CosmicPI: Detecting Cosmic Rays with a Raspberry Pi

Get your 4-layer PCBs for 2$ at jlcpcb.com/​​
Win Test Gear at Keysight University Live bit.ly/MR_KULive​
Miniware Hot Plate www.banggood.c...
Support me on / marcoreps
PCBite DMM kit www.welectron....
NASA Ultra High Definition footage www.nasa.gov/c...
Cosmic Pi website cosmicpi.org/
Cosmic Pi OHWR ohwr.org/proje...
Gigabecquerel on plastic scintillators gigabecquerel....

[@Maarrk]

Do you even realize how advanced of a hacker this man is to make THIS VIDEO'S ADDRESS START WITH "PCB"?

[@DrTune]

That's why each video takes so long to post, he has to keep uploading and deleting it until the URL is right

[@rubiconnn]

Nice, the chance of that happening in the same case are 1/262,144

[@John_Smith__]

Lol! Well he is sponsored by jlcpcb .... :)

[@forloop7713]

@@rubiconnn your math is wrong

[@rubiconnn]

@@forloop7713 how so?

[@lbgstzockt8493]

Two uploads in less than a year? I love it

[@gavengriesel4675]

Yeah same, I was like this can't be right..

[@magnus0re]

F

[@Breakingcraft]

Bless!

[@Gigabecquerel]

better not get used to it

[@ikbendusan]

he posts more often on patreon

[@reps]

One day late for pi day 😥

[@tomaszkarwik260]

xd

[@kapteniglo7404]

Approximating Pi to be 3.15 is good enough for engineers though

[@EgonSorensen]

piTY it is the day after that (m.ππ.TY) ~ 3 14 *15* 926535....

[@SplicesAndCelluloid]

​@@kapteniglo7404 I think you mean pi=e=3 is good enough for engineers. He's in the right month at least :P

[@namewarvergeben]

@@SplicesAndCelluloid Well, if 3 is good enough for engineers then 3.15 is clearly good enough too because the absolute error is smaller.

[@whatdamath]

amazing stuff! I hope they have a version for those of us who are solder challenged

[@AlanCanon2222]

I was thinking the same thing: my hands shake too much to do all that SMC stuff.

[@RayIrv]

Anton you here?! ... Love your content

[@heylel1841]

Neat Wonderful Anton is also here

[@teknoman117]

@@AlanCanon2222 I used to be terrified of the SMD parts. Eventually I just bit the bullet and made a board with 0603 components and some small pitch SMDs. The thing I ended up learning is that even if you short the pins on accident, decent solder wick and a bit of patience will let you fix it. Within about 6 months I got up to being able to hand solder 0.5mm pitch TQFPs fairly quickly. I guess I just had to learn how to be patient and that small mistakes along the way typically won't actually break anything.

[@pirminborer625]

Collegue: What did you do this weekend? Marco reps: Nothing special, just built a muon detector for CERN just for fun...

[@TomKappeln]

No no ... this is only what he let us see ! In reality he GENERATES black holes ...

[@AsymptoteInverse]

You're always building the kinds of projects I would build if 1) I had a little more money to spare, and 2) my brain worked a little better. Always a pleasure to watch.

[@Sx107music]

By the way: the particular signal amplitude does not matter that much in this exact application apart from distinguishing cosmic ray muons/electrons from each other and from other radiation sources. The particular muon energy does not contain much information, since after cosmic rays hit the atmosphere a huge particle shower is produced (just as you said), in which most of the information is, unfortunately, lost (for example, it is much easier to detect the initial particle mass (A) than it's charge Z). Most of the information is restored from the number of particles in the shower, not from their energy. Besides, the signal amplitude won't change very much with changing muon energy due to the particular "particle energy - energy losses (which signal amplitude is nearly proportional to)" curve - the bethe-bloch curve. That is why calorimeters, which absorb the particle, are way more precise in energy detection. Absorbing a muon requires a shitton of tungsten/lead and a huge number of detectors to acquire the precise path the particle has traveled and it's energy losses along the way. In short, there are ways around this problem (not only with a calorimeter straight approach), but certainly not cheap and easy ones. Signal amplitude usually in these applications contains more information about the particle type, not it's energy (for example, the signal amplitude is proportional to Z^2 and detecting the particle charge is way easier). Just to be clear: I am talking about registering cosmic ray muons, which are relativistic. At low energies the signal amplitude do rely on the particle energy - see bethe-bloch curve, again.

[@Sx107music]

Probably the first time experimental nuclear physics education paid off

[@tiporari]

@@Sx107music lol

[@sonotthere]

interesting. thank you for posting

[@PeterMilanovski]

What's a muon or electron look like? How can anyone know what is being detected if no one knows what is being detected?

[@Sx107music]

@@PeterMilanovski First, don't deny the existence of things that you can't perceive with your own senses, that's solipsism. Second, you can calibrate the detector, for example, with a known beta-radiation source (that is, it emits an electron) and see to which of the signal peaks it more or less corresponds in the signal amplitude distribution (but you will also have to consider the energy). Third, the understanding of cosmic rays, atmospheric particle showers and nuclear physics in general is so developed today, that you can just run a simulation of an atmospheric shower (in CORSICA or FLUKA, for example) and a simulation of particles passing through your detector (in GEANT4, for example) to see which signal amplitudes are you expecting, though at least some experimental calibration would be great - and, as it happens, a classical way of calibrating nuclear physics equipment is using cosmic ray muons, so in our case that's no problemo.

[@Toleich]

Looks like a fun project to participate in. * sees PCB * nope.jpg

[@TomKappeln]

Same here ... lol

[@TechGorilla1987]

Literally nobody: Have you heard about the Dry German Scientist Comedian? Me: Yes, I am subscribed with all notifications.

[@drdiesel1]

Love your content and sense of humor, hope one day to buy you a cold beer.

[@STriderFIN77]

i have been here before, butt Now i subcribed o.O

[@MarionMakarewicz]

That is my level of Patreon support for @MarcoReps. $3.00 a month. So I get him one beer a month. Reflecting on this it seems a bit stingy.

[@pe5erbarnes]

I'm using some BME280's for my dissertation: they're really nice sensors, 10/10 would use again!

[@funtechu]

Yep, definitely a favorite. I'm surprised they are out of stock right now, but I guess they are just that popular (and there is the chip shortage).

[@tomaszkarwik260]

me: :( me after seing this a Marco Reps video : :)

[@fffmpeg]

pozdrawiam

[@tomaszkarwik260]

@@fffmpeg pozdrawiam także

[@teknoman117]

I really hope that the component supply clears up by 2022. A project I was working on that I was hoping to turn into a kickstarter is completely on hold because like everything that isn't a dip part from 20 years ago or a passive is out of stock. FPGAs are gone, HDMI drivers are gone, ethernet PHYs are gone, USB2 ULPI and USB3 PIPE chips are gone, etc. I'm trying my luck with utsource at the moment...

[@sarkasaa]

ive lost count as to how often i have already rewatched all your videos, yet they always keep me hooked

[@pissmilker2313]

The Geiger counter sounds in the beginning fooled me. I thought I had a dying smoke detector in my house.

[@killsalot78]

same

[@DrTune]

Yeah on headphones; I thought "f-it! I just changed the batteries"

[@jean-pauldemars9588]

I just replaced mine today too and then same thing happened to me too. Lol

[@Sheevlord]

Just yesterday I was rewatching some of your older videos, thinking "when is he going to release a new one?" Talk about serendipity!

[@SuperSmashDolls]

For a second I thought this involved turning the DRAM on a Raspberry Pi into a cosmic ray detector.

[@piconano]

You are insanely talented. I love and watch your videos at least twice. Your videos are on another level.

[@brucewilliams6292]

Marco Reps hits the air; a smile hits my face! Thanks as always for the laughs and the projects. You really make my day.

[@muonpi778]

Just like Gigabecquerel, we are also investigating the possibility to make our own scintillators, but we still need to get the process down. Right now we are struggling to get it free from bubbles but we do already see some photons. Lots of work, but very fun and interesting.

[@TomKappeln]

A oil bath with ultrasonic does the trick. ;-)

[@TazkoHoZabit]

"Blitzortung" is a nice project, similiar network, but for lightning detection. I recommend a look.

[@mikemines2931]

Many thanks Marco, always worth dropping in. More projects like this please, how I miss the Amateur Scientist in Scientific American.

[@ruadeil_zabelin]

Plastic scintillator SLAPS. "WHAT?!" :D

[@Gameboygenius]

So Marco is a slapper...

[@vathevladimir]

Checkmate! 👆

[@JoanDoe-pq5fu]

We can now measure and study the holy slap.

[@thesciencefurry]

What the hell, I'm just researching on this topic and this video was just uploaded today :o I actually bought some photodiodes already, to build some simple circuits. I did look at scintillator materials. But for now I want to try it with x ray intensifying screen, because I have it there and I'm only able to get large chunks of the scintillator plastic.

[@jheregreign]

Same,. crazy at least cosmic pi

[@labiadh_chokri]

Nice video as always, I found a sat view of my country Tunisia at 2:27 .

[@ToTheGAMES]

Speaking of Bionerd, havent seen her around in ages. Hope she's still ok.

[@dtiydr]

Something is up for sure, either she is dead or very sick (not totally unlikely due to her highly unhealthy interest in radiation in a very unhealthy way) or "people" has told her to not putting up any more videos or things will go bad if she do and the same if she would give of any information about that someone have told her not to, would be my guess.

[@ryanmalin]

@@dtiydr What? Thats a lot of guessing. Hopefully none of it is true.

[@dtiydr]

@@ryanmalin Yea I thought its better to know if more ppl see this so I just did some digging. It seem that she made and put up some (unauthorized) videos of some radioactive things or such inside an forbidden abandon hospital and was blacklisted for those videos (now deleted). Then later she was found by the police in a forbidden area. Among with videos that she had posted around that time probably illegally as it seem (deleted) Ukraine now thought that enough is enough and banned her for life for ever entering again. This was her second home so she had a personal crisis after that and the place where the videos came from. I also found a person that had tried to hire her for a video in Ukraine but the person that was in charge for it said that she was banned from ever entering again. Found some info from the previous year that she is continue her educating in Germany. So she is ok as such but as it seem so would it probably unfortunately never be anymore videos posted on that account again.

[@TomKappeln]

@@dtiydr As far as i know, all true ... Greets from Germany. ;-)

[@TomKappeln]

@@dtiydr Her biggest mistake was to take the small Plutonium piece with her to the hotel, this is a no go. Filming that all and upload it, the second no go ...

[@sjekx]

amazing to see someting related to my work at this site. We use Nal crystals with photomultipliers to look at radiation underground. Very cool Marco!

[@luppa79]

I like that technique of getting an SMD part out with flame-heated small pliers! I gotta build something like those pliers with heat-resistant handles.

[@DouglasFish]

Hey! I have a ton of those BME things in my house. Cool to see you de-solder one. Thank you!

[@jabinstateresa]

I'm waiting for the moment that reps is finally sponsored by keysight.

[@nyceyes]

I'm always excited to see a video posted here. 🎊🎉 Thank you for this video uploader! 🤗

[@fr3shlama]

"Cheap (

[@TomKappeln]

Cheap ! If you had try to build this 5 years ago, you know why. (add 2 zeros)

[@gregorymccoy6797]

The humor... Oh my. Awesome video and project.

[@mitfreundlichengrussen1234]

As always entertaining like hell, including a cool foot warmer hack. By far beyond my personal capabilities still triggering some neurons in my brain. Thank you.

[@HouseGurke]

I NEVER expected a Davie504 reference in a Marco Reps video (at 5:29)

[@jebstalp]

Nice Bass and Davie504 reference! EPICO! And nice video as allways

[@sawmebabe]

Love your narration and sense on humor. Don't understand the details but love the projects.

[@Gigabecquerel]

Great video, love it!

[@420oppa69]

5:28 Davie504?

[@wallaceshackleton1889]

What a beautiful little video. Totally boffin-ish from my point of view and out of my technical reach. More power to you Pi-rotons!

[@ablebaker99]

There is also the "CosmicWatch" project that may be of interest. Unlike the "CosmicPI," you need to build at least 2 "CosmicWatch" units to detect the muons.

[@teknoman117]

Ordered some 4 layer board from jlcpcb to make an HDMI adapter for a FPGA board I have that only spits out LVDS. The redriver/level shifter chip is an 0.4mm pitch QFN that has 30 pins under a 4.5mm by 2.5mm package. I hope the stencils are good...

[@Toxicity1987]

I'm already on the waiting list for an MuonPi, but i probably can get it to work with the CosmicPi project too.

[@muonpi778]

nice :)

[@Sazoji]

MARCO: quick reminder, pixel-based censors on commandline text and numbers are easily decoded, text deblurring is like a beginner network to teach people how to make effective datasets. like, you almost should reupload before someone takes it as a challenge to find your GPS coordinates.

[@PhG1961]

By far one of your most amazing projects ! Really awesome. Since I have a some scintillator recycled stuff... I'm tempted to do some tinkering...

[@rowlandcrew]

Hi Marco, even if your broken NaI crystal were not broken, it and your two slab plastic scintillator are more like two digit spectrometers. If you want a real 9 digit radiation detector, the only one smaller than a truck (e.g. smaller than grand sasso macro) is the Fission Meter, which easily can identify a cosmic ray interaction. No need to rely on magical thinking that small uncalibrated scintillators will be better than a random number generator. Try collimating a gamma ray source perpindicular to your plastic slabs and proving that coincident pulses look different from the thorium and uranium in your house concrete/masonry. Let me know if you want some help.

[@ulwur]

In a previous career I repaired many a scintillation detector and I'm surprised you got it light tight enough with just black gaffatape. These things are pretty sensitive to ambient light.

[@brokepilot6969]

I hope I can one day get all the tools & such needed to build one of these one day! Looks very cool & it seems to be helping the scientific community.

[@provuksmc6619]

11:49 Those lines in the script 🤣

[@appercumstock3017]

Marco...we need new video in these barren times!!

[@nefariousyawn]

I'm really glad you still post on KZbin. I really enjoy your channel, but I frequently forget to check for uploads on Odysee. Even when your video topics dip into depths that I have a hard time following, your narration and sense of humor makes it enjoyable nonetheless.

[@damustermann]

Amazing project, thanks Marco for the heads up. Btw, have you seen Zack Freedmans video about modding the Keysight Oscilloscope? I couldn´t stop thinking about your reaction to that an laughed out loud.

[@whiteflag2366]

The best stand-up technical comedian.

[@YukuriuddoHerusaizu]

Dont pixelate black it out completely! using deeplearning people made some software with you can recover pixelated blocks of text. (I havent tried it out yet my self but I think people should be aware of it) Great video though!

[@MrYukon2010]

You are the German "This Old Tony". Love it!

[@nicknorthcutt7680]

"That tells you when something spicy is in the air" 😂

[@jokerace8227]

I like it. Do you think a Schumann Resonance meter is practically possible with Pi? ☕🐸

[@larcomj]

that hot plate is sweet. gonna have to pick one of those up. Awesome video as always.

[@antoineroquentin2297]

I've been following the cosmic pi project for some years and already wondered how that single cosmic pi in germany got up there. So I assume that's yours on the map :) I hope there will be soon a model for normalsterbliche.

[@TomKappeln]

Will be the same as with "Blitzortung", takes a while and than BOOOOM.

[@wackyvorlon]

You can also use the scintillator+PMT to build a gamma spectrometer.

[@asm_nop]

Is this project dead or on-hold or something? one project update in the last 2 years, and all the links to the data portals are dead..

[@Spacefish007]

8:25 good to see that some people have the same idea :). I use a 3D-Printer Hotbed for the same purposes :D

[@LegacyVision.]

Due to the public link to the Grafana and the limited data set on the map, blurring the GPS data doesn't do much unfortunately when the location is pinpointed on the Grafana page.

[@SpartasPhalanx]

I do the hardware design on a similar project at GSU. github/muontelescope has some stuff, a bit different topology and design, but still fun! Looking to have 40 detectors deployed this year.

[@Doctorbasss]

Marco, was simple fluorescent lamp or neon lamp triggered first by the passage of cosmic particle when HV is applied?

[@TheSwiip1]

i love the mini hot plate, mine arrived a week ago, now i can reflow small boards at home :)

[@astrophotographyenthusiast5273]

Didn’t even watch the full video yet but holy shit I am so excited. This looks so cool!

[@yaroc07]

What a wonderfull project! I would appreciate a radon only detector too :)

[@marcmarc172]

No way! I was just thinking about this idea!! Thanks Marco!

[@2jpu524]

Marco, what was your build cost? Also, an interesting point of fact, if you use a Thallium doped KI crystal as your scintillator allows you to build a spectrometer. This, in conjunction with a multi-channel analyzer that you can do in the STM32, would allow you to identify radioactive sources. I bet with a little more tweaking using your own x-ray source, you could create your own x-ray fluorescence detector, and identify, which high precision the composition of metal alloys, etc.

[@leloctai]

You should backup your data frequently. Sdcard don't like being written to constantly, mine fail after just 2 year, and I don't even log that much. I heard some people keep their for 5-10 years but no reason to risk it, especially when time series data are usually very compressible

[@1kreature]

I always use Eurocircuits eC stencils and already have their eC stencil fix for €15. No manual alignment of stencils, great result every time. Not quite as low as china prices though.

[@chalkchalkson5639]

Just FIY Geier-Müller counters are not really ideal for tissue dosimetry. The physics of how they detect different types of radiation is more or less independent from the physics that makes radiation harm us. PMMA as a scintillator coupled to a good photo-multiplier tube/diode is technically preferable since PMMA is tissue equivalent. The cross-sections of the various interactions in PMMA are very similar to tissue, so when you get a reading in PMMA you can be reasonably confident it will be no more than 2% (usually about 1%) different from that in tissue. With Geier counters you just calibrate and hope that the radiation you encounter has been adequately accounted for my the manufacturer calibrating your Geiger counter to display mSv. The disadvantage is that you need an expensive photo detector and the whole assembly tends to be more bulky for similar sensitivity. But there is a reason we often use PMMA phantoms in testing new CT scanners or radio therapy equipment. Kinda cool that plain old plexiglass glows when irradiated, isn't it? Now that I'm saying that, I don't think I've ever seen a photo of that, only plots and data in papers..

[@johnk7302]

I saw another technology used where it was a big device that required cooling or something i don't really remember but it was much bigger and relied on a different tech I believe not sure anymore

[@lllllll396]

Just wondering did you use optical gel for the SiPM and the scintillating plastic?

[@rlrfproductions]

Do you happen to have a stereo/3d microphone? I found myself repeatedly looking around the room for what the hell a noise was only to realize it was in the video

[@lernenmitrobin]

Thumbs up after the first second. I'm sure I won't get disappointed

[@navid9852]

Another great project of yours.

[@rpavlik1]

Wait, how did you actually buy a bme280 on eBay? I only manage to get the cheaper humidity-less bmp280. The Arduino Nano 33 BLE is an nrf52840 (an m4f like the f401, though a little lower clock), with a built in lsm9ds1, which might be a nice match for the needs of the project. It's handy that it's castellated with no bottom components for easy integration too. Adafruit has an nrf52840 feather "sense" that also has an st imu, though a slightly different model I think, plus a bmp280 and a separate humidity sensor. Of course, these nrf chips also have Bluetooth LE, though no wifi built in. What weird looking gaffer tape, guess we can't trust 3m for that particular product. Pro Gaff is the best one allegedly, though gaffer power is not far off (and it's the one I use). Also, black hot glue is cool. Not sure if I spotted that around your light guides, but if not, give it a try.

[@column.01]

Was just binging all your videos, are you reading my mind?

[@value8035]

What is the cost for entire project?

[@TheSakzzz]

JLC PCB seems to be the Raid shadows of the electronics world :D

[@Arsonist42]

torch and tweezers to remove smd components is a new one to me, frigging awesome !

[@Tedlasman]

OMG yesssssssssss I wish I had money to get on your patreon.

[@yasurikressh8325]

Very exciting project. Thank you

[@goldenpiston8449]

tbh, I would legit solder everything for this guy if he sent the shit to my lab, my way of supporting the channel

[@sethery7826]

I’ve got a Scintillation Detector from the 1950s. It is a Precision Radiation Instruments Model 111B. I got it working at one point, but it has since sat around :(

[@spokehedz]

While it doesn't work for every situation, I find that having a hot glue gun with a stick of black colored hot glue makes quick work of light leakage without any sticky residue everywhere... Just where you put the glue. I'd love to run one of these, if the dev's want to have someone from the USA running a board.

[@John_Smith__]

Marco Great Awsome video once again!! ... and on your last video you told me about the CERN repo of cool projects .. .a Gold Mine for everyone to explore no doubt! But I also told you last time ... I do not like vapor phase cooking to solder the chips ... never liked that process and now you also know why :) ...

[@K.D.Fischer_HEPHY]

You have my full Repspect for you work, Sir.

[@Veptis]

I do have some Germaium scrap pieces from thermal imaging lenses. Would these provide me an option to detect radiation?

[@martinstraub4336]

Marco, where did you get the scintillator "slabs" from? I could not find them at Ketek's.

[@crusaderanimation6967]

8:20 So.. you can use it as foot warmer *NOTED*

[@964cuplove]

There is a hint of Werner Herzog in that Voice - like the bad guy in Jack reacher....

[@emrechomko747]

You should try making a Hydrogen Line radio telescope. It’s basically an antenna and an RF front-end that’s tuned to receive and measure the strength of 1.42GHz radio waves that are naturally emitted by Hydrogen atoms. If you mount this contraption on a pan and tilt mechanism you can raster-scan the sky and create an image! If you wanna go further, you can measure the frequency deviation of the incoming radio waves from their centre frequency of 1,420,405,751.7667MHz, plug it into a Doppler Shift equation and see the direction and speed of the objects in the sky! Just the idea of it is enough to turn me on XD

[@weerobot]

Wow..Marco always Blows my Mind..

[@frosty129]

Marco, please update us on your OSMU project. It’s been a while and I am really hoping you get around to finishing it or at least make some progress. I was reading LT1970 datasheet earlier this week, and it seems very capable all in one solution.

[@jakobwilmesmeier9512]

Nice Project and really well done! If tried an similar project using six photodiodes to detect cosmic rays, but didn't ever get to finish it :(

[@oriole8789]

You seem to be using some heavy de-noising - what software/plugin are you using? Maybe Resolve's denoiser would serve you better since it does motion estimation as well, not just simple temporal denoising with a median filter.

[@vejymonsta3006]

Fantastic project. Love it