A CD Spectrometer (2006) | Hacker News Reader

A CD Spectrometer (2006)

cs.cmu.edu

215 pointswcrossbow48 comments

I recently got a spectrometer with a fibre input from fleabay, which I've been enjoying playing with.

I got a tiny ruby stone and shone a tungsten halogen light behind it - https://www.anfractuosity.com/files/ruby.png I could see a dip around 694nm like someone else's spectrum - https://en.wikipedia.org/wiki/Ruby_laser#/media/File:Ruby_tr.... This is the same wavelength that ruby lasers emit light apparently.

I'm planning on using it to try to measure ABV of beer/spirits. I noticed some of the cuvette holders with fibre inputs seem to be crazy expensive though, so probably need to DIY that!

A while ago I got a little spectroscope that looked a bit like - https://shop.wf-education.com/science/op66595.html. And simply attached to a camera, to generate graph from.

A $500 DIY near-IR spectrometer that would sell for $10,000

https://caoyuan.scripts.mit.edu/ir_spec.html

Current discussion: https://news.ycombinator.com/item?id=37498142

Looks very impressive!, thanks

Hey, try heating up that ruby! It'll turn darker as the bandgap moves toward the IR. You can shift your dip with a hair drier.

It should move from ~1.15eV at room temp 75F/25C to ~1.10eV at 120F/50C, which should be about 35nm longer.

Ooh, cool! Will definitely have to give that a shot.

Use liquid nitrogen and you can get it to turn light pink!

What sort of wavelength of sensor do you need for ABV measurements? Size of ethanol particle? Half it because of Nyquist? Or what does it have to relate to?

I'm just wondering how fancy vs. DIY you need for that sort of purpose basically.

This is the paper I've been looking at - https://www.ncbi.nlm.nih.gov/pmc/articles/PMC3356901/ they make use of a tungsten lamp and IR leds, I tried with just a tungsten lamp and only really saw a plateau around 900nm when ethanol was present. Their spectrometer goes further into IR than mine, mine tops out around 1000nm, but hopefully I'll still be able to get some results when I add IR LEDs.

I'm not sure what the peaks correspond to regarding the ethanol molecule though afraid. It's been a while since I read the paper, but maybe that indicates this.

I believe my spectrometer just has a linear CCD sensor.

For ABV you'd usually use refractive index, there are low cost non-electronic refractometers, you can also use density. Alcohol doesn't absorb much light so generally not done by UV or IR.

FYI, the ruby.png image is 19200 x 14400 pixels big. Luckily it's mostly white.

Thanks for the heads up, just regenerated the graph

Aseq? I've had good luck with these for a few applications.

Oh, for a while Public Lab's Paper Spectrometer [0] was very popular in "citizen science" workshops or forums. It was made peeling the diffraction grating of a DVD-R (not a +R) disc.

It's a fun thing to do in an afternoon.

  0: https://publiclab.org/wiki/papercraft-spectrometer

At my current company we build satellites with imaging spectrometers as payload and wanted to propose building a bunch of this on our upcoming team day. I was hoping for the non-technical team to get a better feeling of what is that we are trying to sell.

Can you give a hint without doxxing yourself?

I work for Kuva Space

P.S. Doxxing me is trivial based on what is already available in this forum

Nice, hyperspectral imagery has some very interesting applications. Cheers!

In my high school physics class we measured the pitch between the tracks of a CD using the same principle. You just need a light source with a known wavelength, like a laser. It was a pretty cool experiment!

That sounds a fun project :) I tried a little peeling of a CD-R under a microscope, where can see the tracks, I should try and measure the pixel distance and convert to track width - https://www.anfractuosity.com/files/cd-r.JPG

what microscope are you using? i've been itching to get into microscope photography for a new hobby, but because i know myself, it'll be a fun but expensive rabbit hole. i've been deliberately putting off on researching because i also know myself and will be just as likely to be shopping than researching

It's an Olympus BHM that I got cheaply from ebay. It's a trinocular metallurgical microscope, that I added an old SLR to, for that photo. I noticed the semi-silvered mirror seems rather scratch though, so might need to look at replacing that.

are you attaching the camera directly with an adapter, or is it projecting on to whatever lens you are using? i guess i'm coming at this thinking of it in terms of astrophotography using T-adapters to connect in place of on eyepiece vs taking an image from the eyepiece

In the trinocular port, there's an eyepiece, which I forget the magnification of afraid. And to that port I added this - https://www.alanwood.net/olympus/photomicro-adapter-l.html and then used an OM to EF mount converter, which the Canon camera directly attached to.

I've since bought a microscope imager, that provides an output over USB. But I think I need to remove the OM converter, and get a shorter adapter to successfully attach to the microscope C-Mount.

Thanks for that. I was hoping that would would be possible. I think I now know what to ask Santa for xmas!

Either works, but trust me, the trinocular part is far superior, ergonomically. You do need to know whether you want a metalurgical scope or a compound light transmission scope. The former is good at looking at opaque samples, the latter for transparent samples (biological stuff mostly).

that is something i have considered, but not known what terms to use to look it up. i would love to be able to look at opaque things in a highly magnified way, so the lighting issue is something i had in the back of mind. i do like the biological stuff in hopes of possible timelapse to show growth.

This is what I have for general opaque and transparent stereo viewing: https://www.amazon.com/AmScope-SM-4TZ-144A-Professional-Trin...

then I get this illuminator: https://www.amazon.com/AmScope-LED-6W-Powerful-Gooseneck-Ill... for top-down viewing, and it can be adjusted for bottom-up illumination.

Then you can buy samples- microscope slides, petri dishes and samples,f rom places like Carolina Biological.

You're my hero!!! I hear Santa's sleigh bells now!

Enjoy. Personally, I buy tardigrades and algae and a petri dish and look at that. Rotifers are cool too, and both can be collected from the natural environment. But, I also use the same scope to look at PCBs and other small things. After a while I got tired of moving the sample around so I motorized it using CNC technology (g-code, stepper motors, linear stages) and then trained an object detector so the scope can just follow the tardigrade around in an area much larger than its FOV.

if I have seen farther, or closer in this case, then I would like to thank all the giant shoulders upon which I have stood (seriously- look up ernst abbe, he's amazing, he created the 8-hour workday!)

This is an awesome picture! Forgive me if this is a stupid question but are we seeing 0s and 1s here? Is that what the black is? Kind of the equivalent to a morse code dit (0) and dah (1)?

That not a stupid question. It's a bit (ha!) more complex. These are called pits and lands.

> The pits and lands do not directly represent the 0s and 1s of binary data. Instead, non-return-to-zero, inverted encoding is used: a change from either pit to land or land to pit indicates a 1, while no change indicates a series of 0s. There must be at least two, and no more than ten 0s between each 1, which is defined by the length of the pit. This, in turn, is decoded by reversing the eight-to-fourteen modulation used in mastering the disc, and then reversing the cross-interleaved Reed–Solomon coding, finally revealing the raw data stored on the disc.

https://en.wikipedia.org/wiki/Compact_disc#Physical_details

These encoding types are used to improve the "tracking" of the laser head, and to keep the timing consistent. We want the medium to regularly have changes between pits and lands to synchronise the timing and speed of the disc. The encoding scheme enforces this.

https://en.wikipedia.org/wiki/Non-return-to-zero#NRZI

https://en.wikipedia.org/wiki/Eight-to-fourteen_modulation

I missed this until now, I really appreciate you tracking down links and sharing this info! Thank you!

It's wild it's as complex as it is! Though it makes sense it seems way more robust

What's the pitch between tracks?

1.6 μm

Ah, I see. Multiple meanings of both "pitch" and "tracks" on a music CD.

Absolutely and utterly brilliant!

I would have never in a million years thought of this use for a plain old CD... but yes, after reading this article, yes, I now see how a CD Spectrometer could indeed work!

Again, absolutely and utterly brilliant!

Upvoted and favorited!

I've been checking the quality of my light sources for years using a CD (florescent vs LED vs incandescent). Didn't DIY'ers know this technique for years?

I’ve been wanting to check my lights and didn’t know about this.

A few years ago we had several cans replaced with solid state LED fixtures and my wife has had a lot of trouble with sleep which roughly correlates. It may be placebo, but it seems like she also may get better sleep when watching an OLED TV vs an LCD iPad. It made me wonder if we should put 460nm filters on our ceiling lights.

How does that work? Are you using a CD to tell if one fluorescent bulb is better than another? Or to tell if one bulb is LED or incandescent or ...?

> Didn't DIY'ers know this technique for years?

I used to live in a rural area (fixing everything in the house by myself) so I'm quite the DIYer but I don't know anything about that!

Using a 0.2 mm slit? Or do you have an easier way that uses just the CD?

First I've ever heard of it, and I went pretty deep into the rabbit hole of spectrum when I switched to LEDs in my home. Don't think this is very common knowledge.

just hold the CD at an angle to the light source and stare at the reflective surface

And here I was, thinking this was about circular dichroism!

Cool article nonetheless..!

same! I was excited to see something so esoteric from my grad school days come up

Constructing a spectrometer using a CD - https://news.ycombinator.com/item?id=10285117 - Sept 2015 (9 comments)

Does using a DVD or Bluray instead of a CD improves the result? The angle might have to be different?

Here's a video showing the construction of one based on the same principle, but using a camera for wavelength measurement:

https://www.youtube.com/watch?v=p3MzQ1OF3lk

Oh my word, that's awesome!

Reply on news.ycombinator.com