DIY webcam spectrometer - IR filter removal, spectra measurements
Vložit
- čas přidán 5. 12. 2022
- Today, I removed the infrared filter from the camera in my webcam + CD based DIY spectrometer. I measure the visible and infrared spectra of daylight, laptop LCD monitor, tungsten lamp, fluorescent lamp, germicidal UVC lamp, sodium and mercury discharge lamp, neon lamp, nixie tube, remote control infrared LEDs and so on.
The previous episode:
• Building a DIY spectro...
Please support my channel on Patreon:
/ diodegonewild
My Instagram:
/ savage_danyk - Věda a technologie
Wow this works better than I first expected with the cheap camera sensor. Very nice filter removal technique, careful as always 😅💪🏻
about 1.5 years this CZcamsr is teaching me how much I can't learn in college or university of course 😍
Thanks for all you teach us.
I can't imagine getting this much education from each of your videos .
I really love this series! From the custom night lamps to the spectrometer. Very fun and inspiring.
Now I really need to build one of these spectral testing units… I have needed one many times !
Nicely done. I was positively surprised when I saw that the calibration is that good with that webcam. Cheers. :)
Yes, the plastic bag, the polycarbonate of the CD and the lens of the camera definitely has losses at certain wavelenghts. Also the color filter matrix on the image sensor subpixels.
Some ideas for further improvements:
- first, try to use a pressed CD instead of a CD-R. The greenish dye doesn't help much with accuracy
- transmissive mode of the CD would probably be better, because in reflective mode, the light passes the polycarbonate twice
- you can try reflective mode with a de-laminated CD, but from the back side. This would eliminate the polycarbonate from the light path entirely (for the price of a significant intensity loss)
- try some actual diffraction gratings
- don't use any diffusor on the slit
- convert the camera to a hole-camera by replacing the lens to a very tiny hole, and find its focal point
- use an old B/W CCD CCTV camera with manual exposure. You can even crack the glass window off of the CCD, so there will be just air between the bare CCD die and the light source.
I can send you some pictures of a cheap old B/W CCTV camera, which uses a bare CCD die out of the factory. Maybe you can find the same model second-hand locally. Hopefully it's not a color CCD in B/W mode (I haven't checked with a microscope if it has a filter matrix). But this cheap one doesn't have manual exposure, so a higher-end model would be better, with a broken-off CCD window.
Hi Mr. @mrnmrn1. Your comment was perfect. Do you know whether is it possible to buy cheap B/W CCD cameras? I´ve tried a lot, but without success. I tried to convert an OV2640 (Very Cheap Camera Sensor), ridding-off the bayer matrix filter. It worked with low efficiency because front illuminated CMOS sensors are not efficient in UV Region. By the other side, backilluminated CMOS Sensors (that are good in UV region) are very difficult to ridoff the bayerfilter without damagging them. Acctually B/W backilluminated CMOS cameras are usually very very expensive, probably because of no mass production.
Awesome! Great little project. Do more please 🥺. I find this very interesting. 😁
Nice man, really enjoying these vids on types of light
That's quite a nice lamp collection you've got there.
I'm really glad you made this video and the last, I was really interested in re-making one of these
Maybe you can also try some laser pointers of different color to check the accuracy of the spectrometer. Solidstate-Lasers/LEDs also change their wavelength a bit, when they are cooled or heated. That might make for an interesting experiment. For image acquisition, a black & white camera (without bayer filter) would probably be best.
Of course I tried to measure a cheap laser pointer I've bought from ebay several years ago. But it doesn't work any more :(.
@@DiodeGoneWild That's unfortunate. Probably the emission spectrum of NaCl in a flame would also work well.
Thank you for sharing all this information. And thank you for sharing the software to make it all happen. You shared that in your last video. Extremely helpful. I'm trying to set up the same sort of thing
Thanks diode! Time to build my own 😊
(Me doing some other work while my phone notifies me about a new DGW video.)
Halts all tasks and starts watching the video.
miluju tě Danyku .... neuvěřitelná bedna ten člověk
Awesome, great project! I love this channel!!
Cool. Someting i can actually build at home. Very well explained
Awesome project!
thans for the video!
you motivated me to continue my own spectrometer project
Cool thank you so much for building this...
Very interesting informative video 🙏🏻🙏🏻❤️❤️❤️❤️👏🏼👏🏼👏🏼🔥🔥🔥🔥 Thank you ❤️🙏🏻
Awesome ! 🔥🔥
Thanks fir the video, makes me want to build my own spectrometer. Btw, you can buy basic camera modules without IR filter - search for Arducam.
Since your recent videos revolved around LED lighting, it might be interesting to check Phillips WarmGlow or similar lamps with variable colour temperature.
btw, for some reason, CZcams doesn't generate CCs on your videos - it's feature I use a lot to watch videos without audio.
Best accent! I like it. Your pronounce is so clear :)
Cool project. 👍🏻😀🇬🇧
Tracing paper works great as a diffuser, btw. I've cut circles out of it for use in light shades.
Good visually, for visible light. Still guaranteed to absorb certain wavelengths.
i think the calibration ofset is due to cd prism is being circle. if the theremina supports or lets you, you should narrow down the spectra rectangle to 3/11 proportion to more like 1/30.
Fascinating, thanks!
Thank you for your support! ;)
Wow super like 👏
you can find tir prism from old cd drives which may work better than cd piece
If I recall correctly, CDs are capable of acting as diffraction gratings (hence why they sometimes take on colors when viewed from an angle). It seems likely to me that this will complicate your spectrometer calibration.
Thanks!
thank you ;)
I wonder if there's a way to use a pinhole lens on the camera so the lens won't interfere with certain wavelengths.
Plastic absorbs UV strongly so that is probably why you are not getting an UV. If the lense was quartz you would get much better sensitivity.
Nice.
Nice 👍
You have to mention that the best CD for this project is any CD of Mariah Carey, but especially the christmas CD's of the world known star. For this experiment only the CD's or Mariah Carey will do, all other CD's of whatever artists will give bad results. It would even give a better result of you cut out a new fresh part of a Mariah Carey CD, every day!
Nice! Pretty much any ordinary glass blocks UVA. You need quartz optics. Maybe you try a pinhole instead. Maybe you are lucky since you have a bare chip camera. Professional spectrometers use line sensors made of tiny PMTs. Not sure how much they cost
Yes I also suggested him a hole-camera, but with a true B/W sensor, and cracked-off CCD package window. This one is not a bare-die sensor, just looks like one. It must have a thin layer of glass or epoxy on top. A true bare-die would not survive the pick-and-place, wave or IR soldering, and flux washing procedures.
Several types of Glass passes a significant protion of UV-A and even some UV-B.
Is the inaccuracy out in the IR range due to the curve of the track on the CD or is it something with the sensor itself?
Also, you need to find a panchromatic camera to use with this, to fill in all of those dips between the three color bands of the CCD. :D
Won't the plastic bag used as a diffuser change the spectrum?
Awesome serie about spectrometers, what's so special about commercial ones that can see into UV? Quartz lens? I still want to make one, but one that can see into UV, it's my main goal.
At the end, that oscilloscope you received is the same as I bought from FNIRSI about 2 years ago, It's very nice, but mine arrived with a problem in the TP4056 charge controller, that I had to replace in order to be able to charge it. I only use it with a 100X probe, and use it mainly to poke direct into 220V sockets of UPS'.
rock salt would be the easiest, it was used in some old spectrometers. You can melt NaCl and it looks like glass.
I copy my comment for Danyk to you, it might help:
Yes, the plastic bag, the polycarbonate of the CD and the lens of the camera definitely has losses at certain wavelenghts. Also the color filter matrix on the image sensor subpixels.
Some ideas for further improvements:
- first, try to use a pressed CD instead of a CD-R. The greenish dye doesn't help much with accuracy
- transmissive mode of the CD would probably be better, because in reflective mode, the light passes the polycarbonate twice
- you can try reflective mode with a de-laminated CD, but from the back side. This would eliminate the polycarbonate from the light path entirely (for the price of a significant intensity loss)
- try some actual diffraction gratings
- don't use any diffusor on the slit
- convert the camera to a hole-camera by replacing the lens to a very tiny hole, and find its focal point
- use an old B/W CCD CCTV camera with manual exposure. You can even crack the glass window off of the CCD, so there will be just air between the bare CCD die and the light source.
I can send you some pictures of a cheap old B/W CCTV camera, which uses a bare CCD die out of the factory. Maybe you can find the same model second-hand locally. Hopefully it's not a color CCD in B/W mode (I haven't checked with a microscope if it has a filter matrix). But this cheap one doesn't have manual exposure, so a higher-end model would be better, with a broken-off CCD window.
@diodegonewild i'm sorry if my comment is unrelated to the video but when i saw the mercury vapor lamp you used to test the spectrometer i tought about it.
Basically today a friend of mine donated to me a vintage mercury vapor lamp, it was of his grandad and it used to be inside of a lampost at his house wich is sadly gone now. The lamp is a 250w MBF-U so i definately need a ballast for it but before purchasing it i looked ibside to see the conditions of the arc tube wich was clear with very minimal gray spots near the electrode, but I noticed something sinister about it, it is bulged! It's not a huge bulge but it's definately there, so my questions are: 1st why or when does this happen. 2nd does it mean that the bulb can explode or that it is better not to turn it on? (The bulb did work last time it was used but that was over 25 years ago as my friend told me)
Hey idol, can you make a list of all lights that are applicable in optical spectrometer? Please
Next we will be doing atomic absorption spectroscopy.
I have two types of it laser diodes one from DVD and other from green laser IR - pumping diode but when i see the ir light of these diode via Camera with its HOT glassed removed I can see the difference in colour in camera display ( not National ir colour of course but in screen ) one is purple and other is light Yellow .
Can you do the measurement ?
So thats why the uvc lamp makes everything in the room light up so weird
The spectrum on some so called ' UV germicidal ' LEDs might be good for a laugh.
was wondering how a linear ccd would react to color bandwidth
09:32 Maybe that lamp uses mostly Argon gas (with a dash of radioactive barium, strontium or Krypton-85 gas to help kick-start it) that produces UV light and internally illuminates blue/green that excites a phosphor coating producing the visable color seen?
Yes, probably argon or other noble gas, maybe a misture of more gases to reduce the ignition voltage. It probably produces UV to excite the green phosphor. I guess there's no mercury, otherwise we would see the 436nm blue.
@@DiodeGoneWild see en.wikipedia.org/wiki/Penning_mixture
Cool video! Where to get the software? Is it for free? Thanks!
Ahh yes mercury lamps, it almost looks like gray light
Good vidio
Naiiise
What if you would peel aluminium film without destroying it and use that as reflection grating. Then only lens would be limiting factor (and sensivity of sensor)
Aluminium film removed from a pressed CD is a great idea! I suggested him a similar solution, just to use the de-laminated CD's back side in reflective mode. But that would cause a big drop in intensity. Your idea is better, but it's probably essential for the diffraction grating to be perfectly flat, which might be a bit hard to do with the peeled-off piece of aluminium film. The lens problem can be solved by not using a lens, just a tiny hole instead.
@@mrnmrn1 If you cut a DVD you can split it in 2 layers and get a reflective aluminum grating on one side of the plastic.
Would you loose the ability to determine spectra if you remove the lens altogether? (shine directly on the sensor) I'm guessing that the lens polycarbonate from it's sharp cut of UVB and C.
The spectrometer works by looking at the light intensity along the horizontal spectrum created by the CD. It maps this intensity to a wavelength based on its position.
Without a lense, you just get a blurry image and cant figure out the positions of where light is coming from, so you also cant figure out the wavelengths.
@@Basement-Science Then perhaps a pin hole camera would be good?
@@franzliszt3195 I think this is where the linear ccd sensor comes in.
@@franzliszt3195 I guess so, if the sensitivity was good enough.
The slit coming into the spectrometer already blocks a lot of light, and an additional pinhole after that blocks even more.
does a webcam camera capture the UV ?
Test some liquids how they absorbs spectrum lines
what is the purpose of the plastic bag diffuser?
Can you do the spectrum measurement by cooling and heating the Ccd or Cmos sensor to see if its wavelength sensitivity shifting or not !
You can use peltier or ice with salt for Lower temperature or maybe Dry ice from Extinguisher .
7:00 and for uv you try to remove the lense and try bare sensor or parabolic lense if possible maybe some pice of parabolic plastic bottle cap with shiny aluminium foil to focus the light to the sensor because most glass and plastic can block uv as you know .
7:58 wondering than why this isn't blocked by lense !!
@@omsingharjit probably the lens is made of quartz glass
@@SadamHusajn82 or of acrylic plastic both glass and plastic blocks uv light quit Significantly .
The surface of the parabolic mirror has to be pretty good, or else the sprectral lines will be smushed. But replacing or covering plastics with other materials like quartz and aluminium might improve the setup a bit anyway. Diffraction grating is also pretty cheap on Amazon.
@@lesliespeaker668 yes , that mirror will not be perfect but that's not for Taking image that is just to see if camera can pick that wavelength or not by discarding other distorted wavelength temporary .
NICE COMEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEENT I LIKE THE VIDIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIOOOOOOOO , THANKS FOR SHARIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIING
change the webcam to lowest fps for more exposure
Cdeeeeeeee and the spectrummmmmmm 🤣
Could you get your hands on a B/W composite video security camera with a CCD and a video digitizer to use it as a webcam? It might cover a wide spectrum quite evenly.
Also, I don’t think you need to warn people against replicating what they see in this video, UVC germicidal tubes have enough warnings on them already.
I already have a B/W camera from the flat CRT door phone system, but I have nothing to connect it as a USB webcam. UVC tubes from ebay tend to have no warnings :).
@@DiodeGoneWild can you remove the Bayer filter from webcam ? It probably produces this dips in the red part of the spectrum.
Would have really liked to see the light coming from a LED grow light.
I don't have any grow light, I have to get some...
Ir absorption becomes an issue as the wavelength gets longer than the biggest dimensions of the pixels
Searching hard, there have been youtube of others, some just place a delaminated DVD directly in front of the lens.
can you calibrate the sensors sensitivity with a knowen source for example a tungsten bulb?
usually a low pressure sodium lamp is used for callibration
@@jg6780 I don’t men calibration for wavelength peaks.
i mean calibration for intensity per wavelength
I think you need a webcam that has a manual mode with fixed exposure to do that. If you're serious about building one careful selection of the webcam and using an actual diffraction grating is probably smart.
@@lambda7652 So just use an incadescent lamp and Planck law formula en.wikipedia.org/wiki/Planck%27s_law
For each wavelength scale the sensitivity of point to the value which should be according to this law.
@@jg6780 Yes tarts in principle what i wan to suggest.
The questin is can the software do it?
and if so do you need to have a calibration for every exposure setting of the cam
Hi, where can i get the software?
www.theremino.com/wp-content/uploads/files/Theremino_Spectrometer_V3.1.zip
no entiendo, como me gustaria que este el subtitulo en español
Adam ya
Did you analyse your sleep lamp?
I think in the last video he did
@@MassimoTava Ah you're right he did, right at the end. Zero blue in it, good!
Jde hodne poznat ze jsi cech. Jsi nebo ne? :D
né, vůbec ;)
thank you i understand the whole concept but why you talk like that
Thanks!
Big thanks for your support!
@@DiodeGoneWild you have one of the best channels on youtube! :)