Hackaday Prize Finalist: An Un-noodly Spectrometer

And so we come to the final finalist bio for The Hackaday Prize. In only three days, we’ll know whether [fl@C@]’s RamanPi Spectrometer or one of the four other projects to make it into the finals round will be making it to space, or only Japan.

There are a surprising number of spectrometer projects out there on the Intertubes, but most of these setups only measure the absorption spectrum – literally what wavelengths of light are absorbed by the material being measured. A Raman spectrometer is completely different, using a laser to illuminate the sample, and measuring the scattering of light from the material. It’s work that has won a Nobel prize, and [fl@C@] built one with a 3D printer.

Bio below, along with the final video that was sent around to the judges. If you’re wondering who the winner of The Hackaday Prize is, even I don’t know. [Mike] and a few Hackaday overlords do, but the rest of us will remain in ignorance until we announce the winner at the party we’re having in Munich next Thursday.

What was the ultimate inspiration to create a Raman spectrometer? 
You say you needed to do spectroscopy for another project, but
why did you  choose Raman spectroscopy? There are several other
photoemissive spectroscopy projects out there. Is it just an
issue of being able to scan everything, or just wanting to a
project for the hackaday prize that replicated work that won a
Nobel prize, or something else?

I’ve been working on my larger project for about 5 years now.. It’s not an open source project unfortunately, maybe some day.. So without going into too great of detail, it worked out that I needed to determine bond angles and Raman shift in a sort of before and after scenario, as that would indicate if I was on the right path while testing. That naturally led me to a Raman spectrometer.. I didn’t have the funds for a used one and the larger project depended on it. I hadn’t seen any other projects out there that provided this type of information, so I just decided to build it. Nothing as glamorous as seeking to replicate any Nobel winning works. :)

You're building complex optical paths with a 3D printer, and 3D
printing is obviously a 'good enough' solution for building a
prototype. Given unlimited funds we know you wouldn't be using a
DaVinci printer, but would you still use filament-based 3D
printers if you were to do this all over again? What problems did
you encounter in printing the spectrometer?

Constructing it with 3D printed parts seemed like a fast and inexpensive way to build it. When the contest came into the picture, the idea of sharing the plans and making it easier to build came more into focus. Keeping the 3D printed parts seemed like a very logical way to make it easy for people to build since more and more people have access to 3D printers than other methods. So, I think to do it all over again I would still keep the 3D printed parts. It seems like the best middle ground, giving the maximum availability to people who want to build one for themselves.

3D Printed optical paths

Having said that, given ‘unlimited funds’… I might choose machining the parts from aluminum, or possibly from some less expensive material like nylon or HDPE (not sure how HDPE compares to aluminum in price/lb though). The biggest problems I encountered in printing were compensation for the ABS shrinkage, the obvious printer issues, and what to do with the countless pieces that had minor mistakes or errors. Shrinkage took a little while to master. This was my first actual project using a 3D printer.

Keeping a local database of spectrometry is insane, and in your
documents you say you're using online spectrometry databases. Is
there an issue in getting the data from these databases into a
coherent format, and what does the future of these proprietary
databases look like, given that the RamanPi will eventually be
released into the wild?

Getting data from the online databases in a coherent format isn’t really an issue. I believe most of them are in comma delimited format anyway, or something similar. I think most of the databases are maintained by big companies that charge money for access to the data. That’s where I see the biggest issue. There are a couple out there that are free, and I’ve gotten permission from them to access the data so long as it isn’t downloaded in full. Keeping a local database is kinda crazy, and that’d be a full time job just cataloging all the spectra from whatever materials you could find. It would be nice to see the future be a sort of peer to peer database where people share their data and contribute to it sort of like a Napster for spectra!

Have you given any thought to the commercialization of your
project? How much are you looking at to turn this into a product
(if you even can), and from the responses, what do you think a
product based on the RamanPi would cost?

Originally, I gave no thought to commercialization. I was just building it for myself. Then, after the contest began, and people started to take an interest, asking if I was planning on selling kits, etc. I’d like to offer a couple versions. I would like to have a low end version with less expensive optics and a Raspberry Pi camera module, a middle version with less expensive optics and a CCD.. and a more expensive version with good optics and a CCD. I would also like to offer just the spectrometer without the Raman system for use as just a regular CCD spectrometer. As for cost, I think the lower end should be around $300, middle around $500, and the high end around $700 – $800. The spectrometer itself being the most expensive part would probably be around $350.

Hypothetical, and we’re not going to hold you to whatever answer
you give. You win the grand prize, a trip to space or about
$200,000 USD. Which one to you take, and what is your reasoning
for doing so?

So, I thought long and hard… If I were to be lucky enough to have to make a choice between space and money… I originally said space, without a doubt. It was a done deal for me.

I think that changed after looking at the options for travel, and this being before the recent accident. I just didn’t trust the Virgin Galactic option. I am all about automation, removing the pilot from the ability to make errors. Being a pilot, I know where this can be a factor.

A normal-looking Rutan design. Not shown: variable geometry sideburns

I’ve also never been a huge fan of Rutan’s work. I didn’t care for the others because I haven’t seen the numbers..and this is my life after all. The idea of taking a trip in the Virgin’s SS2 doing mach 3.5 sounds more fun than space to me. But at this point… I would have to honestly say the cash option. Investing the cash into my larger project will probably pay for a trip later.

17 thoughts on “Hackaday Prize Finalist: An Un-noodly Spectrometer

  1. Aside from the annoying robot voice on the video, it does an awful job of explaining what the hell this thing actually does.

    The article here at HaD doesn’t do much better. Focussing more on the history/materials etc and not giving a simple explanation to what this thing does, how it does it, what you’d use it for, etc.

    After some Googling I feel a bit closer to understanding it. Maybe you can improve this a bit Brian?

    1. Thanks for bringing this up. Glad to see I’m not the only one who finds the understanding and explanation of the ‘science’ side of things a bit lacking.

      Contrary to the explanation of ‘measuring scattered light’, there’s ALOT more going on here.

        1. What I hate is that it only takes 15 minutes to do BOTH, yet it seems like no one could be bothered. Hey, if you don’t know the science, that’s cool. Not everyone is an expert.. BUT.. I thought this was the idea behind having a community of people who DO know this kind of stuff. If you don’t have an expert, FIND ONE. ;)

    2. For posterity, I figured I’d post a short explanation on Raman spec here. A little disclaimer: I am molecular genetics technologist, and more familiar with flow cytometry than Raman spec.

      From what I understand, Raman spectroscopy is a technique based on the inelastic scattering of monochromatic light. In other words, the frequency of the photons from the light source will change once they interact with a sample. Comparing the frequencies of these re-emitted photons to the initial photon light (called the Raman effect or Raman shift) can yield information about the molecular composition of a sample: its structure, moiety, orientation, etc. As Brian mentioned in the article, this differs from how IR spec works, which measures the absorption of photons by the sample.

      This can be used in chemistry, forensics, and biochemistry (Raman Spec seems to be making its way into flow cytometers). Raman spec can be used to analyze a sample in any state (gaseous, aqueous, solid), as well as use water as a solvent. IR spec cannot be used when water is a solvent and gaseous samples can be very difficult to analyze. Unfortunately, Raman spec requires expensive instrumentation – so a homemade one would certainly be a boon.

    3. The much higher-level explanation is that if you shine a laser onto an object (for example, a green laser), it will cause that object to emit light back at you (for example, different wavelengths from green to blue). The specific pattern of colours that’s given off is unique for a given compound, and so if you record it and compare it to a library of known spectra from different compounds, you can begin to tell what that object is made out of.

      One of the challenges is that the light that’s emitted back is about one million times less bright than the intensity of the laser, so you can’t see it with your eyes, and need extremely sensitive spectrometers to detect it.

      Correct me if I’m wrong fl@C@, I think the video is just showing a normal absorbance spectra instead of a Raman spectra (and it doesn’t mention what the compound being tested is), but the project logs show the first Raman spectra displayed on a scope, which is very cool. It’s very technically challenging, and this is being done quite inexpensively by comparison.

      1. This video from my understanding was for a non technical audience….and was supposed to be a sort of commercial… And I could have done better, I will not argue… The first two videos explain more in detail of what’s actually happening..

        The lower half of the details section explains how the system works…(after you click read more) http://hackaday.io/post/2978

        There’s Rayleigh scattering (the reason the sky is blue, etc) and Raman Scattering… Raman scattering happens based on the vibrational and rotational states of the molecules that the monochromatic light is hitting.. Peter is correct, based on those states…the light will be shifted in frequency and be reflected along with a ton of the rayleigh scattered light..You have to filter out the rayleigh light to see the tiny amount of raman light returned.. But that light will provide useful information about the material and it’s state…

      2. Thanks for the easy to understand explanation Peter!

        A high level explanation is good for those of us who aren’t molecular genetics technologists. Then we can research further if we wish :)

      3. @peter
        I’ve taken enough physics courses to have a general understanding of how spectroscopy works, but it still feels like voodoo every time scientists determine the atmospheric makeup of another planet. We live in an era so interesting that you need a panel of people with PHDs to summarize it for you.

  2. This is a very cool project. Does anyone know if this could be used to detect lead in paint samples? Or asbestos in insulation materials? I know that small samples can be sent to labs for about $30, but it’d be really cool to be able to test for that stuff at home. Or determine all the compounds that are in my tap water?

    For the first task (detecting lead based paint) I had looked into X-ray Fluorescence Spectroscopy, which is similar and very cool. It seemed much harder to pull off safely in my garage though.

Leave a Reply

Please be kind and respectful to help make the comments section excellent. (Comment Policy)

This site uses Akismet to reduce spam. Learn how your comment data is processed.