between your three cheapos - then at the very least you can ask the most connected person you know to take yours and at least match them to a calibrated set and give you some correction notes, or if they have any moderate fabrication tools available to them, see if they can/will correct your standard directly.

[@russellzauner]

Actually, not so much jitter as conducted environmental noise goobering up your signal. So you need to isolate/decouple too or you may get seizures from the noise twitching.

[@liwenchang3260]

Two german companies make the exact same instrument with with single wavelength and achieved 1 pm resolution

[@movax20h]

There are also specialized range finders, used in geodesy, that use multi wavelength interferometry and time of flight delay, to automatically compensate for atmospheric refraction, and get millimeter accuracy over kilometers. Some are still in developement, but techniques used are awesome.

[@SmarterEveryDay2]

Excellent Macro Videography on this one.

[@MatthewStauffer]

I see you're in early! Subscribed to this channel with alerts too? Good move. There's another channel I do that with, can't remember the name of it though...

[@peteabc1]

there was this report on seeker a week ago about the fastest camera able to capture propagation of light, hmm

[@AdityaMehendale]

I missed this - where exactly is the macro videography? Destin and Ben, you both create amazing content, but this comment just seems patronizing.

[@mabdialibeki3728]

@Aditya he probably refers to what the oscilloscope makes visible - very cool indeed!

[@rogervanbommel1086]

Wait, why did you comment with your second channel?

[@preddy09]

When an Applied Science video comes out, you know it's time to drop everything else

[@Ringer1982]

Yes, sleep is overrated any way.

[@abdulalhazred5924]

パンツドロプ

[@CodeParade]

You can also get sub-micron distance measurements from this setup! Basically, you could calibrate the exact distance to a target reflector one time, and then use peak counting in software to determine the distance as it moves closer and farther. Kind of like the speaker, but free-floating. As long as peaks are never miscounted, it will maintain its accuracy to sub-micron levels, and you can move the reflector arbitrarily far.

[@Steve_Just_Steve]

Sounds like I just got me a new DRO for the mill. lol

[@mckenziekeith7434]

Cool idea! Very sensitive to any change in wavelength, though, and relies upon the diode maintaining phase coherency without interruption over the entire operating period. If the laser diode looses coherency, you will miscount. It is my understanding that lasers do have coherency dropouts from time to time.

[@giggawc8457]

@ Steve just Steve You have to use 2 photodiodes phase shifted the half wavelength to count up and down. Or just buy a interferometer dro sios-de.com/ I believe renishaw sell them too

[@TheRainHarvester]

Wouldn't ambient light affect the measurement?

[@CodeParade]

It shouldn't as long as the ambient light isn't also coherent at the laser wavelength. That said, it would still be a very sensitive setup, not practical for most applications, but doable in lab conditions.

[@Totalis1989]

I believe that the movement of the reflective target by 1/4 of a wavelength results in a path difference of 1/2 wavelength which changes the light from constructive to destructive (or visa versa).

[@Graham_Wideman]

Good thinking. So peak-to-peak of the waveform corresponds to displacement of 1/2 wavelength.

[@yuanfutan1719]

so the displacment equals to number of fringes multiply the wavelenth or half of the wavwlength ?

[@dedskin1]

@@yuanfutan1719 half the wave length , i dont know much about light but i know sound , that is radiation as well , and if obstacle is bigger the half the wave length it will reflect ALL of the wave , if its smaller it will reflect nothing . So you have 1 and 0 , digital type of thing , so you can in theory build Sound computer. Is that not crazy . You would need however millions of transducers , but i think it can be done , and that from it a lot of interesting inventions will come . So by it self is not much , but it will lead to a lot of development .

[@Wayne_Robinson]

Really interesting experiments and illustrations! At first I was envious of your oscilloscope, but then I simply enjoyed being a voyeur watching you use it. Tektronix is still killing it after so many decades.

[@lhxperimental]

You are really living up to the name of the channel. Each episode of yours has technology with business potential of millions or even billions of dollars if one considers how many applications it can be put to use in.

[@iangrant9675]

You know, it is really _weird_ how, when I come back to watch one of these videos, after a few months, it seems like there's _more_ stuff in it! How am I ever going to understand all this?! 😂❤️💓💕

[@JoshuaPalley]

I didn’t understand how a TIA worked until you explained it so simply. Thank you!

[@Afrotechmods]

Holy crap. I can imagine this might form the basis of an optical servo tweeter. Or a new way of characterizing drivers at least.

[@BillySugger1965]

Afrotechmods Might be a challenge at tweeter frequencies. But my application is monitoring the performance of low frequency drivers, up to 2kHz, and there this might be a practical proposition if enough signal could be extracted from the photodiode output.

[@miniwarrior7]

@@BillySugger1965 subwoofer distortion monitor aye?

[@vk2zay]

Laser vibrometry *is* the standard method of characterising acoustic drivers.

[@Steve_Just_Steve]

@@vk2zay you have an interesting channel also vk2zay. subbed

[@vk2zay]

Thanks mate. I've been ignoring it lately, time to make more content!

[@RajasPoorna]

I love this. The entire comment section is filled with the same kind of nerd as me. I've been wanting this company for a long time now. And obv Ben is out of the world, I don't have words to describe your amazingness. I would like to be like you. I'm a physicist/bioengineer and this is exactly the sort of thing I like. But you walk in and out of chemistry like it's nobody's business and I'm envious of that. I would like to learn. Again, you're amazing.

[@electronicsNmore]

Fantastic as always!

[@Hlkpf]

word

[@jcybarra1]

This is very helpful for understanding signal processing in vibration measurements. I am currently using a laser doppler vibrometer to characterize a MEMS device I built for my PhD project. These are $100,000+ systems and I heard that cheaper methods were possible. This was a nice demonstration of one such method.

[@TheMovieCreator]

The offset between constructive and destructive interference is indeed half a wavelength, but since the wave has to travel both to and back from the reflector, the displacement of the reflector has twice the effect on the offset of the wave. The difference in reflector displacement between full constructive and full destructive interference is therefore the quart of a wavelength, not half a wavelength. But pretty impressive stuff!

[@eelcohoogendoorn8044]

Came to comment on the same topic. Indeed the path-length difference in the reflected ray is double the displacement of the surface; but the destructive interference would only result in an intensity maximum every half wavelength if counterpropagating relative to the non-reflected reference beam. If the reference and reflected beam are coaxial and traveling in the same direction (which seems more likely), then a full wavelength shift of the reference beam is required to cycle the intensity pattern once. So assuming the reflected and reference beam hit the diode from more or less the same direction, id say 2 maxima per single wavelength displacement of the reflective surface. But it really depends on how the beams meet, and if that is in a nontrivial way, all bets are off; though 4 maxima per displacement is indeed a theoretical maximum. Best to get out a micrometer and measure it!

[@Refthoom]

You need to take into account the distance between the LD and PD as well, right? Or is that to be ignored because it's a constant?

[@eelcohoogendoorn8044]

​@@Refthoom Yeah pretty much. Should you vary that distance, you would not observe any variation if the light bundles are travelling in the same direction; or two maxima per wavelength if they are travelling in opposing directions. But as long as you dont vary it the actual distance does not matter since the light forms a periodic pattern.

[@michellesoinne9219]

I noticed that also and saw you already had commented. Important detail.

[@talbakish9479]

Actually, in commercial implementation (such as in Vocalzoom IC) we modulate the laser and we get much higher fringes speed per movement

[@pppppierre]

The set up act also as a microphone. Whenever the experimenter talk a signal can be seen. It appears that the loudspeaker, acting as a microphone with natural modes that amplify some frequency. It is a pleasure to follow the clear thinking of the experimenter.

[@beforebefore]

Nice... similar to the principle that I co-developed for a microwave motion detector back in the late 1970's. It just used a simple/cheap Gunn diode microwave source (10GHz), high-impedance power supply, and detect Doppler shift as AC/audio across the Gunn diode. It used about a 6" parabolic dish antenna, which was also the reflector for the light... as this was an automatic motion sensing yard-light.

[@tiberiu_nicolae]

Did it fry the neighbors cat?

[@RobertSzasz]

You can now buy the same thing for a couple dollars, most seem to just connect the RF circuit to a board designed for a PIR sensor and it works well enough.

[@hawke2325]

I'd like to know more about this I find unpublished development history fascinating

[@RobertSzasz]

Big Clive shows the modern version kzbin.info/www/bejne/fJjHiYWCn76emZY

[@hexarith]

The technique presented at the end of the video, sweeping the wavelength and measuring the interferometric fringe spectrum is essentially Swept Source Optical Coherence Tomography (SS-OCT). The axial resolution is proportional to the sweep bandwidth, i.e. the higher the bandwidth, the higher the resolution. If you scan the laser over a sample, and for each scanning point take the spectrum of the fringe signal you get depth resolved reflectivity. Do this with large enough bandwidth, and with fast enough sweeps and you can create volumetric scans like this: kzbin.info/www/bejne/rHbKo6OtqbVji7M

[@erikisberg3886]

Thanks for a very interesting video! Have made a few reflective sensors using laser diodes and very thin fibers for micro mechanical measurements inside inkjet heads. Used small half mirrors from ES and external photodiodes. These worked surprisingly well but calibration was difficult. Wish I had known this self mixing principle then, since Interferometry gives an absolute measurement.

[@britttullos8119]

One of the best, most interesting and thorough videos that I’ve seen in a while. And he just simplified what others before failed to completely explain in such a way that So many answers I’ve had and gaps in my understanding of laser light and optics

[@rarelycomments]

On the square wave you are hearing the sharp rise and not the small ringing. Basic fourier stuff, a rapid change requires high frequencies. The steeper the change, the higher the frequencies produced. This is why the "ramped" square wave was not audible.

[@KravchenkoAudioPerth]

That's correct. Might I add that the driver used, or pretty much any driver will have a hard time reproducing a clean square wave at anything other than low midrange and down without a great deal of jiggery pokery in phase compensation techniques.

[@rarelycomments]

@@KravchenkoAudioPerth Fortunately that doesn't really matter too much, as we are only sensitive to the magnitude of the spectrum and not phase distortion.

[@pingvinac]

Don't know where else to put this, so : in my life and here on youtube i saw on numerous ocasions "cleaning" PCBs with compressed air. Being in refrigeration industry for years i realized that there are few reasons for not using compressed air : - either dry or not it will get the parts wet because of low temperature of expanding gas - it can push the dirt under components like ICs and connectors - combined dirt with moisture makes a pcb bad in moment. Vacuuming with integrated brush sorts all that out (low pressure sucks dirt and moisture). Now i wait to hear from average Joe ...blablabla.. but they all do it... blablabla :) But, it often happens before trying the part/system so you can always say that it was already bad ;) Maybe this could be a subject of interest for you, for one more brilliant video ;) Greetings from Croatia, keep up the good work. p.s. sorry to all for off-topic letter

[@johnnyhammersticks1695]

You can see if your AC electroluminescent displays are actually oscillating at the driving frequency. It would make sense that the width between the top and bottom electrode is expanding and contracting due to the strong applied electric field.

[@YodaWhat]

If you make the laser light bounce off both the top and bottom of the EL strip at the same time, you may get double the amplitude, and see double the count of interference fringes. Comparing that measurement to single-sided will provide a nice double check on results.

[@davidparrish1133]

Wow. Great work! One of the first projects I worked on back in the early 80's was programming a laser dilitometer we built. It used a laser interferometer to measure the change in length of dental material samples in a precisely controlled optical furnace.

[@Heksu77]

Wohoo! Every day when Ben releases a new video is like christmas and birthday party at the same time.

[@C2H5OHist]

So you feel like Jesus?

[@drewlarson65]

lmao I was the last comment and here I am two years later rewatching all sorts of Applied Science vids, you learn more every time.

[@IDoNotLikeHandlesOnYT]

Some optical drives use self-mixing interferometry. I was confused when I disassembled them and didn't find any apparent sensor to detect the light reflected by the disc. I don't remember how I looked it up online, but I somehow found self-mixing interferometry in my searching, and thought it was pretty cool that you could use the free space between the laser diode and the target as a kind of secondary laser cavity, and measure how well that cavity lases as a proxy for the distance. I had no idea it was this easy to do, though. (I say easy, but I hope it's even easier than it looks, because I don't have a fancy SMU or DSO…) The FMCW macroscopic range measurement is a cool bonus, too. Now I'm imagining building a pocket tool like those commercial laser distance measurers, but that measures centimeters to micrometers instead of meters to centimeters (and also has a tachometer function).

[@AureliusR]

It's probably more likely that they use that technique just to determine if the disc is there or not. Even if it appears there's no separate sensor, the laser and photodiode can be packaged very close together with some optics to reflect the light coming back into the receiver.

[@smallmoneysalvia]

This is the best youtube channel. I honestly mean it. You make the coolest stuff, and you don’f treat us like we’re dumb. Thank you so much.

[@jmpattillo]

Holy jeez that is a big oscilloscope screen.

[@MysticalDork]

They charge by the square millimeter.

[@feelx92ger]

@@MysticalDork DPI actually means Dollar per inch, a highschool tech supplier once told me. Pretty impressive stuff Tektronix makes today for that amount of money, nonetheless.

[@TheSethcoleman]

The price tag on that scope is comparable to new mid-range SUV...

[@jmpattillo]

It better tidy up my bench for that price

[@T2D.SteveArcs]

when your trying to show 120000 people whats going on needs to be big lol

[@JustSushi0]

You have one of the most interesting and educational channels on KZbin. Please never stop making content. ☺️

[@w2aew]

Very cool Ben! Always get some great physics visuals from you! Hmmmm - I'll have to get in on this Nerd Thunder thing... ...or maybe I'm not nerdy enough..

[@AppliedScience]

You'd be a most welcome addition to Nerd Thunder! Dean Segovis (Hack-a-week) is the organizer. I added your channel to the list in my video description, so you're in!

[@w2aew]

Thanks!! I’m honored to be among the KZbin Nerd elite!!

[@kingsman428]

@@w2aew *"...not nerdy enough..."* As if. 😁

[@tetrabromobisphenol]

A major contribution to the "stair step" waveform shape is caused by the transfer characteristic of the transimpedance amplifier itself, which is effectively an integrator with a reset time proportional to the RC time constant of the feedback resistor and feedback capacitor. You want to match the feedback capacitance as closely as possible to that of your sensor (the photodiode). Note that at these small capacitance values, the wires themselves can start to dominate the total effective capacitance. So you want lead lengths as short as possible. You also need good (low ESR and low ESL) decoupling capacitors between each rail and ground. Great video, this is a really cool use of the monitor photodiode!

[@johnnyhammersticks1695]

You can cover the laser diode with a smooth thin reflective membrane and maybe make a microphone. Maybe gold foil or aluminum/silver deposited on transparent membrane. The membrane would oscillate with the surrounding air and reflect the laser light back into the photodiode. They're close proximity (right in front of it) would minimize unwanted vibration or movement.

[@sc0or]

You can better light up a distant window and hear what they say =)

[@ee6lpzfzj023]

Hands down one of the best engineering channels ever. Thanks very much, it was very entertaining for my inner nerd.

[@jauld360]

The Sony SLD3134VL laser diode includes a photodiode, according to the spec sheet. The cost is about a dollar a piece on eBay.

[@rizdalegend]

I'm happy that I understood 4% of what you said, keep it coming! BTW do a demonstration about light speed, and how its unimaginably fast. That will help

[@TheCaphits]

Seems like there could be some ******REALLY****** useful uses with this. Extreme precision from a distance? So cool. So useful. This could be a game changer in measurement. All the fancy scope stuff could be streamlined and put onto a chip for pennies. Laser diodes can be extremely cheap. This could be everywhere, and everything could have extreme precision.

[@MusicBent]

Caphits a really common use is in optical media like CR-ROMs and BlueRay discs. 👍🏼

[@YodaWhat]

@@MusicBent > Caphits a really common use is in optical media like CR-ROMs and BlueRay discs. Yes, but those only need to sense the _change in distance_ of ONE-QUARTER WAVELENGTH, and within a predictable range of frequencies, while ignoring the much larger but much slower changes due to mechanical vibrations and non-flatness of the disc. They also have to ignore the laser mode-hopping and other changes to coherence length, which is not an issue for those optical discs, but IS a problem for many other applications. The coherence length of laser diodes is usually mere centimeters. Much better stability of the laser output is necessary for general-purpose laser measurements, so much more complex optics are needed. Long cavity lengths and/or external mirrors are usually part of the solution, but difficult to apply with laser diodes. HeNe lasers, on the other hand, are very stable after warmup.

[@MusicBent]

YodaWhat true. Thinking about it again, maybe a more related application is the lunar ranging experiment. I believe they use a short burst and then use timing to measure the distance. Also, still waiting for the moon bounce ruby laser video 😂

[@williamogilvie6909]

I have built and tested FT-IR spectrometers. They use a Michelson interferometer with an internal laser interferometer that generates a sampling waveform. The repeatability of sampling is very precise - often less than 30 nanometers of drift over a 24 hour period. The laser signal resembles the scope pattern you are showing, although most of it is a sine wave. At each end of travel the laser signal frequency drops down to zero Hertz, similar to the discontinuities your waveform shows. One company I worked for, in the '70s, made a film thickness gauge for semiconductor films. It was a low resolution IR spectrometer that only had to do peak detection.

[@hrtlsbstrd]

Great stuff as always man, love the way you break down common devices to get at some crazy physical principles 👌 (As a cognitive neuroscientist, you've definitely inspired some of my instrumentation)

[@thetruthexperiment]

It’s crazy. I remember when these were new and I wanted some so bad. Who knew that they’d come and go and be rare before I could tell time passed. Insane.

[@scanlime]

great video! I want to know what the signal sounds like! and can you use the self-mixing interferometer as a laser microphone? or a distortion pedal? :D

[@AppliedScience]

That's a great idea! We could down-convert the signal at several different frequencies, and mix them all together. Maybe get a chorus effect?

[@alexwang007]

And for the distortion effect, mabe a recording one period of the waveform and using it for a waveshaper is more viable; i believe you are suggesting to feed the input as the current modulation signal, and the output taken from the PD?

[@trey1531]

Is this what spies use to hear people in buildings by reading the vibrations on a pane of window glass?

[@41ch]

There was a research paper, and sample video, using an extremely high speed camera to capture vibrations induced by sound on various objects to infer or capture whatever was on the other side of a glass window. kzbin.info/www/bejne/fHy7gKiZjadnl5o

[@Lossanaght]

I might be imagining it but around 8:30 and 25:50 but there seems to be a strong correlation between lots of little spikes in the signal and his voice, especially sibilants.

[@soulwynd]

I like it how we can easily see the vibrations caused by your voice too.

[@skylerlehmkuhl135]

So if I understand this correctly, the distance measurement is the same thing as modulated continuous wave radar, except in the visible spectrum? Really cool!

[@talbakish9479]

The distance measurement is actually like a FB interferometer where the fringes rate (or distance between fringes) is correlated to the distance from the target

[@Steve_Just_Steve]

This is so over my head but I understand enough of it to grasp how interesting and cool it is thanks to your awesome ability to explain things in a common sense way. I'm not just saying that, I've always struggled with electronics and you really do have a gift of -dumbing it down enough for me to get it- = ) breaking something complex down into the simpler mechanics of it and explaining it in a way that my mechanical mind can understand. Thanks Ben

[@myself248]

I always thought it would be neat to have a laser microphone like this, but with the membrane some distance away from the source/measure, connected by a fiber. The fiber itself would be insensitive to EMI, which could be a virtue in some environments.

[@rewolff2]

Very interesting video. One small error... To go from constructive to destructive interference requires a half-wavelength difference in path-length. To go from constructive to constructive requires a full wavelength in path-length difference. But because the path is to and from, for example, your speaker, the speaker only moves half a wavelength for each full cycle. So, for most of the video, your speaker is only moving around 3 microns!

[@VladyYakovenko]

I have some basic in electronics through school, signal theory because I make music and lasers cause I stydied some physics from a book. This video put all of my knowledge together in a complete delicious package. Super cool to follow, and very clean. This is was 100% gourmet food for my brain.

[@ARVash]

Next on applied science, making a desktop gravitational wave detector

[@ARVash]

Does make me wonder if you could just loop a TON of fiber optic cables instead of shining the light a very far distance.

[@theharbingerofconflation]

@@ARVash yes, yes you could

[@drmosfet]

Make you wonder, if monitoring the sum collections of errors rate on a high speed fibre optic back bone trunk of a communication provider, will be of used?

[@anteconfig5391]

@Robert Slackware Holy shit I think that's totally possible. If you could direct a beam up that high and then back from there without the beam dispersing so much.

[@alfredjodokusquack2]

Your way of explaining and the balance of details are brilliant. This had exactly what i was searching for plus some context on top. Thank you so much.

[@LazerLord10]

ahhhhh, why midnight? Sleep or this?

[@poptartmcjelly7054]

sleep is for the weak

[@user255]

False dichotomy, you can sleep and watch the video when you wake up.

[@marksmod]

sleep, trust me

[@thewolfin]

Turn off your notifications.

[@davetriplett8109]

@@thewolfin... they are 😅

[@tom_something]

17:17 - This is an important phenomenon to understand, so I'm glad you mentioned it. I was watching a video about a very, very low-pitched flute. They said the lowest notes are outside the range of frequencies the human ear can perceive. Many people in the comments said that they _could_ hear those low notes in the video, but what they were really hearing was a bit of acoustic "slapping" at the frequency of that note. Similarly, we cannot "hear" a 1Hz sine wave, but we can easily hear someone clapping their hands once per second. This is also the reason I'm not a _huge_ fan of allowing vocal fry in low-note-singing contests. They're just clapping... with their throats.

[@mrlithium69]

clapping would be impacting two surfaces. compared to one surface vibrating the air.

[@tom_something]

@@mrlithium69 I'm using the term as a gross simplification. We could go with "clicking" instead. What matters is the underlying principle that separates "hearing" a frequency from hearing an audible event that repeats at that frequency. If our ear canals don't have the hairs for it, we're not hearing it.

[@riaan_za932]

this would be really cool if you use this laser diode to read vinyls and play it back over a speaker, laser diode vinyl player nice!

[@benjaminmiller3620]

Not impossible, but very hard. Challenges: Focusing the laser (need to read depth information from a very tiny spot at any time) Stereo signal. ( there are actually 2 analogue signals in each groove. recorded at ~90 degrees from each other. tracking (the original grooves were laid down mechanically, and there is no way they are a perfect spiral.) If you can build a dual read head CDROM drive with active vibration compensation, you can do it!

[@y__h]

You are basically reinventing Audio CD :D

[@zlotvorx]

It's done. Google laser vinyl player.

[@riaan_za932]

@@zlotvorx yeah, i see a "LASERPHONE" haha.

[@zlotvorx]

@@riaan_za932 I forgot the name and brand, only remember it's Japanese and was shown to me by a HiFi enthusiast.

[@Hirudin]

Mind-blowing stuff! Thanks! While watching I was being reminded of chatting with a land surveyor who told me that laser distance measurers ascertain distances by comparing the wave(s) of the light emitted to the wave(s) of light that bounce back. Although I think he did a good job of telling me the gist of what's going on, your video really made it much more clear. I believe the surveyor left out any mention of current sweeps or light interference since they're somewhat deep topics for a casual on-the-job conversation. He also said that you can get better accuracy by emitting more "waves" at once. In other words, you can get better accuracy with three simultanious frequencies than with only two. I think most people assume laser distance measurers are "simply" measuring how much time it takes for the light to bounce back, learning the way they actually work is incredibly fascinating!

[@MusicBent]

A wave reflected from a mirror will have inverted polarity, so I believe the top picture shown at 11:00 is incorrect. Destructive interference would happen as shown in the bottom picture, but constructive interference would occur after a lambda/4 shift (causing a lambda/2 difference between source and reflected waves) A retro reflector would also do this because it has and odd number of reflections (three). I think also that the measurement would be twice as precise as you say. There should be two points of constructive interference, and two points of destructive interference per wave cycle. So you have two maximums on the oscilloscope per cycle. Great video as always!

[@MusicBent]

This website has a great animation of wave interference at boundaries. The behavior depends on the refractive indexes of the two mediums, but what I stated above should be true for light going from air -> mirror/paper -> back to air. www.animations.physics.unsw.edu.au/jw/light/reflection-and-phases.html

[@malachilandis9542]

"The physics is a bit above my head" - Well there isn't hope for us then. Really cool how you can take such a neat concept and break it down. I feel like I'm discovering it with you!

[@kyle4224]

Was the piece of paper picking up your voice as you were speaking? Could this be used as a very sensitive microphone by turning the distance measured into a wave function?

[@Kalanchoe1]

I've heard of lasers being used to pick up acoustic vibrations on a pane of glass but I can't imagine its easy.

[@coletrain9173]

Check out vibrometers. You can use them to easily measure acoustic vibrations in materials.

[@mckenziekeith7434]

I think it was vibrating in response to his voice. I think in principal it could be used as a microphone. But it may not be practical, except for spies and whatnot.

[@wreckingangel]

@@mckenziekeith7434 Not difficult at all, in fact it is a very nice electronics beginner project: www.instructables.com/id/Laser-Beam-Microphone/

[@flomojo2u]

I noticed the same thing, particularly the “ess” sibilant sounds seemed to be picked up the best. I wouldn’t call it so much of a microphone as an effect of the pressure waves from speech modulating the movement of the speaker cone. Very cool though, and just another indication of how sensitive this laser/circuit is.

[@bastiengrimaldi6510]

Great Showcase of self-mixing well explained. At 12:10 you said that if you count 10 fringes it means that the speaker moved by 6500 nm but actually it's half because of the round trip.

[@AxGxP]

Thank you! Very interesting.

[@FesixGermany]

So the laser diode saga continues. Very interesting to see that the waveform is even responding to your voice when you pronounce a harsh "s". By the way thank you for retweeting my photos of the blue laser diode. Had a huge impact!

[@naikrovek]

Jesus that scope is huge.

[@trombre]

Just looked it up, it's a Tektronix MSO58....bout $35,000...and that probe is $1,800.

[@naikrovek]

@@trombre Yep, it was a gift from Tektronix.

[@NickFoxQuixand]

This video makes me want to buy another oscilloscope

[@Debraj1978]

I am considered technical specialist, but this video is out of my limits. Thumbs up.

[@AnotherGlenn]

15:46 "...because the physics is slightly over my head." Sure, right! HA! :)

[@brandysigmon9066]

Right, this is one of the smartest guys on youtube when it comes to this stuff.

[@rosebarnes9625]

Exactly! I'm sitting thinking "dang I'm pretty smart, but I'm only understanding 2/3 of what he's saying, so I don't stand a chance on that!"

[@ritobt]

This is great, I work with interferometers for research, this is rather cool demo and explanation of a really cheap but great tool!

[@ToukoMies]

You could probably control the laser with pulse width modulation, and then measure how much photosensor data is lagging behind to determine the distance between the laser and the measured object.

[@MatthijsvanDuin]

And how would you perform the femtosecond-resolution time measurement required to achieve sub-micron distance resolution?

[@ToukoMies]

Sub-micron resolution wouldn't be possible, but accuracy of few millimeters would be easily achievable.

[@clusterfork]

"Pretty cool," you say. I say this is absolutely brilliant, probably my favourite project of yours yet.

[@AtlasReburdened]

What beautiful physics.

[@johnhosky2931]

The best part is at 16:14 when you say retroreflector you can see the reflection of your hand in the screen of the Tektronix... a reflection in a reflection in a reflection... btw I love all your videos!

[@alexwang007]

PLEASE DO A QRNG!! for single photon counting avalanche photodiodes, vl6*** series laser proximity sensors from ST-electronics could be a cheap source(sub $20 in small quantities here in canada).

[@stevethul1]

I did notice that the laser was actually responding to the vibration of your voice. Great stuff.

[@BaldBozo]

I watch these videos hoping to pick up even one percent of what this guy knows.

[@jetkingknight]

when the stuff you learn in physic 2 is applied in the real world.... I am speechless; thank you for the amazing video

[@NourMuhammad]

I think your voice was affecting the measurements and disrupting the paper! 8:42

[@dtiydr]

Cool looking, and surprisingly advanced, flat screen TV there.

[@kengineer_au]

Any reason you can't use FFT to extract "the number of steps" per cycle in the final experiment/measurement setup?

[@kengineer_au]

I'm going assume from the ❤ that there isn't any obvious reason...

[@talbakish7045]

Actually, in the commercial version of this, it is done with FFT.

[@GimmickEffects]

Please use this to record sound that is vibrating objects!!! Everybody else, thumbs up this if you want him to do it!

[@nattsurfaren]

Question: Is it possible to turn incoherent light like light you get in a cloudy weather into coherent light?

[@MusicBent]

nattsurfaren yes! In fact Dennis Gabor won a Nobel prize for using coherent light to create the first hologram before the laser was invented. Lasers have both spatial coherence and a very narrow frequency range (essentials just one frequency). Narrow frequency light can be created from a light source like a mercury vapor arc lamp (similar to the orange ones used in some street lights). Special coherency can be achieved by passing the light through a pinhole to create a point source of light. (I can not explain the reason why this is true in this comment). A device like this is called a ‘spacial filter’ which are explained on google.

[@nattsurfaren]

Thanks @@MusicBent

[@RobertSzasz]

No. Lasers are are special because they produce phase coherent light. There is no way (that I know of) to filter for a single phase of light.

[@nattsurfaren]

​@@RobertSzasz To my understanding that if you use a magnifying glass to make fire from the sun light you can only get as high temperature as the temperature on the surface of the sun. So if there were a way to "filter" light from a cloudy sky to coherent light you would only get the temperature that you get from the surface of the clouds. But this is kind of ideas that are just spinning around in my head and I really don't know how it would work for real.

[@tolipwen1487]

@@nattsurfaren If you used a huge magnifying glass(or an acre of solar mirrors) to pump a laser you can accomplish in phase polarised monochroamtic light. You might need to pump some non lasing medium as an interim stage though. A ruby laser video exists on this channel.

[@JoseSilveira-newhandleforYT]

Extremely interesting video. Opens up a box full of possible applications. Impressive!

[@drummwill]

"without any external components" he says as he hooks up everything in his shop to the laser diode XD 27:31

[@seanl9313h]

Most of my recent experience is using photo diodes to validate laser pulse counts and shape. This is a very cool idea that I will have to play around with! But as usual this is the internet and people are jumping to some really crazy conclusions for how this can be used. It's a creative re-purposing of an existing component that tinkerers will have fun with, which is my plan. It will not measure vehicle speeds, open a black hole, or revolutionize industry. It is simply measuring the power of the reflected light it collects(which is not limited to lasers, by the way), so you have to have a fixed and repeatable amount of laser power and a target which cannot vary in its reflective properties. If you had an application where you could ensure that your laser diode output power and target reflection was consistent then this would work. Maybe some extra filtering for ambient light "noise", but that is easy. I would not be surprised if this worked quite well given the minimal parts count and cost. Most laser distance measuring is done with time-of-flight measurement. Like range finders you can get from Home Depot, SparkFun, etc... As far as I know, which might not be much, anything other than time-of-flight methods get into some really cool commercial stuff like what SICK, Banner, Keyence, others make. At the extreme end would be the stuff Laser Depth Dynamics makes.

[@eugenew2]

I noticed when you froze the scope with the first since wave, the voltage between peaks were phase slanting (moving forward and reverse in phase). The Doppler effect?

[@MatthijsvanDuin]

Side-effect of AC coupling most likely

[@philoso377]

Doppler effect constitutes that interference cycles count occurs (within each positive and negative slopes of speaker wave) should be uneven. I counted, they were equal. So Doppler is less likely or absence.

[@DietterichLabs]

The abnormal level of technically involved content on your KZbin channel is refreshing

[@fp4303]

I wonder how far can you still get a relatively good sensitivity with this setup?

[@talbakish9479]

with a VCSEL (laser diode) you can go up to 3 meters (maybe 4) because of coherence length. for velocity. (as far as distance measurement, 1 meter is the maximum)

[@n1352-m1i]

I was revisiting your always inspiring videos, and noted a possible factor of two improvement in actual accuray as the beam is actually doing a round trip, so the required motion for peak-to-peak on the scope is lambda/2 in sensor to device distance (cf. instant ~12')

[@Muonium1]

Well at least it's only midnight and not 3am this time so I won't have to call in to work for lack of sleep.

[@Veptis]

I got some kind of laser diode from a maritime imaging system and this video helped me understand what the diode does and might help me allow house the little board I found it on.

[@rizdalegend]

Kinda like a laser range finder?

[@graealex]

Laser range finders either use triangulation, or modulate the laser with different lower frequencies and measure the signal resulting from mixing the original with the received signal. Varying frequencies are used to tune in the exact range.

[@titter3648]

No. Most laser range finders use time of flight.

[@graealex]

@@titter3648 and they measure the actual time by mixing the original, modulated signal with the received one. None of these directly measure light interference. Andreas Spies has a detailed review.

[@cogoid]

@@graealex Laser distance measurers sold in hardware stores (and Leica surveying equipment on which some of them are based) use the continuous beam with the clever modulation / heterodyne method that you describe. But there are also plenty of other rangefinders that actually measure the time it takes for a reflection of a pulse of light to return back to the unit -- many rangefinders used for sports, for example, use this principle. They might be accurate to a foot with a range of some hundreds of feet. Also, "Hughes Tank Rangefinder AN/VVS-1" is particularly famous, because it is based on a ruby laser, and cheap surplus units were in the past wildly available on surplus market. Curiously, some mobile phones today also use tiny single chip time-of-flight laser range finders to detect when they are next to the ear -- to lock the keypad out, so that the ear pressing on it would not create a nuisance.

[@graealex]

@@cogoid It's very hard to build TOF sensors that work with a large variety of distances and are still very accurate. Also the distance sensors in smartphones usually do not employ TOF, rather they measure amplitude and give out a binary signal (near, far) like any other IR light barrier. As I said, most devices DO NOT employ TOF, even with many people claiming otherwise.

[@boblansdorp8479]

Cool video! One minor correction: because you are measuring interference of the laser in reflection, a 330nm deflection of the retroreflective speaker will cause a path-length shift of 660nm of the reflected beam. So the good news is that you're actually twice as sensitive to displacement as you reported! :)

[@3er24t4g1]

Why not use this to level a 3D printer bed?

[@AntonBabiy]

Cause it's way too sensitive to the environment

[@ARVash]

@@AntonBabiy Nothing that can't be compensated for via signal analysis. This is a decent enough idea. The real reason why it hasn't been done is because it's challenging from a mathematics/software perspective also because nobody has thought to do it. The idea of shifting the frequency to get an absolute distance for example is pretty darn clever, and i suspect it has more precision than firing a laser and identifying how long it takes for the light to return.

[@peteabc1]

Because it's overkill and there are probably more cheaper, simpler and reliable sensors. Like a digital indicator. But you can of course.

[@ARVash]

@@peteabc1 well it's not overkill from a cost perspective. It is from an effort perspective but if one person does the legwork then it's done.

[@peteabc1]

​@@ARVash Said you can :). But I warn you there are some hidden hard to solve problems on the path.. Btw that time measurement is called time of flight and it usually isn't precise because light is too fast.

[@volodymyrsirous2430]

Applied Science You got to have the capacitor across the feedback resistor of the transimpedance amplifier to keep it stable. Otherwise, parasitic capacitance at the negative input of the opamp, alongside with the capacitance of the photodiode, would create uncompensated pole which would cause the transimpedance amplifier to oscillate. Compensation of the pole is the main purpose of the capacitor. I enjoyed the video!

[@Redafto]

Today on applied science I'm going to build a nuclear reactor and show you a funny feature of it

[@quarkmarino]

During the whole video I've been all wondered, but you just blew my mind when you proposed to vary the current to change the frequency to measure a fixed distant (shame I didn't think about that at that moment), "of course!" I yelled.

[@alexwang007]

I think I just witnessed a FMCW Li-DAR, WITHOUT a "tunable" laser.................. MOM GET THE CAMERA

[@uwezimmermann5427]

yes, you did!

[@DumbledoreMcCracken]

I hate the term FMCW. It is a stupid conjunction of two meaningful terms. I prefer Continuous FM (CFM), which actually says what I think FMCW thinks it is saying.

[@danielweber9738]

Well it is tuneable probably below 1 nm - but most likely not without spectral mode hopping. This is actually (one of) the trick(s) getting a nice and reliable signal out of your setup. Anyway nice video, what i find funny here is the fact that you use a

[@tazking93]

Well done! It’s not every day you see someone build a diy Michelson Interferometer!

[@rizdalegend]

This is way out of my pay grade

[@Dragonmastur24]

I thoroughly enjoyed that one!(as all the other ones ;D) It was really cool to catch a glimpse of the sound wave your voice produced as the table was picking up your voice!!

[@graemedavidson499]

Very interesting indeed! It made me wonder if a phased array laser had been created yet but on searching I found that in one project, heterodyne interferometry played a part in calibrating the individual laser path lengths rather like we see here!

[@rottenjonny88]

Awesome video as usual. What came to my mind was using this technique for a seismograph.

[@viesturssilins858]

Absolutely astounding as usual, you definitely are a modern renaissance man! (And props for Sci-Hub link!)

[@dimtraveler589]

Very interesting observation. This method is used in RADARs and is called FMCW. After the signal is detected, the FFT needs to be taken and the peak position is proportional to distance. For best results, the current probably needs to be a rising ramp on a pedestal, followed by darkness. It would be interesting how such prototype will work?

[@VINICIUSBH100]

I am a fan of this guy.... very clever and super kind to share with us what he knows.

[@tipfox9212]

Very interesting ! I can see not only the tapping on the desk but also your speech - very sensitive device