How to build a MINIATURE Spectrometer for the Raspberry Pi
Vložit
- čas přidán 19. 06. 2021
- Episode 27
#spectrometer
#raspberrypi
Code at the end of the Description! Check out my other videos: / leslaboratory
Please don't forget to like,subscribe and comment for more great content!
If you found this content useful, and would like to support this Channel, please consider supporting this work on Patreon: / leslaboratory
Or donate directly: paypal.me/leslaboratory
Alternatively, please share this content on your social media platforms, it really helps me out!
This is a follow up video to Episode #20 where I built an inexpensive Raspberry Pi Spectrometer: • How to build a DIY Ras...
In answer to comments about miniaturizing this, I build a tiny lightweight Spectrometer, suitable for adding into smaller projects such as Tricorder projects, or maybe even mounting on a Drone.
As an aside, I also look at the spectra from some Different lamps (as well as Ultraviolet) , including: Fluorescent, Mercury vapor,, Xenon, and Deuterium!
The pocket spectroscope used in this build, was a kind donation from Paton Hawksey: www.patonhawksley.com/product....
This uses readily available components and easy to use Python software I have written specially for this application.
All code for this video as well as additional information is available at my GitHub:
github.com/leswright1977/PySp...
#RaspberryPi
#Spectrometer - Věda a technologie
Being from a low income country, I will be building this small version with the aim of using it to teach students in spectroscopy and if possible use it in research for identification of elements.
Thanks a lot for this.
Awesome! Glad it helps. Science should be in reach of everyone!
When I think what the spectrometer we use for in reaction measurements coasts, and then I look at this little gem...
This is red hot wonderful continent! Truly what CZcams was made for! Imagine having access to stuff like this back in the 70s or 80s.
Thanks again for making this project!!!
Personally I would love to see a high resolution more traditional version. Possibly something based on an ocean optics enclosure which can be had with the slit and grating for under 50 bucks. Or maybe a 3-D print body matching the ocean optics bench geometry. I think the files have already been created and are extremely accurate and forgiving. I’ve tinkered with and re-calibrated over a dozen spectrometers of that form factor… And was able to get ridiculously accurate results in a hobby optics lab setup.
Would be awesome to see what you could do with a setup like that. With optical grade glass diffraction grading instead of plastic film. And a fully adjustable/customizable optics bench in that smaller form factor.
Thanks again for this wonderful series!
Here’s a great example. Check out this listing and this guys other listings.
www.ebay.com/itm/143933041252
This is the optics bench including everything except the sensor… or for under $50 you can get the optics bench and mirrors without the grating. Or like I said previously… I’m almost certain several people have designed this and files are available to 3-D print. I think it’s peak accurate and customizable. Considering the optical quality of components that can be had for extremely cheap. It’s so easily adjustable and customizable anyone can calibrate this thing to within factory specs with a bit of practice.
The Guys Ebay page is a treasure trove. He even sells the linear CCD mount it on the board ready to go for about 40 bucks. But I have a stack of those linear CDs out of every kind of spectrometer and device imaginable… So hit me up if you want a sampling of them, and let me know where to ship them.
Disclaimer-I HAVE ZERO Affiliation with this guy or his eBay page. I’ve never talked to him or bought from him, so I don’t vouch for the service or quality…
Very cool....cheers!
This is FANTASTIC! WELL DONE!
Thank you very much!
The simple calibration in the software is just brilliant!
Thanks! There is 3 point cal in the new version, for a 3rd order polynomial fit, but it is just as simple!
Well done, sir.
Very nice little addition to its larger brother... thanks for another great video
Thanks Flavio!
Neat, some of the commercial micro-spectrometers are really cool, like the Hamamatsu C12880MA
This is really fascinating!
Thanks!
What a great video! :)
Thanks! :-)
Nice
Hi Les. Kudos on your mini spectroscope and the covering software. I have built one in PLA and it works very well.
Can I perhaps make you a suggestion for a small addition to your software in some eventual update to v3.1? Would it be possible to make the central white line, that I suppose represents the sampling horizon, adjustable up and down so one may make up for some degree of offset of the spectrum above or below the central field? This would allow one to sample at the visibly best slice of the spectrum. Thanks again for your excellent and instructive videos!
Very nice build, the aluminium makes it look very professional.
It would be interesting to integrate it into a standalone bit of testequipment.
Something with the formfactor of a small oscilloscope maybe.
You could even integrate the screen and UI ... and maybe even two sources for an "auto calibration" function.
I would propably do it this way, if I only had a use for a spectrometer. ;)
Greetings,
Michael
Cheers!
Yea, a compact build would be cool. Yep, an integrated laser would be cooler. the suppliers of the Hamamatsu breakout boards have a 405nm diode mounted on them for fluorescence experiments as well I believe.
Great build! Are you using the version of the raspi camera without the IR filter (NoIR version)? Do you think that the pocket spectroscope will actually work for up to at least 900nm? Thanks for sharing!
you do it?
Absolutely love what you have made.
Are you ever going to use your Spectrograph as a Spectrometer. I looked at your Channel and don't see any attempts at making one. I would love to see the plots for materials around us.
See what components can be identified.
For plastic threads, put them on top of each other with just a bit of pressure. Unscrew until you feel it click, then screw in. Easier to feel with coarser threads, but...
Great tip, this is also what I do. Plastic mounts are still a bit poor though, easy to strip and/foul the threads.
Man, I totally missed this video. Qosmos Chem YT channel video was suggested by YT when searching most recently and he mentioned your video which I assumed was of the larger diffraction grating spectrometer. Didn't realize you also had this gemologist mini handheld version as well. Neat, good fun.
Cool! I hope this is useful! Certainly less expensive than commercial units!
How good that really is? previous vs versus another build. Is there somewhere captured "raw picture" or video directly from raspberry camera available? Having library of captured reference video clips from various light sources would be nice thing for unit testing software etc..
Do you think that it is able to see more UV; for example, using an 8mm lens instead of the 12mm? It might able to see further down the UV range? I doubt you will see below 300nm, but I'm a bit curious if you can see the 313nm peak of the Hg lamp. Thank you for a great video!
Great project as always!
Do you think a 6-22mm zoom with F1.6 is good for a spectrometer? Or maybe it's better a 2.8-12mm?
That's what I have used, but feel free to experiment!
Program the software to read corner-to-corner instead of just horizontal, and tile the spectrum accordingly - that will give you 20%+ additional bandwidth/resolution
I have several UV filters… They can easily be found online… Apparently they transform deep UV into visible light. Might be a way to extend the range down into the 265nm range
I gathered them a year or so ago when the first UVC LEDs were showing up on the market. To help convert, verify, and measure the light. It worked a treat!
They won't be filters, they are probably fluorescent targets.
@@cambridgemart2075 I definitely have some fluorescent targets in there… But supposedly these are filters. Similar to those found in all types of track lighting. “UL UV FILTER 001”
I don’t know if these do what they claim… But they sure are everywhere, and there are several patents and ul listings regarding them. They claim to convert UV light into the visible spectrum or block it all together.
images.app.goo.gl/Ah3p3DLzv1zxUuqU9
Not sure about the claim of converting Vis to UV, that would be a nonlinear process. But sure, the internet is abundant with visible cut filters, that allow only UV and IR though. Woods glass is by far the cheapest. I have some in, because I want to see what the Picam can see in the near UV.
@@LesLaboratory sorry I meant to say they claim to convert UV to visible light. I’m not sure about the claim either… I can totally understand UV blocking filters but to convert UV to visible light without phosphors or other targets… Sounds fishy. But there are a lot of UL listed “ conversion“ filters that can be found on lights in most home stores. So who knows.
Woods the glass is great stuff. Was Lucky enough to stumble upon on a college optics lab auction bundled in with a MACHINE SHOP auction. That’s where I got my optical breadboard stuff and it had all kinds of wonderful UV and visible light/ir filters. And I regularly by visible light filters when building some of my custom hi power UV flashlights. To remove any visible light and only allow the wavelength desired.. And thanks to the new production of ultra high power UVC LEDs… I’ve had a lot of cool flashlight projects this year. This new UVC led Stuff is wicked crazy….
I stole an acebeam X80uv from eBay last year for under $100. Considering they go for over 500 I’m very lucky. Has 16 of the most powerful UV LEDs ever made. And inspired many of my own UVLED flashlight builds.
Really enjoy messing around with it, I became interested in it because I’m always needing a various wide range of UV wavelengths In my musical instrument repair business. To examine and locate previous repairs, cracks, or markings in several hundred year old violins and such. It requires a very wide range of UV wavelengths because each task is a bit different and reveals itself under different light
Also a big collector and hobbyist regarding radioactive materials and geiger counter/survey meters. Scintillation detectors in particular. So when hunting for uranium in old mines in tailing piles… High power UV really comes in handy!
Hii, I like your work much. Could you give me information about the resolution of your setup?
Do you have any quick guide/advice for how to operate these discharge dicharge/vapor lamps ypu used for testing? Will cheap ballasts work? I'm not finding many tutorial/guides and the high voltage less inclined to make a diy solution
I just use an old HV power supply from a plasma lamp. Cheap, easy to use and pretty safe (the worst you would get is a small shock and a burn or two)
Hi! The Lenses I've found in ebay seem to have an IR filter in the bottom? Did yours have one too? Not sure if it's removeable :)
Hi Les and thank you for another video. My diffraction grating has arrived from Paton Hawksley and I am now just waiting on some lenses. Being new to this, I am curious as to what/where you obtained the lasers you are using to calibrate. Are they simply over the counter laser pointers that are accurate enough to use, or are they something more than that? Thanks again!
Over the counter pointers will do, however be aware the wavelengths quoted aren't very accurate, with the exception of any DPSS lasers (Green laser pointers). these are always bang on 532nm.
It is possible to calibrate using a fluorescent tube though. See: github.com/leswright1977/PySpectrometer
@@LesLaboratory I have ocean optics spectrometer.can i use deutrium lamp or halogen lamp for its calibration for UV range specific? their light source devise are too expensive.
@@nerd9992 Sure, any light source that emits discrete lines will do, so stay away from LED's and Halogens (they emit contunnua), and use old style Fluorescent tubes. If you want to calibrate UV, you can get away with Cheap UV-C lamps (Just look up the emission lines for Hg as a reference) mind and protect your eyes from UV though. Corneal burns are exquisitely painful (voice of experience from a welding incident!)
@@LesLaboratory 🧡🧡🧡 thanks Doc!
What are yo running your deuterium lamp from? I have one and I tried using my variac with a ballast resistor and could not get it to run on that. I can sort of power it on my HV PSU but I don't think I am getting it to it's full current with what I currently have set up. I would love to have a solid PSU solution for my lamp.
Same as you, I just powered it from a small HV power supply from a plasma lamp. It's not something I have had time to look into at the moment, too many other projects on. :-S
So how does it detect iR spectrums, and does all lasers have a bit of iR leaks?
Do the spectroscopes contain glass optical grade diffraction grating? Or are they plastic film gratings?
I’ve looked but can’t find clarification on this anywhere online, nor a tear down video
I would say almost certainly plastic film given the transmission style design.
We don't use grating film for the spectroscope as it is not of high enough quality for resolving fine spectral features adequately. We use optical resin and high precision original ruled diffraction master gratings and copy them directly onto the glass with the optical resin, the linear spectrum is centred using a prism and the slit is brought into focus with a lens. Hope this helps! :)
@@rampantmoss4528 Awesome! Thanks for the reply, that clears that up!
@@rampantmoss4528 that’s incredible thank you for the details! I come from a long history of resin and mold making. Grew up around the stuff, my dad was experimenting and searching for a patentable material to be used as a synthetic ivory, mother of pearl, and stag. And eventually he succeeded, gaining many of the major knife, gun, and other manufacturers needing those Materials.
As well as pioneering the process of stabilizing wood and other materials with resin in a vacuum/pressure chamber.
Good Lord some weeks he would take delivery of two dozen 50 gallon drums of resin. So grateful for those experiences…My entire life until about 20 years old revolved around that stuff.
Really appreciate you taking the time to let me know some of the details. I would really enjoy seeing pictures or video of your set up. You would make incredible content!
can you apply it on an apple, to check the quality if fruit
can you do a demo how to calculate SO2 in tonnes/day using that spectrometer? TY
This is a great project. Could you give any details on the pocket spectroscope? Particularly the optics involved. Is it as simple as a prism or is there a grating in there so where? Thanks. Doug
Thanks! Yes, these are diffraction grating spectroscopes, but there is also a prism to straighten up the optical path. Best guess the optical elements front to back are: slit > Collimating lens > Grating > Prism > Objective lens
@@LesLaboratory any interest in doing a “tear down” ….. you know, just for completeness? :)
@@LesLaboratory I actually found a video where the instructor detailed these gem spectroscopes. Appears that there is a slit, collimating lens, split prism sandwich with the diff grating between two pieces of the pricing. Very nice
how far does it go into the infrared? It seems that you're cutting off around 700nm. is it a camera with IR filter or is there a different reason it cuts off? would a shorter focal length allow to extend it to ~1000nm provided there's enough light?
I have had it down as far as 800nm, but the spectroscope itself was designed for the visible. You could build your own spectroscope, there is no reason why the software and picam shouldn't handle it, though response might be a bit funky in the IR.
can you also measure liquids? using 2 ? measuring the wavelength of the source and then the liquid?
Yes, in principle. It depends what you want to measure. Assuming absorption, you would first take a spectra of your light source, then spectra of your sample in the path of the lightsource. You could then load these into matlab for example and subtract them from each other.
Hello sir, can I observe doppler shift with this system??
for playing with deuterium lamps (if not interested in the UV range), why not simply cover it with glass?
That would probably work just fine!
have a question , Can you with rasperry pi , build a laser in , so you Can meassure the colour on Coffe beans , realtime , when they are in the roaster. , so you Can stop it roast , when reach the right colour , i dont know much or anything about this , so have to ask the right person 😊thanks, i write laser , in one of your other video , but maybe it is this to use in the projekt i wrote about , very interisting Channel , so you Got a new sub from 🇩🇰thanks
Can you make your graph display farther into both ends of the spectrum?
Yes, by changing the lens, then calibrating, it can 'see' into the near IR and near UV, but not too deeply. the Spectroscope it uses as designed for the visible spectrum.
Is there a way to see spectra below 280nm (UV-C)? I'm interested in figuring out which UV lights have germicidal capabilities with your spectrometer :)
Sensors from GenUV should do it, but they don't give the spectral range in real time, just presence/intensity.
Not with the current spectroscope attachment. There is too much glass and plastic in the optical path, For UV work, you need fused silica optics. Additionally the cameras themselves absorb to much UV. This was part of the motivation for this video: czcams.com/video/y39UKU7niRE/video.html
Did you had the chance to look for "single array cmos" or "linear cmos array" on internet ?
At 7:15 it's showing some vacuum UV at 109nm. Is this an artifact of some kind? Does PySpectrometer work by looking at the distance between the calibrated peaks on the horizontal plane?
For sure, that will be light leakage from somewhere. There is no way 109nm is making it through the optics.
Would you please mind sharing sources of "calibration" lasers, you are using?
Most of the lasers he uses are standard cheap Chinese lasers that take 2 AAA batteries.
Considering all red, blue, and green lasers and diode‘s are going to fall within a very close tolerance of their stated spectrum. It’s a really good start. Then you can always branch out to more accurate sources like helium neon laser tubes, nitrogen lasers, deuterium lamps, and stuff like that. Which can be bought very cheaply if you watch the websites closely.
You can get a three pack containing a DPSS 532nm Green, 650nm red, 4045 nm violet/blue in a set for under $10 on eBay and even less on alli express.
The DPSS green laser will show peaks at 532 nm, as well as infrared from the pump diode. Very interesting experiments can be done and I highly suggest picking up a pack.
BUT Be careful!!!! They are NOWHERE NEAR five MW. The green lasers can easily reach 100mW and burn dark surfaces, and most don’t have an infrared filter meaning instant eye damage or blinding us faster than you can blink. Even from a reflection!!!!
DEFRACTION GRADING SPECTRSCOPE what is it made out of?
I wonder about the absorption of the lens? Are they plastic, glass?
Way back I speculated on the spectra of salt solutions, I think our collective opinion was incorrect. Sources indicate IR absorption lines midst the broad admittance in the near IR.
Anyone else verify?
Is this a prism-grating-prism setup?
Or does this just take the -1,0,1 order onto the sensor together and waste a lot of resolution?
Internally I believe it may prism-grating prism (never torn it down, though looking through the end at odd angles you can see the optics in there). The spectroscope is high quality educational toy, and the resolving ability for Fraunhoffer lines is really very good. That said, I do not think you would likely resolve the two Sodium D-lines with it, but truth be told I have never tried!
Could be possible to measure transmittance with this device? I am trying to figure out how to make a spectrophotometer that works on an specific wavelenght and measure transmittance of an electrochromic thin film.
Yes, you would just need to have a nice clean stable light source, take a baseline measurement without the sample, then a measurement with the sample.
Are you willing to share your parts list and Software? Id like to make one but dont have the programming skills..
The link is in the description, I open sourced it: github.com/leswright1977
Wondering: what is the magical power source to drive all kinds of tubes? Especially the Xenon flash tube, AC / DC ? V?
For this video, I just used the PSU from a broken plasma lamp that happened to be under my bench! That's it. It produces about 8-10kV high frequency AC, from 12v in, so no mains voltages and in the scheme of things, relatively safe, unlikely to cause serious injury.
I wonder its feasible to do this with an esp32cam rather than Raspberry pi. This way it could be even smaller, run from a small battery and be accessed via wifi.
Maybe, I have not looked into it. It depends on what libs are available in terms of grabbing a row of pixels, and processing and presenting the data. If I get time I will look into it.
Let me know if you do have a look at it as I now have an Arduino Sketch for esp32cam which grabs a grayscale image and extracts a single horizontal line as raw 8 bit data which should be a good starting point@@LesLaboratory
are electronic techs in the uk as underpaided as the one in the us?
Probably!
What are the specifications of the smaller spectrometer ? Did I miss that ? They are selling for half the price as your patron, elsewhere online.
It is linked to in the description. They are visible light scopes but allow the picam to 'see' from about 360nm to 800nm or so. Interesting, do you have a link?
Has anyone done any work with the AS7265X?
I wonder what is the limit of detection at 254 nm 🤔
Hello, can you send me a link to the xenon flash tube lamp you are using? And what kind of power supply do you use for it. The xenon flash tube lamps I found on the internet had 330 V output.
I don't remember off the top of my head, though in this application any old xeon lamp will do. I just used the power supply from an old plasma lamp to drive it, however this is not really the correct way to do it, as it will run very hot. That said, lamps are cheap!
@@LesLaboratory For our project, we are trying to find the right light source. So we decided that the xenon lamp is the one that shows the best spectral results. Yes, the lamps are cheap but the problem for us is the power supply. Do you know what voltage you used? We are currently using LED light source, do you think we should use xenon or should we continue to use LED light source.
14:13 how it is more dangerous than previous tube if that tube 's wavelength was lower than this ?
The spectrometer won't read lower than about 350nm or so, but the UVC lamp also emits strong lines in the UVC region as low as 254nm. At this wavelength the light will burn skin an damage DNA.
Kindly give a circuit diagram or assembling procedure for the same. Thank you. What is that software? Software link..Please provide. How to operate the software is not clear at all. Good work.
The links to the software and the requirements for the build is in the description.
Can it be treated as raman spectra?
How to place the sample? For detection
Can it be treated as FTIR or NMR
how can i make the aluminum case like you
The Case was just made from scrap aluminum out of the parts bin. The black portion was a drilled out Aluminium corner cube like this: www.amazon.co.uk/RanDal-Connector-Aluminum-Extrusions-Profile/dp/B083PSQ72F
does it work with pico?
It might work on the Zero, but the Pico is an entirely different beast. The same basic principles could be applied, assuming you can even get a decent camera to talk to the Pico, and you wouldn't be able to use the OpenCV libs on the Pico either.
"So this light gives you skin cancer and does permanent damage to your eyes.... We'll try it out in a bit!"
...alright. But all joking aside, nicely done, very interesting build!
Point your camera lower so we can read your shirts!
LOL Ok! :-)
Just a comment but you should take care physically handling that D2 lamp without gloves on; the MgF2/SiO2 envelope gets exceedingly hot, and the oils from your skin absorb the UV strongly causing localized hot-spots (The UV emissions extend down to the VUV but the envelope absorbs anything less than ~ 220 nm); touching without gloves will greatly increase the likelihood of envelope failure.
Also you don't see the UV because of the plastic/glass elements in the spectrometer, but ~ 60% of the output is a continuous emission between ~ 180 - 350 nm; the pink is mostly from the "Balmer Lines". Be very careful with that thing.
For sure. I am driving it well below its specified current rating for exactly these reasons. Corneal burns are deeply unpleasant.
Whoever owned it before me wrote on the envelope in in permanent marker (facepalm) , so it would likely explode if driven correctly.
do not expose bare skin to deuterium lamp...
...proceeds to touch it
:-D it was run at a very low current, so not too much in the way of UV to start worrying about. If I ran it off of a real PSU, I woudl not be in the same room as an exposed lamp like that!