0:00:00.000,0:00:13.630
rc3 prerol music
0:00:13.630,0:00:19.840
Herald: All right, fellow creatures, to be[br]honest, I never thought that I would be
0:00:19.840,0:00:25.440
introducing a talk on measuring[br]radioactivity like ever in my life, but
0:00:25.440,0:00:30.480
then again, considering the world stage,[br]current state at large, it might be not
0:00:30.480,0:00:35.600
such a bad idea to be prepared for these[br]things. Right? And gladly, our next
0:00:35.600,0:00:42.000
speaker, Oliver Keller, is an expert in[br]detecting radioactive stuff. Oliver is a
0:00:42.000,0:00:47.040
physicist and works at one of the most[br]prominent nerd happy places. The CERN
0:00:47.040,0:00:53.760
since 2013 is also doing a PhD project[br]about novel instruments and experiments on
0:00:53.760,0:01:00.800
natural radioactivity at the University of[br]Geneva and to even more to add even more
0:01:00.800,0:01:06.480
C3 pizzazz. Oliver is active in the open[br]science community and passionate about
0:01:06.480,0:01:12.960
everything open source. All that sounds[br]really cool to me. So without further ado,
0:01:12.960,0:01:18.400
let's give a warm, virtual welcome to[br]Oliver and let's hear what he has to say
0:01:18.400,0:01:24.800
about measuring radioactivity with using[br]low cost silicon sensors. Oliver, the
0:01:24.800,0:01:28.070
stream is yours.
0:01:28.070,0:01:32.560
Oliver: Thanks. That was a very nice[br]introduction. I'm really happy to have the
0:01:32.560,0:01:38.310
chance to present here. I'm a member since[br]quite some years and this is my first CCC
0:01:38.310,0:01:44.438
talk, so I'm quite excited. Yeah, you can[br]follow me on Twitter or I'm also on
0:01:44.438,0:01:50.156
Mastodon, not so active, and most of my[br]stuff is on GitHub. OK, so what will we
0:01:50.156,0:01:55.695
talk about in this talk? I'll give you a[br]short overview, also about the
0:01:55.695,0:02:02.035
radioactivity, because yeah, it's a topic[br]with many different details and then we
0:02:02.035,0:02:08.183
will look at the detector more in detail[br]and how that works in terms of the physics
0:02:08.183,0:02:13.160
behind it and the electronics. And then[br]finally, we look at things that can be
0:02:13.160,0:02:19.156
measured, how the measurement actually[br]works, what are interesting objects to
0:02:19.156,0:02:28.122
check and how this relates to silicon[br]detectors being used at CERN. So the
0:02:28.122,0:02:33.400
project is on GitHub called DIY Particle[br]Detector. It's an electronic design, which
0:02:33.400,0:02:39.437
is open hardware. There's a wiki with lots[br]of further details for building and for
0:02:39.437,0:02:45.363
troubleshooting. There is a little web[br]browser tool I will show later, briefly,
0:02:45.363,0:02:52.115
and there are scripts to record and nicely[br]plot the measurements. Those scripts are
0:02:52.115,0:02:58.056
BSD-licensed and written in Python. There[br]are two variants of this detector. One is
0:02:58.056,0:03:03.094
called electron detector, the other one[br]alpha spectrometer. They use the same
0:03:03.094,0:03:08.471
circuit board, but one is using four[br]diodes, the other one one photodiode...
0:03:08.471,0:03:14.180
There's a small difference between them,[br]but in general it's pretty similar. But
0:03:14.180,0:03:18.509
the electron detector is much easier to[br]build and much easier to get started
0:03:18.509,0:03:25.678
using. Then you have complete part lists[br]and even a complete kit can be bought on
0:03:25.678,0:03:31.638
kitspace.org, which is an open hardware[br]community repository, and I really
0:03:31.638,0:03:36.366
recommend you to check it out. It's a[br]great community platform and everyone can
0:03:36.366,0:03:44.105
register their own GitHub project quite[br]easily. Now, this is a particle detector
0:03:44.105,0:03:51.025
in a tin box, so you can use the famous[br]Altoids tin box or something for Swiss
0:03:51.025,0:03:57.342
chocolate, for example. You can see it's[br]rather small, the board about the size of
0:03:57.342,0:04:03.615
a nine volt block battery. And then you[br]need, in addition, about 20 resistors,
0:04:03.615,0:04:10.356
capacitors and these silicon[br]diodes plus an operational amplifier,
0:04:10.356,0:04:15.117
which is this little chip here, this[br]little black chip here on the right side,
0:04:15.117,0:04:19.520
you can see is all old school large[br]components. This is on purpose, so it's
0:04:19.520,0:04:24.635
easy to soldier for complete electronic[br]beginners. And this by the way, this
0:04:24.635,0:04:30.176
picture is already one user of this[br]project who posted their own build on
0:04:30.176,0:04:38.084
Twitter. OK, so natural radioactivity. So[br]I would say it's a story of many
0:04:38.084,0:04:44.080
misconceptions. Let's imagine we are this[br]little stick figure here on the ground.
0:04:44.080,0:04:50.644
Below us we have uranium and thorium. We[br]also have Potassium-40 in the ground and
0:04:50.644,0:04:56.809
Potassium-40 is is pretty specific and[br]peculiar. It actually makes all of us a
0:04:56.809,0:05:05.138
little bit radioactive. Every human has[br]about 4000 to 5000 radioactive decays
0:05:05.138,0:05:10.258
every second because of the natural[br]potassium and natural potassium comes with
0:05:10.258,0:05:14.979
a radioactive isotope, which is just[br]everywhere, it's in bananas. But it's also
0:05:14.979,0:05:20.320
in us because we need it for our body[br]chemistry. It's really important and even
0:05:20.320,0:05:26.224
some of those decays are even[br]producing anti-matter. So how cool is
0:05:26.224,0:05:32.432
that? OK, so what would we be measuring on[br]the on ground? Well, there could be some
0:05:32.432,0:05:39.429
gamma rays or electrons. Those are from[br]beta-decays. Or from the Uranium, there is
0:05:39.429,0:05:46.956
one radionuclide appearing in the decay[br]chain, which is called Radon, and Radon is
0:05:46.956,0:05:53.320
actually a gas. So from the ground the[br]Radon can diffuse upwards and travel with
0:05:53.320,0:06:01.137
air and spread around. So it's a bit like[br]a vehicle for radioactivity from
0:06:01.137,0:06:08.183
the ground to spread to other places. And[br]that Radon would decay with alpha
0:06:08.183,0:06:14.148
particles producing electrons and beta-[br]decays and also gamma radiation further
0:06:14.148,0:06:21.683
down in the decay chain. So just to[br]recapitulate, I've said it already twice,
0:06:21.683,0:06:29.828
so alpha particles are actually helium[br]nuclei, so it's just two protons and two
0:06:29.828,0:06:38.150
neutrons and the electrons are missing.[br]And in beta decay basically one neutron is
0:06:38.150,0:06:43.095
transformed into a proton and an electron.[br]And there's also an electron-anti-neutrino
0:06:43.095,0:06:47.694
generated. But this is super hard to[br]measure. So we're not measuring those.
0:06:47.694,0:06:54.524
Mostly we will be measuring electrons from[br]beta-decays. That's why you see all these
0:06:54.524,0:07:00.334
little e's indicating betadecays. Ok, if[br]you would go to the hospital here on the
0:07:00.334,0:07:08.965
left side, we would probably find some x[br]rays from checking our bones or something
0:07:08.965,0:07:16.484
like this, or even gamma rays or alpha[br]particles being used in treatments or very
0:07:16.484,0:07:22.814
modern even proton beams are sometimes[br]generated for medical applications. Now,
0:07:22.814,0:07:28.179
here on the right side, if you go close to[br]a nuclear power plant, we probably measure
0:07:28.179,0:07:33.886
nothing unless there's a problem in this[br]case, most likely we would find some gamma
0:07:33.886,0:07:40.639
radiation. But only if there is a problem.[br]OK, and then actually that's not the whole
0:07:40.639,0:07:46.886
story. This is terrestrial radiation. But[br]we also have radiation coming from
0:07:46.886,0:07:50.863
upwards, showering down on us every[br]minute, and there's actually nothing we
0:07:50.863,0:07:56.571
can do against it. So protons are[br]accelerated from in the universe.
0:07:56.571,0:08:02.416
Basically, the biggest particle[br]accelerator nature has. And once they hit
0:08:02.416,0:08:10.159
our atmosphere they break apart into less[br]energetic particles and it's many of them.
0:08:10.159,0:08:15.850
So in the first stage there's lots of pions[br]generated and also neutrons. But neutrons
0:08:15.850,0:08:21.625
are really hard to measure, so I'll ignore[br]them for most of the talk. Then those
0:08:21.625,0:08:29.045
pions can decay into gamma rays and then[br]trigger a whole chain of positron electron
0:08:29.045,0:08:34.877
decays, which again create gamma rays and[br]so forth. And this goes actually the whole
0:08:34.877,0:08:40.183
way down to the earth. We will have a[br]little bit of that on the sea level.
0:08:40.183,0:08:46.040
And the other more known part of[br]atmospheric radiation is actually muons.
0:08:46.040,0:08:52.374
So some pions decay into muons, which is[br]kind of a heavy electron and also
0:08:52.374,0:08:57.639
neutrinos. But neutrinos are, again, very[br]hard to measure. So I'll ignore them for
0:08:57.639,0:09:03.464
most of this talk. And if you look here on[br]the right side on this altitude scale,
0:09:03.464,0:09:08.420
you'll see an airplane would be basically[br]traveling where most of the atmospheric
0:09:08.420,0:09:13.323
radiation is produced. And this is why if[br]you go on such an airplane, you have
0:09:13.323,0:09:19.879
actually several times more radiation[br]in there than here on earth. And, of
0:09:19.879,0:09:24.210
course, on the ground, it also depends[br]where you are. There are different amounts
0:09:24.210,0:09:30.140
of uranium and thorium in the ground and[br]this is just naturally there. So but it
0:09:30.140,0:09:36.773
depends on the geology, of course. OK, so[br]I've talked quite a bit about radiation,
0:09:36.773,0:09:43.045
and I'm saying I want to use silicon to[br]detect it. So what radiation exactly?
0:09:43.045,0:09:48.514
Maybe. Let's let's take a step back and[br]think about what we know maybe from
0:09:48.514,0:09:54.840
school. So we have this rainbow for[br]visible light. Right. This is in terms of
0:09:54.840,0:10:03.000
wavelength. We have 800 to 400 nanometers[br]spanning from the infrared/red area to all
0:10:03.000,0:10:12.857
the green to blue and into the violet. And[br]lower down those wavelengths or let's say
0:10:12.857,0:10:17.744
bigger millimeter waves, meter waves and[br]even kilometer, that would be radio waves,
0:10:17.744,0:10:22.370
radio frequencies for our digital[br]communication systems, wi-fi, mobile
0:10:22.370,0:10:27.720
devices and so forth. But I want to look[br]actually more towards the right because
0:10:27.720,0:10:33.441
that's what we are measuring with these[br]detectors. It's shorter wavelength, which
0:10:33.441,0:10:39.797
actually means higher energy. So on the[br]right side, we would be having ultraviolet
0:10:39.797,0:10:47.328
radiation, which is kind of at the border[br]to what we can measure. And these 800 to
0:10:47.328,0:10:54.085
400nm translate into 1.5 to 3 eV, which is[br]a unit that particle physicists really
0:10:54.085,0:11:02.931
prefer because it basically relates the[br]energy of an electron after it has been
0:11:02.931,0:11:09.270
accelerated by 1 Volt and makes it[br]much easier to work with nuclear
0:11:09.270,0:11:15.385
particle physics, because everything, all[br]the energy is always related to an
0:11:15.385,0:11:21.208
electron. And this energy, this formula[br]here is just a reminder that the
0:11:21.208,0:11:25.937
wavelengths can be always converted into[br]energy and it's inversely proportional. So
0:11:25.937,0:11:30.551
wavelength increases to the left and the[br]energy to right. And if you increase
0:11:30.551,0:11:35.962
energy more from from the visible range,[br]so let's say thousands of electron volts,
0:11:35.962,0:11:43.400
then we arrive here. Millions - mega[br]electron volts, even GeV. And there is now
0:11:43.400,0:11:51.810
a pretty important distinction between[br]those two areas, and that is the right one
0:11:51.810,0:11:58.020
is ionizing radiation and the left one is[br]non ionizing radiation. UV is a little bit
0:11:58.020,0:12:03.150
in the middle of that. So some parts of[br]the UV spectrum can be ionizing. It also
0:12:03.150,0:12:09.567
depends a lot on the material that the[br]radiation is interacting with. For these
0:12:09.567,0:12:13.820
detectors I'm talking about today and[br]alpha, beta, gamma radiation, this is all
0:12:13.820,0:12:21.326
ionizing, so some examples, lowest energy[br]on the lower spectrum would be x rays than
0:12:21.326,0:12:29.440
electrons, gamma rays from radioactive[br]nuclides that already talked about in the
0:12:29.440,0:12:34.877
previous slide, alpha particles, and that[br]muons from the atmosphere would be more on
0:12:34.877,0:12:40.302
the GeV range and so forth. And for these[br]higher energies, of course, you need
0:12:40.302,0:12:46.184
something like the LHC to accelerate[br]particles to really high energies. And
0:12:46.184,0:12:56.415
then you can even access the TeV regime.[br]OK, silicon diodes. What kind of silicon
0:12:56.415,0:13:02.592
diodes? I'm using in this project, low[br]local silicon pin diodes, one is called
0:13:02.592,0:13:09.034
BPW34 it's manufactured from Vishay or[br]Osram, costs about 50 cents. So that's what
0:13:09.034,0:13:15.167
I mean with low cost. There's another one[br]called BPX61 from Osram. It's quite a bit
0:13:15.167,0:13:19.555
more expensive. This is the lower one here[br]on the right. It has a metal case, which
0:13:19.555,0:13:23.451
is the main reason why it's more[br]expensive. But it's quite interesting
0:13:23.451,0:13:28.523
because that one we can use for the alpha[br]detector. If you look closely, there is a
0:13:28.523,0:13:35.800
glass on top, but we can remove that. We[br]have a sensitive area. So this chip is
0:13:35.800,0:13:43.076
roughly 7mm² large and it has a thickness,[br]a sensitive thickness of about 50
0:13:43.076,0:13:49.681
micrometer, which is not a lot. So it's[br]basically the half of the width of a human
0:13:49.681,0:13:55.401
hair. And in total, it's a really small,[br]sensitive volume. But it's it's enough to
0:13:55.401,0:14:02.200
measure something. And just as a reminder,[br]how much of gamma rays or X-rays we will
0:14:02.200,0:14:09.057
detect with this, not a lot because it's[br]high, energetic photon radiation kind
0:14:09.057,0:14:15.471
doesn't interact very well in any kind of[br]matter. And because a sensitive area is so
0:14:15.471,0:14:21.399
thin, it would basically permeate through[br]it and most of the times not interact and
0:14:21.399,0:14:28.400
doesn't make a signal. OK, what's really[br]important, since we don't want to measure
0:14:28.400,0:14:35.120
light, we have to shield light away. We[br]need to block all of the light, that means
0:14:35.120,0:14:40.080
easiest way to do it is to put it in a[br]metal case. There is electromagnetically
0:14:40.080,0:14:44.880
shielded and completely protected from[br]light as well. Electromagnetic radiation
0:14:44.880,0:14:49.840
or radiowaves can also influence these[br]detectors because they are super
0:14:49.840,0:14:55.360
sensitive. So this sould be a complete[br]Faraday cage, complete metal structure
0:14:55.360,0:15:03.120
around it. There's a lot of hints and tips[br]how to achieve that on the wiki on the on
0:15:03.120,0:15:10.080
the GitHub of this project. OK, let's[br]think about one of those PIN diodes,
0:15:10.080,0:15:18.720
normally there is one part in the[br]silicon which is n-doped
0:15:18.720,0:15:23.920
negatively doped, and the other part[br]usually, which is positively dropped. And
0:15:23.920,0:15:28.080
then you arrive at a simple so called p-n-[br]junction, which is a regular
0:15:28.080,0:15:33.840
semiconducting diode. Now, pin diodes add[br]another layer of so-called intrinsic
0:15:33.840,0:15:41.920
layer, here shown with the i. And that[br]actually is the main advantage. Why this
0:15:41.920,0:15:50.800
kind of detector works quite well and have[br]a relatively large sensitive Sigma's. So
0:15:50.800,0:15:58.320
if you think about, let's say, a photon[br]from an x ray or gamma-decay or an
0:15:58.320,0:16:03.920
electron hitting the sensor. So by the[br]way, this is a cross-section view from the
0:16:03.920,0:16:09.840
side, but that doesn't really matter. But[br]let's say they come here from the top into
0:16:09.840,0:16:16.480
the... into the diode and we're looking[br]at the side then we have actually
0:16:16.480,0:16:22.000
ionization because this is ionizing[br]radiation, so we get free charges in the
0:16:22.000,0:16:26.960
form of electron-hole pairs. So electrons,[br]which here the blue ball and the red
0:16:26.960,0:16:34.240
circle would be the holes. And depending[br]on the radiation kind, how this ionization
0:16:34.240,0:16:39.760
takes place is quite different, but the[br]result is if you get a signal, it means
0:16:39.760,0:16:45.520
there was ionization. Now, if just this[br]would happen, we could not measure
0:16:45.520,0:16:53.120
anything. Those charges would quickly[br]recombine and on the outside of the diode,
0:16:53.120,0:16:58.720
it would be a little signal. But what we[br]can do is we can apply actually a voltage
0:16:58.720,0:17:06.320
from the outside. So here we just put a[br]battery. So we have a positive voltage
0:17:06.320,0:17:12.000
here, a couple of volts. And then what[br]happens is that the electrons would be
0:17:12.000,0:17:18.480
attracted by the positive voltage and the[br]holes will travel to the negative
0:17:18.480,0:17:26.720
potential. And we end up with a little net[br]current or a small bunch of charges that
0:17:26.720,0:17:35.200
can be measured across the diode as a[br]tiny, tiny current. The sensitive volume
0:17:35.200,0:17:41.040
is actually proportional to the voltage,[br]so the more voltage we put, the more the
0:17:41.040,0:17:44.880
bigger is our volume and the more we can[br]actually measure with certain limits, of
0:17:44.880,0:17:49.520
course, because the structure of the pin[br]diode has a maximum thickness just
0:17:49.520,0:17:56.800
according how it is manufactured. And[br]these properties can be estimated with
0:17:56.800,0:18:02.720
C-V-measurements. So here you see an[br]example of a couple of diodes, a few of
0:18:02.720,0:18:06.240
the same type. The two that I've[br]mentioned, they're different versions. One
0:18:06.240,0:18:11.360
has a transparent plastic case. One has a[br]black plastic case. Doesn't really matter.
0:18:11.360,0:18:16.800
You see, basically in all the cases, more[br]or less the same curve. And as you
0:18:16.800,0:18:21.760
increase the voltage, the capacitance goes[br]down. So it's great and basically shows us
0:18:21.760,0:18:26.960
those silicon chips are very similar, if[br]not exactly the same chip. Those
0:18:26.960,0:18:34.880
differences are easily explained by[br]manufacturing variances. And then because
0:18:34.880,0:18:39.280
this actually, if you think about it, it[br]looks a bit like a parallel plate
0:18:39.280,0:18:45.120
capacitor and actually you can treat it as[br]one. And if you know the capacitance and
0:18:45.120,0:18:50.240
the size, the area, you can actually[br]calculate the distance of these two plates
0:18:50.240,0:18:58.080
or basically width or the thickness of the[br]diode. And then we arrive at about 50
0:18:58.080,0:19:06.800
micrometer, if you put something like 8 or[br]10 volts. OK, now we have a tiny charge
0:19:06.800,0:19:11.600
current, now we need to amplify it, so we[br]have a couple of diodes, I'm explaining
0:19:11.600,0:19:16.640
now the electron detector, because it's[br]easier. We have four diodes at the input
0:19:16.640,0:19:21.360
and this is the symbol for an operational[br]amplifier. There are two of those in the
0:19:21.360,0:19:25.840
circuit. The first stage is really the[br]special one. So if you have a particle
0:19:25.840,0:19:31.200
striking the diode, we get a little charge[br]current hitting the amplifier. And then we
0:19:31.200,0:19:34.880
have here this important feedback[br]circuit. So the output is fed back into
0:19:34.880,0:19:40.880
the input, which in this case makes a[br]negative amplification. And the
0:19:40.880,0:19:46.320
amplification is defined actually by this[br]capacitance here. The resistor has a
0:19:46.320,0:19:51.360
secondary role with the small capacitance.[br]It is what makes the output voltage here
0:19:51.360,0:19:57.200
large. The smaller the capacitance, the[br]larger the output and it's inverted. Then
0:19:57.200,0:20:02.320
in the next amplifier step, we just[br]increase the voltage again to a level that
0:20:02.320,0:20:08.160
is useful for using it later. But all of[br]the signal quality that has been
0:20:08.160,0:20:13.120
achieved in the first stage will stay like[br]that. So signal to noise is defined by the
0:20:13.120,0:20:18.880
first stage. The second one is just to[br]better adapt it to the input of the
0:20:18.880,0:20:24.480
measurement device that's connected. So[br]here, this is a classic inverting
0:20:24.480,0:20:29.120
amplifier with just these two resistors[br]define the amplification factor. It's very
0:20:29.120,0:20:35.360
simple. It's just a factor of hundred in[br]this case. And so if you think again about
0:20:35.360,0:20:39.760
the charge pulse and this, the circuit[br]here is sensitive, starting from about
0:20:39.760,0:20:50.320
1000 liberated charges in those diodes as[br]a result from ionization. We get something
0:20:50.320,0:20:55.920
like 320 micro Volt at this first output,[br]and this is a spike that quickly
0:20:55.920,0:21:01.600
decreases. Basically these capacitors are[br]charged and quickly discharged with this
0:21:01.600,0:21:07.360
resistor and this is what we see here. And[br]then that is amplified again by a factor
0:21:07.360,0:21:14.080
of 100. And then we arrive at something[br]like at least 32 mV, which is conveniently
0:21:14.080,0:21:20.000
a voltage that is compatible with most[br]microphone or headset inputs of computers
0:21:20.000,0:21:25.600
or mobile phones, so that the regular[br]headset here has these four connectors and
0:21:25.600,0:21:31.520
the last ring actually connects the[br]microphone. The other is ground and reft.
0:21:31.520,0:21:39.360
Left, right for the earbuds. OK, how do we[br]record those pulses? This is an example of
0:21:39.360,0:21:46.400
1000 pulses overlayed and measured on an[br]oscilloscope here. So it's a bit more
0:21:46.400,0:21:52.400
accurate. You see the deposits a bit[br]better, kind of like the persistence mode
0:21:52.400,0:21:58.160
of an oscilloscope. And the size of the[br]pulse is proportional to energy that was
0:21:58.160,0:22:03.680
absorbed. And the circuit is made in such[br]a way that the width of the pulse is big
0:22:03.680,0:22:08.560
enough such that regular sampling[br]frequency of a sound card can actually
0:22:08.560,0:22:14.800
catch it and measure it. Yeah, this is[br]Potassium Salt. This is cut here. This is
0:22:14.800,0:22:18.720
called a low salt in the UK. There is also[br]a german variance, you can also just buy
0:22:18.720,0:22:26.320
it in the pharmacy or in certain organic[br]food stores as a replacement salt.
0:22:26.320,0:22:33.120
On the right side is an example from this[br]small Columbite Stone, which has traces of
0:22:33.120,0:22:38.720
uranium on it. And this is measured with[br]the alpha spectrometer. And you see those
0:22:38.720,0:22:42.640
pulses are quite a bit bigger here. We[br]have 50 microseconds and here we have more
0:22:42.640,0:22:52.880
like one milliseconds of pulse width. Now[br]there's a software on a browser. This is
0:22:52.880,0:23:00.720
something I wrote using the Web Audio API[br]and it works on most browsers, best is
0:23:00.720,0:23:06.640
Chrome, on iOs, of course, you have to use[br]Safari and that records once you plug the
0:23:06.640,0:23:13.120
detector, it records from the input at 48[br]or 44.1kHz the pulses. Here's an example
0:23:13.120,0:23:18.560
with the alpha spectrometer circuit, you[br]get these nice large pulses. In case of
0:23:18.560,0:23:22.800
the electron detector the pulse is much[br]shorter and you see it, you see the noise
0:23:22.800,0:23:28.880
much more amplified. This red line is kind[br]of the minimum level that the pulse needs
0:23:28.880,0:23:32.240
to trigger. This would be better. And[br]that's like the trigger level of an
0:23:32.240,0:23:38.160
oscilloscope. And you can set that with[br]those buttons in the browser. You need to
0:23:38.160,0:23:42.960
find a good value. Of course, if you[br]change your input volume settings, for
0:23:42.960,0:23:49.840
example, this will change. So you have to[br]remember which, with which settings it
0:23:49.840,0:23:55.600
works well. And it is pulsed, for example,[br]is even oscillating here. So for electron
0:23:55.600,0:24:01.440
detector, it's basically nice to count[br]particles. For the alpha detector it's
0:24:01.440,0:24:06.240
really the case where the size of the[br]pulse can be nicely evaluated and we can
0:24:06.240,0:24:11.120
actually do energy measurements. And these[br]energy measurements can be also called
0:24:11.120,0:24:17.520
spectrometry. So if you look closer at[br]these many pulses that have been recorded
0:24:17.520,0:24:26.560
and we find that there is really like much[br]more intensity, which means many more same
0:24:26.560,0:24:32.160
pulses were detected, we can relate it to[br]radium and radon. If we use a reference
0:24:32.160,0:24:35.920
alpha source and I have done this, I have[br]measured the whole circuit with the reference
0:24:35.920,0:24:41.360
sources and provide the calibration on[br]GitHub and you can reuse the GitHub
0:24:41.360,0:24:47.280
calibration if you use exactly the same[br]sound settings that I have used for
0:24:47.280,0:24:53.600
recording. And for example, these two very[br]weak lines here from two very distinctive
0:24:53.600,0:25:02.560
polonium isotopes from the uranium decay[br]chain. The top part here which is really
0:25:02.560,0:25:08.240
dark, corresponds basically in the[br]histogram view to this side on the left,
0:25:08.240,0:25:12.160
which is electrons. Most of these[br]electrons will actually enter the chip and
0:25:12.160,0:25:18.800
leave it without being completely[br]absorbed by it, but alpha particles
0:25:18.800,0:25:22.960
interact so strongly that they are[br]completely absorbed within the 50
0:25:22.960,0:25:29.520
micrometers of sensitive volume of these[br]diodes and OK here is a bit difficult to
0:25:29.520,0:25:35.440
see peaks. But the far end of the high[br]energy spectrum, you see two really clear
0:25:35.440,0:25:40.560
peaks and those can only stem from[br]polonium, actually. I mean, we know it's
0:25:40.560,0:25:46.960
uranium and that can only be polonium,[br]which is that isotope that produces the
0:25:46.960,0:25:56.080
most energetic alpha particles and[br]which is natural. And I said, if you use
0:25:56.080,0:25:59.840
the same setting like me, you can use it.[br]So the best is if you use actually the
0:25:59.840,0:26:04.560
same soundcard because they're if you put[br]it to hundred percent input sensitivity,
0:26:04.560,0:26:08.880
you will have exactly the same result,[br]like in my calibration case. And this
0:26:08.880,0:26:13.120
soundcard is pretty cheap, but also pretty[br]good. It costs just two dollars and has a
0:26:13.120,0:26:18.560
pretty range and resolves quite well, 16[br]bits and think, oh, you could do that with
0:26:18.560,0:26:24.640
Arduino as well, is actually a bit hard to[br]do. A really well defined 16 bit
0:26:24.640,0:26:31.280
measurement, even at 48 kHz. It's not so[br]easy and this keeps it cheap and kind of
0:26:31.280,0:26:34.960
straightforward. And you can have just[br]some Python scripts on the computer to
0:26:34.960,0:26:40.960
read it out. And this is as a reminder, in[br]order to measure alpha particles, we have
0:26:40.960,0:26:44.720
to remove the glass here on top of the[br]diode. So I'm doing it just cutting into
0:26:44.720,0:26:50.160
the metal frame and then the glass breaks[br]away easily. Is not a problem, there's
0:26:50.160,0:26:56.800
more on that on the wiki. Now we[br]can kind of compare alpha and gamma
0:26:56.800,0:27:03.920
spectrometry. Here's an example. This is[br]the uranium glazed ceramics. The red part
0:27:03.920,0:27:09.840
is uranium oxide that was used to create[br]this nice red color in the 50s, 60s, 70s.
0:27:09.840,0:27:15.040
And in the spectrum we have two very[br]distinctive peaks and nothing in the high
0:27:15.040,0:27:20.800
energy regime. Only this low energy range[br]has a signal. And this corresponds
0:27:20.800,0:27:27.920
actually to uranium 238 and 234 because[br]they use actually purified uranium. So all
0:27:27.920,0:27:33.920
of the high energy progeny or daughters of[br]uranium, they're not present here because
0:27:33.920,0:27:38.720
it was purified uranium. And this[br]measurement doesn't even need vacuum, I
0:27:38.720,0:27:43.280
put it just like this in a regular box. Of[br]course, if you would have vacuum, you
0:27:43.280,0:27:48.400
would improve this peaks by a lot. So this[br]widening here to the left, basically, that
0:27:48.400,0:27:55.200
this peak is almost below the other one.[br]That is due to the natural air at regular
0:27:55.200,0:28:00.960
air pressure, which already interacts a[br]lot with the particles and absorbs a lot
0:28:00.960,0:28:06.960
of energy before the particles hit the[br]sensor. So in terms of pros and cons, I
0:28:06.960,0:28:12.080
would say the small sensor is quite[br]interesting here in an alpha spectrometry
0:28:12.080,0:28:18.400
because it's enough to have a small[br]sensor. So it's cheap and you can measure
0:28:18.400,0:28:25.280
very precisely on specific spots. And on[br]the other hand, of course, the conditions
0:28:25.280,0:28:29.440
of the object influence the measurement a[br]lot. So, for example, if there's some
0:28:29.440,0:28:34.560
additional paint on top, the alpha[br]particles might not make it through. But
0:28:34.560,0:28:40.560
in most of these kind of samples, alpha[br]radiation actually makes it through the
0:28:40.560,0:28:46.800
top, a transparent paint layer. In terms[br]of gamma spectrometry, you would usually
0:28:46.800,0:28:51.760
have these huge and really expensive[br]sensors. And then the advantage, of
0:28:51.760,0:28:56.800
course, is that you can measure,[br]regardless of your object, you don't
0:28:56.800,0:29:01.040
really need to prepare the object a lot.[br]You might want some lead shielding around
0:29:01.040,0:29:06.400
it. That's again, expensive, but you can[br]do it. You can improve the measurement
0:29:06.400,0:29:14.000
like that. And it's basically costly[br]because the sensor is quite expensive.
0:29:14.000,0:29:19.680
Vice versa in the set setup for 15 to 30[br]euro. You have everything you need and
0:29:19.680,0:29:28.160
here you're looking at several hundred to[br]several thousand euros. OK, now measuring
0:29:28.160,0:29:34.880
I have to be a bit quicker now, I noticed.[br]So I talked about the potassium
0:29:34.880,0:29:39.440
salt. There's also fertilizer based on[br]potassium baking powder. Uranium glass is
0:29:39.440,0:29:44.800
quite nice. You can find that easily on[br]flea markets. Often also old radium
0:29:44.800,0:29:50.080
watches. Here's another example of a[br]uranium glaze, the kitchen tile in this
0:29:50.080,0:29:54.400
case, this was actually in the kitchen. So[br]the chances are that you at home find
0:29:54.400,0:29:58.240
actually some of those things in the[br]cupboards of your parents or your
0:29:58.240,0:30:01.840
grandparents. It is an example of[br]thoriated glass, which has this
0:30:01.840,0:30:08.400
distinctive brownish color, which actually[br]is from the radiation. And a nice little
0:30:08.400,0:30:12.720
experiment that I can really recommend you[br]to look up is radioactive balloon
0:30:12.720,0:30:17.920
experiment. Here, you charge the balloon[br]electrostaticly and then it would catch
0:30:17.920,0:30:21.840
polonium from the air. And it's really[br]great. You basically get a radioactive
0:30:21.840,0:30:30.803
balloon after it was just left for 15[br]minutes in a normal regular room. OK, now
0:30:30.803,0:30:36.928
the last kind of context of all of[br]this to end this presentation, I want to
0:30:36.928,0:30:43.280
quickly remind how important these silicon[br]detectors are for places like CERN. It's a
0:30:43.280,0:30:48.815
cross-section of the ATLAS detector. And[br]here you have basically the area where the
0:30:48.815,0:30:53.950
collisions happen in the ATLAS detector.[br]So this is just a fraction of a meter. And
0:30:53.950,0:31:02.049
you have today 50 to 100 head on collisions[br]of two protons happening every 25
0:31:02.049,0:31:08.380
nanoseconds. Not right now, but soon[br]again, machines will be started again next
0:31:08.380,0:31:15.112
year. And you also can, by the way, build[br]a similar project which has a slightly
0:31:15.112,0:31:19.492
different name. It's called Build Your Own[br]Particle Detector. This is Atlas and made
0:31:19.492,0:31:25.210
out of LEGO. And on this website, you[br]find a nice plan, how to build or ideas,
0:31:25.210,0:31:32.504
how to build it from LEGO to better[br]visualize the size and interact more with
0:31:32.504,0:31:38.434
particle physics. In case of the CMS[br]detector. This is the second biggest
0:31:38.434,0:31:43.648
detector at CERN. Here you see nicely that[br]in the middle, at the core of the
0:31:43.648,0:31:49.362
collision, you have many, many pixel and[br]microstrip detectors which are made of
0:31:49.362,0:31:59.464
silicon. And these are actually 16 m² of[br]silicon pixel detectors and 200m² of
0:31:59.464,0:32:04.870
microstrip detectors also made of silicon.[br]So without basically that silicon
0:32:04.870,0:32:10.649
technology modern detectors wouldn't work[br]because this fine segmentation is really
0:32:10.649,0:32:16.611
required to distinguish all of these newly[br]created particles as a result of the
0:32:16.611,0:32:24.609
collision. So to summarize the website is[br]on GitHub, there is really this big wiki,
0:32:24.609,0:32:29.349
which you should have a look at, and[br]there's a gallery of pictures from users.
0:32:29.349,0:32:34.307
There's some simulation software that I[br]used as well. I didn't develop it, but I
0:32:34.307,0:32:38.920
wrote how to use it because the spectra[br]can sometimes be difficult to interpret.
0:32:38.920,0:32:44.440
And there's a new discussion forum that I[br]would really appreciate if some of you had
0:32:44.440,0:32:49.648
some discussions there on GitHub. And most[br]of the things I saw today are actually
0:32:49.648,0:32:54.594
written in detail in a scientific article,[br]which is open access, of course. And I
0:32:54.594,0:33:00.198
want to highlight two related citizen[br]science projects on the one hand, as the
0:33:00.198,0:33:07.382
safecast, which is about a large, nice,[br]sensitive Geiger-Müller based detector
0:33:07.382,0:33:12.640
that has the GPS and people upload their[br]measurements there. This is quite nice.
0:33:12.640,0:33:17.318
And also opengeiger is another website,[br]mostly German content, but also some of it
0:33:17.318,0:33:23.412
is English, that also uses diode[br]detectors, showed many nice places. He
0:33:23.412,0:33:29.971
calls it Geiger caching, places around the[br]world where you can measure something,
0:33:29.971,0:33:35.032
some old mines, things like this. And if[br]you want updates, I would propose to
0:33:35.032,0:33:40.156
follow me on Twitter. I'm right now[br]writing up two other articles with more
0:33:40.156,0:33:46.889
ideas for measurements and some of the[br]things you have seen today. Thanks a lot.
0:33:50.949,0:33:57.840
Herald: Well, thanks a lot, Oliver. I hope[br]everyone can hear me now again. Yes,
0:33:57.840,0:34:02.880
thanks for mentioning the citizen science[br]project as well. It's really cool I think.
0:34:02.880,0:34:09.840
We do have a few minutes for the Q&A and[br]also a lot of questions coming up in our
0:34:09.840,0:34:18.240
instance at the IRC. So the first question[br]was, can you talk a bit more about the SNR
0:34:18.240,0:34:24.000
of the system? Did you pick particular[br]resistor values and or Opamps to optimize
0:34:24.000,0:34:29.280
for noise? Was it a problem?[br]Oliver: Yeah, so noise is the big
0:34:29.280,0:34:36.880
issue here. Basically, the amplifier is[br]one I found that this around four, four
0:34:36.880,0:34:45.040
euros, trying to find the slide. Yeah, you[br]have to look it up on GitHub to the
0:34:45.040,0:34:49.600
amplifier type, but this is the most[br]important one. And then actually the
0:34:49.600,0:34:54.800
resistors, here, the resistance in the[br]first stage, sorry, the capacitors is the
0:34:54.800,0:34:59.200
second important thing. They should be[br]really small since I'm limited here with
0:34:59.200,0:35:07.520
hand soldarable capacitors. Basically I[br]choose the one that were just still
0:35:07.520,0:35:11.200
available, let's say, and this is[br]basically what is available is basically a
0:35:11.200,0:35:15.840
10 pF capacitor. If you put two of them,[br]one after another, you half the
0:35:15.840,0:35:20.800
capacitance, so you get five. And this, by[br]the way, is also then the capacitor. So I
0:35:20.800,0:35:29.280
kind of tried to keep the same[br]resistor values as much as possible, and
0:35:29.280,0:35:32.720
here at the output, for example, this is[br]to adjust the output signal for a
0:35:32.720,0:35:37.680
microphone input in the alpha[br]spectrometer, I changed the values quite a
0:35:37.680,0:35:43.440
bit to make a large pulse. But, yeah,[br]it's basically playing with the time
0:35:43.440,0:35:47.670
constants of this network and this[br]network.
0:35:49.425,0:35:56.000
Herald: All right, I hope that answers for[br]the person. Yeah, but people can get a
0:35:56.000,0:36:02.480
contact to you right after the show maybe[br]as well. So there's another question. Have
0:36:02.480,0:36:12.400
you considered using an I²S Codec with a[br]Raspberry Pi? radiation H80 should be
0:36:12.400,0:36:17.120
almost no set up and completely[br]repeatable. So last ones are for comment.
0:36:19.520,0:36:25.280
Oliver: I don't know that component, but,[br]yeah, as I said, using a sound card, it's
0:36:25.280,0:36:31.280
actually quite straightforward. But of[br]course there's many ways to get fancy.
0:36:31.280,0:36:35.440
And this is really this should actually[br]attract teachers and high school students
0:36:35.440,0:36:41.040
as well, this project. So this is one of[br]the main reasons why certain technologies
0:36:41.040,0:36:45.399
have been chosen, rather simple than,[br]let's say, fancy.
0:36:45.399,0:36:51.612
Herald: Yeah, so it should work with a lot[br]of people, I guess, and one another
0:36:51.612,0:36:58.169
question was how consistent are the sound[br]cards? Did you find the same calibration
0:36:58.169,0:37:04.733
worked the same with several of them?[br]Oliver: So if you want to use my
0:37:04.733,0:37:11.701
calibration, you should really buy this[br]two dollar card from eBay, CM108. I
0:37:11.701,0:37:20.620
haven't seen a big difference from card to[br]card in this one. But of course, like from
0:37:20.620,0:37:25.790
one computer to the mobile phone, it's a[br]huge difference in input, sensitivity and
0:37:25.790,0:37:30.782
noise. And it's very difficult to reuse[br]the calibration in this case. But you
0:37:30.782,0:37:39.344
still can count particles and the electron[br]detector is anyway, um, mostly it actually
0:37:39.344,0:37:43.124
just makes sense for counting because the[br]electrons are not completely absorbed. So
0:37:43.124,0:37:47.526
you get an energy information, but it's[br]not the complete energy of the electron.
0:37:47.526,0:37:52.971
So yeah, you could use it for x rays, but[br]then you need an x ray machine. So yeah.
0:37:52.971,0:37:58.685
Herald: Who doesn't need an x ray machine,[br]right? laugs So maybe one question I
0:37:58.685,0:38:05.123
have, because I'm not very familiar with[br]the tech stuff, but what actually can be
0:38:05.123,0:38:11.881
done with it right in the field. So you[br]mentioned some working with teachers with
0:38:11.881,0:38:17.523
these detectors? What have you done with[br]that in the wild so to say?
0:38:17.523,0:38:23.800
Oliver: So what's quite nice is you can[br]characterize stones with it, for example.
0:38:23.800,0:38:29.840
So since you can connect it to a[br]smartphone this is completely mobile and
0:38:29.840,0:38:34.379
it goes quite well in combination with a[br]Geiger counter in this case. So with a
0:38:34.379,0:38:38.440
Geiger counter, you just look around,[br]where are some hot spots and then you can
0:38:38.440,0:38:44.665
go closer with the alpha spectrometer and[br]actually be sure that there is some traces
0:38:44.665,0:38:51.883
of thorium or uranium on the stone, for[br]example. Or in this type of ceramic, these
0:38:51.883,0:38:58.952
old ceramics, you can go to the flea[br]market and just look for these very bright
0:38:58.952,0:39:04.310
red ceramics and measure them on the spot[br]and then decide which one to buy.
0:39:04.310,0:39:12.072
Herald: OK, so that's what I'm going to do[br]with it. Thanks for for highlighting a bit
0:39:12.072,0:39:18.846
the practical side, I think it's really[br]cool to educate people about scientific
0:39:18.846,0:39:25.734
things as well. Another question from the[br]IRC. Didn't you have problems with common
0:39:25.734,0:39:30.999
mode rejection by connecting the device[br]through the sound card? If yes have you
0:39:30.999,0:39:38.345
tried to do a AD conversion digitization[br]on the bord itself already? Transfer
0:39:38.345,0:39:42.440
transfer wire SP dif?[br]Oliver: Yeah, so, of course, I mean, this
0:39:42.440,0:39:48.154
is the thing to do, if you want to make a[br]like a super stable, rock solid
0:39:48.154,0:39:54.224
measurement device, but it is really[br]expensive. I mean, we are looking here at
0:39:54.224,0:40:00.837
15 euros and yeah, that's the reason to[br]have this separate soundcard just to
0:40:00.837,0:40:08.206
enable with some very low resources to do[br]this. But I'm looking for these pulses. So
0:40:08.206,0:40:15.660
this common mode rejection is a problem.[br]And also this is kind of Überschwinger -
0:40:15.660,0:40:22.610
I'm missing the English term. This is kind[br]of oscillations here. If you design a
0:40:22.610,0:40:27.654
specific analog to digital conversion, of[br]course, you would take all of that into
0:40:27.654,0:40:32.671
account and it wouldn't happen. But here[br]this happens because the circuit can never
0:40:32.671,0:40:38.097
be exactly optimal for certain soundcard[br]input. It will always be some mismatch of
0:40:38.097,0:40:44.508
impedances and.[br]Herald: All right, so maybe these special
0:40:44.508,0:40:52.000
technical issues and details, this could[br]be something you could discuss with Oliver
0:40:52.000,0:41:00.112
on Twitter of maybe Oliver or want to join[br]the IRC room for your talk as well. People
0:41:00.112,0:41:07.120
were very engaged during your talk. So is[br]this always a good sign. In that sense I'd
0:41:07.120,0:41:14.680
say thank you for being part of this first[br]remote chaos experience. Thanks again for
0:41:14.680,0:41:21.433
for your talk and for taking the time and[br]yeah, best for you and enjoy the rest of
0:41:21.433,0:41:28.474
the conference of the Congress and a warm[br]round of virtual applause and big thank
0:41:28.474,0:41:33.024
you to you, Oliver.[br]Oliver: Thanks, I will join the chat room
0:41:33.024,0:41:34.553
right now.
0:41:34.553,0:41:39.272
rc3 postrol music
0:41:39.272,0:42:14.000
Subtitles created by c3subtitles.de[br]in the year 2021. Join, and help us!