Identifying Elements with Backscattering [a DIY project]

Поділитися
Вставка

КОМЕНТАРІ • 82

  • @Posh-Arthur
    @Posh-Arthur 2 місяці тому +4

    We are all hoping that @Neptunium will help us to understand why increasing the proton count in a linear way should result in a straight line on a log-z graph.

  • @raymitchell9736
    @raymitchell9736 2 місяці тому +9

    Success or not, It was an interesting experiment and it was educational as well!

    • @project-326
      @project-326  2 місяці тому +1

      Thank you, It was a learning experience for me too, and that's why we make these videos!

  • @brucewilliams6292
    @brucewilliams6292 2 місяці тому +3

    You did a great job on explaining the back-scattering and I love the knowledge you impart. Thank you!

  • @eddydogleg
    @eddydogleg 2 місяці тому +6

    Time, distance, shielding brings back memories of sitting in the doghouse joking with the Schlumberger hands about how cute a three head kid could be.

  • @BRUXXUS
    @BRUXXUS 2 місяці тому +2

    This is a really fun and interesting video! If anything, it's a great demonstration of why the handheld analyzers are so expensive. There's so much that has to be taken into account to properly measure each thing. 😮 Well done!

  • @TerryLawrence001
    @TerryLawrence001 2 місяці тому +7

    Great video as always. I like the introduction of characters, it's a bit more entertaining. I suspect a few "serious" people will have problems with it. My "Little Garden" Spectrometer is on the way!!

  • @Muonium1
    @Muonium1 2 місяці тому +6

    I actually don't think I've ever seen anyone else here do this before, and that's really saying something considering how obsessed I am with exactly these kinds of science videos on the site! (I too have many a time wished to own and simultaneously lamented the insane cost of XRF devices) Thanks for your review and recommendation of the "Little Garden" spectrometer again, I just got mine and the spectral resolution is EXCEPTIONAL, I'm thinking of making it a little better by using a pair of razors for the entrance slit instead of the printed pla. There is another industry tool used to do what you are attempting here which doesn't use XRF or beta backscatter, it's a "spark spectrometer" or "spark optical emission spectrometry". Bruker and some other companies sell high end units to the metals industry for process analysis, but I think you could do a rudimentary version of it with none other than our LG spectrometer itself! It's just a spectral analysis of the plasma light emission from a low voltage spark contacting the metal in question. It's destructive, but only very slightly. I will also echo the comments of others urging the introduction of a 1cm thick piece of plastic before the lead in the shield portion of the device to suppress brems production in the lead (may as well keep the lead to block the few Y90 gammas).

    • @project-326
      @project-326  2 місяці тому +1

      Thank you for the ideas, I will add this to the already crazy long 'wish list' for videos.
      Really pleased you like the LG spetro too, if you look on the post on the community tab for this channel, we included the creators email address. He is a really nice (and modest) guy, it would be great if you sent him a quick email letting him know, it would make his day...
      As for the Bremsstrahlung radiation, I did a background measurement without any sample to deduct from the total from each sample that hopefully accounted for a lot of this (except what occurs in the sample itself). You idea about alternative shielding materials is also good, perhaps HDPE would be good too.

    • @PeterMarchl
      @PeterMarchl 2 місяці тому

      @@Muonium1 oh that sounds very interesting indeed! I hope you keep us updated on the progress!

  • @dwathen9
    @dwathen9 2 місяці тому +2

    e
    Exceptional video with clear, crisp narration!

  • @Neptunium
    @Neptunium 2 місяці тому +1

    hey wow! that is some very cool stuff my friend! I whish I thought of that! I am always amazed how talented people can do so much with so little!

    • @project-326
      @project-326  2 місяці тому

      Actually, I have a question for you. I have been trying to understand why this effect 'should' plot a straight line on a graph with a log-z scale, with a linearly increasing z-number. I'm not saying that I think this is wrong, I just don't understand why.
      I did some simple geometric analysis to see how the charge density per unit volume changes with z-number (assuming equal P and N), but I didn't find anything that was logarithmic in response, mind you, its wasn't linear either...
      Perhaps my approach is wrong, ie starting with the math and trying to apply it to the physics.
      Any idea where I might be able to find an answer to that question, or at least some clues? The AI engines I tried basically had a 'brain-fart' when trying to answer that question. All of the published data I managed to find (admittedly, not that much), on this subject clearly states a log-z response with count rate but never says why...

    • @Neptunium
      @Neptunium 2 місяці тому

      @@project-326 this sounds counter intuitive I agree! This smells like a math problem to me ! Let me make a fool of myself again and venture a wild guess! The log scale represent vastly separated numbers in your cases counts, but they can still be "aligned" or linear, since the radiation behave like light, it's intensity increase according to the inverse square law, I would expect the count to be far apart and a log scale would make sense. But why would the data stay linear? Why not? As long as Z increases at a steady pace there is no reason to deviate.. of course I could be completely wrong here ! This is just a first thought... More like rambling!

    • @Neptunium
      @Neptunium 2 місяці тому

      I guess the additional charge in the nucleus doesn't change the path of a single electron too much but with astronomical numbers of atoms the statistical results is very significant.. what do you think?

  • @mcwolfbeast
    @mcwolfbeast 2 місяці тому +3

    11:00 It actually makes sense that backscattering of beta particles from light elements consists of slower (less energetic) particles; smaller nuclei will have a weaker force acting on the particles meaning faster particles will not be curved as strongly and will not hit the detector. When measuring heavier elements, the opposite is true.

    • @project-326
      @project-326  2 місяці тому +2

      that was our assumption too, but it was very reassuring to be able to test this with the PIN diode spectrometer...

    • @Posh-Arthur
      @Posh-Arthur 2 місяці тому +1

      It does make complete sense that the heaver elements will be able to interact with faster betas, but it was good to be able to check that assumption. Physics has a habit of throwing up effects that often don't seem important in one situation but suddenly become dominant in another.

  • @PeterMarchl
    @PeterMarchl 2 місяці тому +2

    Now that was a very good video! I did try something like this with the KC761B but didn't make this nice rig and much less scientific patience... So it didn't bring the results I hoped. Now, I'll try again.

    • @project-326
      @project-326  2 місяці тому

      Thank you, I managed to get some results from the KC761(A) with the modifications to the PIN diode but there was less discrimination at lower energy, and hence with lower z-number elements. There seems to be a minimum activation energy for the PIN sensor.

    • @PeterMarchl
      @PeterMarchl 2 місяці тому

      @@project-326 yes... Beta mostly gets kind of dumped in the low energy channel and does not discriminate well. But the latest update did help with compensating the alpha particle energy quite well in comparison. Thanks again for all that info I learned in that video.

    • @project-326
      @project-326  2 місяці тому

      @@PeterMarchl The basic physics of the PIN diode is the main reason for the limited range of the beta energy spectrum, even with the update. The FW limits to a lower range bound of 100 KeV and the limits of the thin intrinsic, means that the response just keeps dropping with increased energy.
      You are right, the alpha energy compensation is a definite improvement with the last FW update.

  • @guygordon2780
    @guygordon2780 2 місяці тому +2

    OK, you got me with that penultimate sentence. I'm in.

  • @gblargg
    @gblargg 2 місяці тому +3

    This would be interesting for testing gold (and silver). Pb and Au are a couple of atomic numbers apart, and Pb is commonly injected into gold bars. The usual tests are density (weight and dimensions), a "ping" test to measure speed of sound (also using an ultrasonic thickness gauge), resistance (using eddy currents, a pulsed electromagnet next to it), and destructive testing using acids.

    • @project-326
      @project-326  2 місяці тому +3

      The down side to this technique is that it really is only testing the surface of the material, beta's don't need to get very far into a typical metal before they will be totally absorbed. For Au and Pb, that is a VERY short distance.
      Still, it was a fun experiment and we are pretty pleased that we got some good results.

    • @gblargg
      @gblargg 2 місяці тому +2

      @@project-326 Oh right, same issue with XRF testing, mostly just the surface.

  • @dbmail545
    @dbmail545 2 місяці тому +4

    "Patience is a finite resource around here"😂

    • @project-326
      @project-326  2 місяці тому +2

      I seem to make a mess of about 80% of the experiments that I perform, probably for that exact reason!

  • @mosmix6680
    @mosmix6680 2 місяці тому +2

    Nice experiments as always! By the way, mica tubes are cool and i am planning to make a counter using one of those. There is a ukrainian experimentator on youtube (Hamster time) who made a great video about them.

    • @project-326
      @project-326  2 місяці тому

      I'll check out that channel, thanks for the info.

  • @rob627x
    @rob627x 2 місяці тому +2

    Thank you for sharing this interesting experiment. Especially your measurements of electron backscatter spectra with the Measall KC761 of Al and Pb raise my curiosity. The energy distribution of Al looks very sharp. Did you compare it with direct beta radiation from Sr 90 source? Maybe not so easy, because of the much higher flux. Increasing the distance would help, but which role plays scattering in air? Primary electron energy should be 546 keV, if I am right? But there can be inelastic scattering in the source itself, maybe broadening the source spectrum (?).

    • @Posh-Arthur
      @Posh-Arthur 2 місяці тому +1

      the basic issue with the spectra of beta radiation is that it forms a continuum and not discrete energy levels. If you take a look at the spectra, you see a maximum energy at 546 KeV but the peak of the energy curve is at 196 KeV. We were working on a beta spectroscopy video (several months ago) but basically gave up over this issue, and the fact that the only device we have that can make any meaningful measurements of the beta spectra is this PIN diode based detector.
      We are working on a new video that includes a few interesting experiments based around this Sr-90 source and will show why we were not able to extract a spectra from our detectors.

    • @rob627x
      @rob627x 2 місяці тому

      Thank you for these interesting results, I would not have expected it to this degree. A lot of inelastic scattering seems to slow down the electrons here. Maybe in the source itself, if it is thick enough? On the other hand, I can hardly imagine that a short air passage and the inactive top layer of the PIN diode make a relevant contribution to such an energy loss. Of course, the low detection efficiency of the thin active layer of the PIN diode for fast electrons will lead to underestimating the high-energy part of the distribution. Anyway, I am curiously waiting for your new video.
      In the meantime, by the way, I found your very interesting and in-depth analysis of alpha spectra. Another video I came across might be an option for people who are interested in this topic and don't have the Measall KC761 / KC761B or are hesitant to void the warranty by removing the protective cover. It describes a DIY approach, so Measall's resolution might be better (?). According the alpha-energy calibration shown in this video I am now a bit skeptical too, after I have seen your analysis:
      ua-cam.com/video/1uoUGNa_g_E/v-deo.html

    • @project-326
      @project-326  2 місяці тому +1

      I think that a lot of the experiments do something similar to Stoppi, they just calibrate at the published energy level, which is fine as long as the reasons for it are valid in the context of the purpose experiment. In my case, I wanted to see if the energy resolution to alpha was accurate on the instrument that I was using, and I'm fairly happy that I managed to account for the differences between the theoretical values that come from the atomic models, and the values that I was seeing.
      For the beta energy, the KC761 cuts off in firmware below 100 KeV, the top layers of the diode are just going to absorb too much of the energy for useful measurement.

  • @presbiteroo
    @presbiteroo 2 місяці тому +2

    It's better to shield beta radiation with plastics, not lead. 5mm plexiglas is a good protection. This avoids the generation of x-ray that will travel farther.
    Also, always use some glasses, betas are terrible for the eyes.
    Nice experiment, maybe it could have also been done by measuring the noise in a camera in the dark caused by the back scattering. For an even more low cost way.

    • @project-326
      @project-326  2 місяці тому +3

      I did a measurement with a webcam (with all of the optics removed) using this same source and there is certainly some dark noise, but far less than I initially expected. On the surface of the CMOS sensor there is a fairly thick SiO2 passivization layer and the Bayer filter that seems to be better than expected at blocking direct (and very close) betas. That experiment will be part of another video that includes of other general beta experiments I have done over the last few months.

  • @DonnyHooterHoot
    @DonnyHooterHoot 2 місяці тому +2

    Cool viddy! What about this experiment with the radiacode? Peace!

    • @leandroebner1405
      @leandroebner1405 2 місяці тому +1

      The RC is unlike a GM tube or the KC Pin diode not sensitive to beta in the same way :). Different approach

  • @pmate95
    @pmate95 2 місяці тому +2

    You should check out the Raman spectroscopy. :)

    • @project-326
      @project-326  2 місяці тому

      that sounds interesting, but yes, we will certainly look into it, I think I probably have the required filters already.
      On this channel, everything is done 'on the cheap'...

  • @migsvensurfing6310
    @migsvensurfing6310 2 місяці тому +2

    Thank you for your time. Your videos are interresting.
    I have watched quite a lot of videos on the topic of radiation but I am still confused about how the different types of radiation has influence on our body and at what levels. Also if you would monitor for a nuclear plant like f.ex. Chernobyl... wich rays to look out for.. It is very cofusing to me.
    I ask you since you seem to have good knowledge and all other channels I have watched seem to copy each others statements and reviewing chinese products wich doesn't seem to do the job.

    • @project-326
      @project-326  2 місяці тому +1

      I was actually thinking of making a video on exactly that subject, there is a lot of confusing information on this aspect. Sadly, I already have a long list of videos to get through...

  • @markkaidy8741
    @markkaidy8741 2 місяці тому +2

    Are there any cheap XRF guns you can review?....I did buy a Radiacode 103 because of your video. Thank you!!

    • @project-326
      @project-326  2 місяці тому +1

      Sadly, even the cheapest seem to be very expensive (about $7K).

    • @gblargg
      @gblargg 2 місяці тому +2

      I think the only hope would be finding a used broken one you can repair. Those are probably still thousands.

    • @project-326
      @project-326  2 місяці тому +2

      @@gblargg Yes, indeed there are broken ones out there, but there also seem to be company's that specialize in buying them up, refurbishing them and re-selling. There is a lot of circuitry to go wrong in these things, the HV x-ray PSU is probably the main culprit...

  • @byronwatkins2565
    @byronwatkins2565 2 місяці тому +1

    What are your error bars? My eye says NONE of the data disagree with the linear model; all of the error bars overlap the line -- much less twice the error bars. One in three will disagree (to one standard error and on average) even when the data are taken from the same distribution.
    Since these counting statistics obey Poisson's distribution instead of Gauss', the total counts +/- sqrt(count) would be a more correct representation than average and standard deviation.

    • @project-326
      @project-326  2 місяці тому

      mean root deviation from total mean value.

  • @csmyfavoritecompany1213
    @csmyfavoritecompany1213 2 місяці тому +2

    Ive been looking for this device and its very supper expensive and you have to ask for a quote just to buy one of this

    • @project-326
      @project-326  2 місяці тому

      Yes, exactly. XRF machines are really cool but just so far away from my budget - hence this experiment with a cheaper technique...

    • @csmyfavoritecompany1213
      @csmyfavoritecompany1213 2 місяці тому +1

      @@project-326 idk if this method will it work on metal alloyed or metal with impurity I only want to separate the right metal that I melted using my furnace so I can separate each metal alloyed or none alloyed

    • @project-326
      @project-326  2 місяці тому

      @@csmyfavoritecompany1213 it needs very careful calibration with reference samples. If you are separating metals from scrap then a lot of the elements will also be transition metals with quite close atomic numbers, so the system accuracy needs to be very high. If you are separating Al from Pb, less so.
      Hope that helps!

  • @leandroebner1405
    @leandroebner1405 2 місяці тому +2

    Very nice!! :)

  • @davehimlin2374
    @davehimlin2374 2 місяці тому +1

    Any idea when you will be uploading a video about the NUKE ALERT keychain radiation detector ? Ty

    • @project-326
      @project-326  2 місяці тому

      There is a video coming up that mentions this, hopefully this one will be the next one but to be honest the testing has been really problematic due to the really poor sensitivity. I have had to make some fairly large investments in equipment and next week I will be able to find out if this has paid off.

    • @davehimlin2374
      @davehimlin2374 2 місяці тому

      @@project-326 I figured it would be difficult to use typical test samples for the NUKALERT, since it is not designed to signal a alarm for very low/ microseivert readings that are often obtained from normal household test samples.

    • @davehimlin2374
      @davehimlin2374 2 місяці тому

      @@project-326 I think you would need a test sample capable of putting out 1000 microseverts minimum per hour , in order to get the nukalert to sound the alarm of 1 chirp..and at this rate, it would equal 100 rems in about 40 days of constant exposure at 1 chirp from the detector ?

    • @davehimlin2374
      @davehimlin2374 2 місяці тому

      Is my above math correct, in regards to getting the NUKEALERT to create a audio chirp to prove it does detect radiation ?

  • @samoime1955
    @samoime1955 2 місяці тому +1

    courage experiment, if you could have compared the incident count rate to diverted, the ratio could have been related to element identification. Also the diode detector or even plastic scintillator could make better gamma discrimination that would also increase beta sensitivity. The question is also, how does the diversion angle depend on core mass? If you could measure on two different angular positions, the subtraction may give angular difference. This reminds on bremsstrahlung, that is principle of x-ray, as the high accelerated electrons do divert over the wolfram core and by reduction of speed do emit x-rays, but here the effect is not as significant likely to have E=mc2 resulted x-ray.

    • @project-326
      @project-326  2 місяці тому

      Thank you for the interesting feedback!

    • @samoime1955
      @samoime1955 2 місяці тому +1

      @@project-326 thank you and as you described the glass tube may not be the best solution is its geometry of cigar gives you a lot of background, regarding that source radiation may have more narrow radiation. So some cylindric window tube like LND712 shielded on sides may collect more source versus background, specially if you intended to reduce the distance your dispersion geometry fits even better.

  • @maxbartoshik
    @maxbartoshik 2 місяці тому +1

    Nice, thanks!

  • @olafmarzocchi6194
    @olafmarzocchi6194 2 місяці тому +1

    If you measure low energy radiation, put a thick copper plate after the lead, otherwise the lead will cause a shower of lower energy radiation as result of the slowing down of the initial particles you want to shield.

    • @project-326
      @project-326  2 місяці тому

      That's a great idea, thanks!

    • @project-326
      @project-326  2 місяці тому

      May I ask, why copper? What property of copper is helpful in this regards?
      Thanks in advance. Always looking to learn!

    • @olafmarzocchi6194
      @olafmarzocchi6194 2 місяці тому

      @@project-326 it's the most simple "graded-Z shielding". In wikipedia "Radiation protection", see:
      "In a typical graded-Z shield, the high-Z layer effectively scatters protons and electrons. It also absorbs gamma rays, which produces X-ray fluorescence. Each subsequent layer absorbs the X-ray fluorescence of the previous material, eventually reducing the energy to a suitable level. Each decrease in energy produces Bremsstrahlung and Auger electrons, which are below the detector's energy threshold"
      It can be lead-steel-copper, or lead-cadmium-copper, or similar. Lead and copper is already good.
      Search also "Homemade Graded-Z Shield for a Gamma-ray Spectrometer"

  • @BombasticVirus
    @BombasticVirus 2 місяці тому +2

    wait, why did you changed the voice. that's feels strange after so many videos I have watched xD

    • @BombasticVirus
      @BombasticVirus 2 місяці тому +1

      ah never mind you added 2 voices

    • @project-326
      @project-326  2 місяці тому

      don't worry, "posh Arthur" can't be fired...

    • @BombasticVirus
      @BombasticVirus 2 місяці тому +2

      actualy thats very nice that there is a robot icon when the original voice is speaking :)

    • @project-326
      @project-326  2 місяці тому +1

      @@BombasticVirus Thanks, hopefully we can develop the two characters going forward...

    • @Posh-Arthur
      @Posh-Arthur 2 місяці тому +1

      @@BombasticVirus I am the original channel host - I'm not getting fired any time soon!

  • @LabiaLicker
    @LabiaLicker 2 місяці тому +1

    Love your methodology!