For those of us interested in optics, how about a spectrometer capable of measuring optical wavelengths?
This would be interesting for astronomy and (probably) other things. The build would require data acquisition and an optical setup which would be useful for other projects in the future (Michelson Morley experiment?)
Materials required:
I think that this one deserves a dual entry for category, as it fits nicely in electronics as well.
I worked in a laser lab in a former life that had a device we used to characterize light permeable media using a sensor that could sense a slightly wider spectrum of light than the visible spectrum. You could shine a visible light laser into it, and it would break the incoming light into a spectrum that extended on both sides beyond the visible light spectrum which was something like 300-700 nm. The device would characterize from say 200 to 1000 nm. The result was a graph showing the intensity curve at each point across this spectrum. It was derived via some version of transform, likely of Fourier variety, as was clear if you watched how it danced on the screen during live operation.
So the idea was, a physicist or engineer could shine the light through nothing and do a capture. The device software would allow you to use this as a calibration reference curve for your source. It had user selectable averaging, which was nice since there was movement from sampled curve to sampled curve.
After calibration, the user could place a filter of some sort in need of characterization betwixt the source and the sensor. A practical filter in need of characterization could be some lens of sunglasses. And it was fruitful to see how the lens would allow certain wavelengths of light to pass at various angles. I think it may have had polarization options as well, but I may have just made that up.
Now I think that you could use something like this sensor to do mass spectroscopy, but I could again be lying, as I am a tad rusty.
And yes, I have interest in Michelson Morley as well. I want to build an inexpensive (sorta) interferometer for commercial sales, something that can be shined upon a surface, and yield data concerning that surface such as Ra.
I think a key to some of the items we are thinking about is the CCD, and CCDs with potential wells that capture photons of differing wavelength on movable platforms, alongside some clever mathematics are a very powerful thing indeed. Mix that will a good understanding of lenses and a data acquisition expert who can code mathematics to massage the acquired image data on the fly in either software or hardware and you have a deadly mix. Untapped profitable applications abound in this space.
So yeah, I like the suggestion. I recommend steering far clear of Arduino solutions here. Go straight into your PC so that you have the full gambit of decades of software development available at your fingertips with no need to write oodles of code to accomplish a simple task. Many like Matlab, however, I would point LabVIEW straight at this and be done coding while the Matlab folks were still arguing about how to get it done.
Do your bring up and debug in LabVIEW with full blown PC. Once you have the secret sauce for your application, pull the PC data acquisition gear out of the way, and then interface using Arduino. Now you have a working alogorithm in the form of your LabVIEW code. And whats even cooler, now it appears that you may be able write your Arduino code in LabVIEW (but you won’t be able to copy the PC code into it, still have to use a new set of Arduino specific LabVIEW functions).
I think we should take your idea and go a step further… identify a hole in the marketplace we can exploit, take your physics knowledge, combine with electronics skills, and go straight for the profitable product. Are you so brazen as to believe that such development can take place in that well equipped electronic room? I am.
Dude, sorry for catching this late. We discussed a very similar idea last night but wanted to do a simple shoebox spectrometer from the MIT OpenCourseWare (big thanks to @wandrson) for an intro class to give people an intuitive and hands on feel to how optics and light work. For me personally, looking at optics arrangements and formulas on paper didn’t come easy but everything clicked and fell into place when I saw the spectrometer components in action during the Raman build.
I can see several benefits and applications for DMS members from having an optical setup and work bench and even just benefiting from a simple visible light spectrometer.
Dude!!! You’re awesome! What you say about the expertise needed and marketplace is dead on! There’s so many little niches in the market for a team of persons to carve out for themselves that a little ingenuity and drive can carry a team of people a long way at Dallas Makerspace!
We’ve been going off this Hackaday Pagefor the Raman build and have acquired a lot of useful knowledge from it! I’d be really curious to get your take on using the Toshiba TCD1304DG Linear CCD Array Detector in conjunction with the STM32F401RE 512K for reading out the detector. The site’s really messy though, if you’re interested and I can help you find something please let me know! We’d love to get your expertise and help with the Raman Spectrometer Build and any other Optics projects people might be interested in.
You may be interested to know that my professional engineering career began at National Instruments in 1999 in their IMAQ (Image Acquisition) group. It was quite an honor to get an offer from NI, and even a greater honor to be placed in this group. Both of these items have contributed to wonderful success in my life. The product of these successes has me here at DMS now today, working to create entrepreneurship opportunities for those interested. And this optics space is of particular interest to me.
So here are a couple of things to note;
I have just began to roll out a business structure to include interested individuals in whatever profits we may be able to conjur with our ideas. Please check out 3e8tech.com. You can create a profile manually, then document your project idea or early notion now. You will have full editing rights to your postings on this site. I turned it on for folks to be able to access the blogs leas than 24 hours ago… How this business operates has yet to be determined, as we need to populate the membership with founding fathers, and then work together to decide the direction of a profitable group of makers through meetings, voting, etc. It is undefined yet, but since I have never worked as an officer of non-profit, I can work on the profitable side. My intentions are to share the profits with those who contribute, with seniority and level of contribution being key in determining the splits.
LabVIEW is powerful beyond measure for these sorts of optical projects we are speaking of. By Thursday, there will be a class on the schedule for folks to learn LabVIEW as taken from the top. Requirements for the class include having a laptop with LabVIEW and Device Drivers installed prior to first class (will be 2-3 weeks out). When you see this pop up on DMS calendar, get on the schedule quick, and lets get some folks some LabVIEW coding experience around DMS!
Thanks for the enthusiasm, and look forward to making some cool things happen!
Do you know what kind of accuracy you expect out of this? Eventually I would like a spectrometer for some of my work, but it needs to be able to take measurements at most every 10nm, at a relative accuracy of 1-2%…
The obvious answer at the moment is a $2000 trip to Thor Labs… ha
Device I saw in use at lab in the past:
Let me dig around online and see what the tool was that the military was using. It was just a USB device. Seems it should cost less than $2k in and of itself. Why do I want to think they paid like $1500 for the instrument. I saw the invoice for it. 1-2% seems very reasonable for that device, provided a good baseline was taken and the light source was nice and constant across wavelength. BTW, I do not believe this was a Raman device. It was something else. I don’t want to shoot down your enthusiasm, but it may be possible to get a device that can break down light into its spectrum for not very much. Tech has advanced a lot and $1500 was for USA govt over a decade ago. Is the interest to build something, or that we need one at DMS?
Hackaday Build:
But your setup should do the same thing. I don’t know for sure how well the CCD will work out, I’ll have to really take time with it to give a solid answer there. I wonder if it might just be easier to go with USB camera with large CCD. This way, no need for processor, its included in camera cost, and also no alignment concerns. Even if you have to push the CCD closer to slit, it stil makes sense as long as we get nice linear image across some row or column of CCD. Best to use monochrome, and we would need to refresh on which type CCD doesn’t have a lot of space between wells. But back to your question, uP will take some time to look at very closely. If these are the hackaday reocmmended devices, then your probably good without strenuous review. Otherwise, we should spend the time so we don’t have to waste cycles.
I like your enthusiasm. And I want to help. Being a LabVIEW guy, I can for sure help, especially when it comes to cranking out a nice GUI for the Spectrometer. And also anything custom we may wish to do with it.
So here is the deal: it will consume my time to help with Spectrometer, and while it is a labor of love, I would like to request something of you. You see, I am dabbling in entreprenuership, and have noticed that DMS is non-profit only. In fact, DMS position holders can’t be involved in a business using DMS facilities, as this would be conflict of interest. However, successful electronics business(es) that form within DMS and grow there initially are a feather in the cap for DMS.
While I am not quite sure where the divide is or even how to structure a small tech partnership, I have drawn up business docs and created a website to see if I can get something going. Check it out at 3e8tech.com. The blogs just went live like a day or two ago. So while there may be some pasting back and forth between DMS and 3e8 for a minute until I figure something better, I would like some help populating the website with project idea talk directed at making a profit.
The manual sign up works, facebook not yet. I would love to see some postings on that site concerning Spectrometer, and would also love to open some dialog there about any business ideas you may have that you would want to partner with other smart maker types to pursue. 3e8 will be a conglomeration of numerous good ideas developing in parallel, while the founders will enjoy the spoils. Think of the company as a way to team up on a few projects with a few great makers/engineers, and make some $.
So in addition to the Raman, what follow on projects look interesting to make $. I have the inferometer idea noted there. I wonder can we find build information?
Great question - I expect the spectral resolution of the spectrometer for the Raman to be 0.3 nm. I’m going off of the information provided by this great resource. We’re using a Toshiba TCD1304AP (3648 pixels, 8 um pixel width), 1200 groove/mm diffraction grating (spectral range: 400-700nm), and a precision mounted entrance slit (Entrance Width: 10 um).
The linewidth for the 532nm laser for the Raman spans from 531-533nm (or 2nm). If I’m understanding everything correctly, I expect the total spectral resolution for our Raman system to be sqrt(0.3nm^2 + 2nm^2) or… 2nm i.e. our spectral resolution for the entire Raman system is really limited by the excitation laser.
Here’s the BoM for our spectrometer:
532 nm laser from ebay: ~$80
Entrance slit from Edmund Optics: $117
80mm F-length spherical mirror from Edmund: $37.50
Diffraction grating from Edmund: $140.00
100mm F-length Concave Mirror from Edmund: $42.50
CCD detector: your guess probably better than Toshiba TCD1304AP
or $417.00 minimum.
Can you say more about what you’d like to use it for? If so, I can most likely point you to several resources.
Awesome! Thanks for the summary!
I would be using it to characterize light sources and color filters, and gathering data about the resulting colors – so the laser is not something I actually need. If I got the one from Thor I would probably use the cosine aperture anyway, and do the testing in a dark room.
The REALLY expensive item are the reference light sources (T, D65…), but fortunately I think I can do without on that 
You’re absolutely welcome to use any of the optics we already have for the Raman… sounds like we have most of what you need but you might have to come up with your own box… hehe. We’re using a ARM Nucleo Board STM32F4 STM32F401RE 512K run to a laptop for data acquisition but I really like @semaphore1999’s idea of doing everything in LabView and having a dedicated computer.
The diffraction grating spans from 400-700nm. I think this singular grating might be limiting to many other applications for us and I’d like us to purchase other gratings and components to construct other optical setups.
I absolutely want an optical bench for Science and want it housed in the coming sterile area to avoid dust. I’m unsure as to exact requirements needed for an optical bench for what we’re going after but I’d be stoked buying some pieces of aluminum like these and punching some well spaced screw holes into them for a start.
Mirrors: we got a concave & spherical mirror
Sources of known intensity and wavelength: we got a 532nm laser (and 4 pairs of safety goggles) and it seems like we’re always coming across excitation lamps or we can buy more
Disclaimer: I’ve mishandled some of the mirrors and the grating but the spectrometer produces a well-defined visible wavelength spectrum across the detector… even in average room lighting!
(pic of this visible spectrum to come once I recover some files off a crashed laptop)
If 400-700nm is enough of a range (i mean… it’s possible), our spectrometer should work for your purposes once we confirm a spectral readout from the detector to laptop. Might be a good introduction before attempting construction of a more sophisticated system for cheap.
How much do those reference light sources cost? They’d be useful in horticultural LED characterization applications.
You already have the firmware written for the device?
How about PC driver to read out data across USB or better yet bluetooth/wifi?
As for my uses, they would be to make $, in any reasonable manner. Think sales of instrumentation equipment. Instrumentation market is a wonderful place to be, and the more innovative the better. I look for items that I cannot purchase easily online. And then look for the cheapest of what is out there and comparable. Sometimes, the most seemingly obscure items become extremely popular and useful. I know a gentleman in California who makes RF amplifiers. Well, his amplifiers have become popular through the years, and now his company has some 25 employees all making amplifiers that now the govt purchases. If you made one, you have the BOM (bill of materials) and you know how you did it. Can you do it again? How can you make it quick/easy/cheap to reproduce? How can you quickly test (automation guy is your friend there as well, think more LabVIEW)? A lot of places in USA have more $ than time so making one of these is not their charter. If they need one, they will purchase it, from somewhere. Oh, look, there it is bwtek.com. They want you to call to get a quote. That means expensive. They need competition.
And since I am about $, now that you have one, why not offer services associated with the item that you now have? Can DMS_Labs perform the testing less that Thor and be profitable?
In order for your build to be extremely useful, we have to know for sure what the light falling upon a given ccd well or range of wells indicates. As such, we need a reference of a known wavelength. Come to think, possibly two references. Now I am getting confused as to exactly how this is working due to the fact that you are planning to work with a laser of specific wavelength. So what would make the light emitted by laser a different wavelength, like ever? I see that these are used for spectroscopy, but how does it extract the nucleual mass? …rusty like i said…
But back to Raman…that looks like a very long linear CCD. I doubt one can easily find a camera with CCD that long. And it is cool to have processing power right there, especially with wireless capability.
Oh, and look, the oompetition was nice enough to list out all of the applicable industries for such a thing… how nice of them…
This is the competition for a basic spectrometer:
http://www.edmundoptics.com/testing-detection/spectroscopy/laboratory-spectrometer/3643/
Apparently the standard illuminants have disappeared from the internet… because I’ve looked at them but can’t find a trace of them right now 
I thought Thorlabs looked familiar!
This is the unit I used in Navy lab… OSA201C
And the same software, which, BTW is callable via ActiveX.
A suggestion for calibrating the spectrometer…
Would burning a hydrogen flame give you something to align and scale the spectrum along the CCD?
I’ve never messed with a CCD before, but why are you using the STMF4 dev board? Is that so you can productize it easier later? That thing can support all sorts of things including USB, SD cards, and even Ethernet… so just curious.
For early prototyping it seems a little over-complicated? But if it works for you it works for you!
Get the thing working first. You have the one laser now?
Also, guess I was wrong on price tag for that spectrometer. Thorlabs wants $25k. Theirs will have lower noise floor. But I bet most places don’t need such a low noise floor. I’d bet they would appreciate inexpensive. Just thinking aloud…
But yes, once working I would think anything of known frequency and narrow band would work.
You know what DMS need to come up with? A laser lab with the sort of optics table that the Navy had in that lab I worked in. The table has a grid pattern of screw holes, and evidently these holes are all sized similarly. And would you believe that there are all these vendors making instruments, filter and reflector stands with screw holes that fit in those holes?
So in effect, you can make anything on that table out of lasers. It turns out that there are endless experimental stuff you can figure out using lasers and such. I have an interest in LIDAR, which are those rotating things you see on top of google car and Boston Dynamics monsters/dogs/mules/whatever you wanna call them. They have multiple sensors and moving mirrors and can map the area around them fully. These sell for not one bit cheap, and it would be very profitable to make them.
Oh, Thorlabs makes those tables as well:
We are gonna need more room for science, electronics, robotics, and such…
Hahaha, you’re a hoot!
We have the firmware written for the device.
PC driver to read out across USB - yes. wireless - not yet, but why not.
We got the BoM… the main part of it is in a reply to Daniel totaling $417. Then we got an optical notch filter, usually like $600, so now BoM is ~$1017 (we got a generous donation/loan of 2 of these though). Then $75-115 for a beam splitter or an ??$$ interferometer, $100 for an optical window, $100 for electronics, and $100 for housing. So let’s call a moderate estimate at $1500. We can save ~$3-400 by switching the notch filter for two band pass filters… I grimace at the thought of it though.
I think the Science Committee can offer analysis services and charge outside companies for them (it’d be a very nice contribution to the overall DMS budget). Now I’m no fancy big city lawyer, but seems to me if we were ever to offer a bad analysis that costs a company money, then I think we it’s possible that we open DMS up to liability issues. It’s still a possibility on the table for now as far as I know that we can look into further.
We almost have a finished Raman spectrometer but not yet. The next main step is to construct a vacuum chamber around the spectrometer with an embedded optical window and then test the readout of the detector. If all’s good, then we go for our first big milestone, proof-of-concept Raman spectrum acquisition. After proof-of-concept, we pause, and discuss aims vs. costs for the next build. I’ll be up there early tomorrow evening if you and/or anybody else wants to stop by for an overview to the Raman.
Pots of Gold
Biology is the Science of the 21st century. Most other areas of Science are past their Renaissance golden eras. However, materials still have quite a ways to come and looks like some improvements can be made on the space exploration front. I like your strategy and it makes since for quick, easy, and continued steady flows of income and I’m all for that as long as it’s also in line with something trying to establish itself 1-2 years out in front.
High-throughput screening analysis for biological research is more and more becoming the norm in place of targeted, strategically planned research and hey why not? This principle can also be applied to materials research and material processing too, check out my post here on using CRISPR for exploring new possibilities for nanomaterial processing. I’m talking stuff like running 4096 well plates in minutes and setting up experiments with >1,000,000 samples… neural nets - check… data management - you betcha. We’ll yet find that antibiotic that kills every last one of any bacteria that pisses us off.
Raman can find surprising, game-changing, 5-yr plan disrupting, niche applications in conjunction with a modifying or adjoining technique like SERS or TERS. If you read the Wiki on SERS, take note of the “single molecule detection blip”. It’s these detection ranges biologists are often looking for. So, take an analyte-detecting enzyme linked to a SERS active surface and throw it in a 4096-well Raman spectrometer plate reader and you just drank somebody’s milkshake.
Yea, I’m wondering for the time being what we really need from an optical table. I think some screw holes drilled in some 1/2" aluminum would be a great start in place of compressed air balanced surfaces… which I’m pretty sure are thousands of dollars.
