So, no, your pulse height tells precisely nothing about the energy (color) of the original photon.
Thanks, Carl, I think I'm finally beginning to understand why that is definitely the case. Thank you, too, for your illustration of "several million throws of the dice". That sums up the process quite well.
Mark, you are correct in thinking that liquid scintillation counters work somewhat like gamma counters. The major difference is due to the energy of the events being detected. A gamma ray (or cosmic ray) can travel through a fairly large amount of matter and still have enough energy to cause a fairly bright flash of light in a sodium iodide crystal. That flash is intense enough to produce a signal that is well above the thermionic noise of the PMT.
A beta particle (free electron) has a much lower energy and can only penetrate a few millimeters of air, so to measure them the beta emitting material is mixed with a liquid that emits light when struck by an electron and placed in a glass vial. This flash of light is so faint that the signal produced by the PMT is in the same range as that produced by thermionic emissions. To detect it a PMT is placed on each side of the vial and only those pulses that appear on both PMT's at the same time are analyzed. Almost all of the thermionic noise is eliminated since those pulses occur at random times, and only rarely happen at the same time in both PMT's.
Both the sodium iodide detector and the liquid scintillation vial and PMT's were placed inside lead shielding to remove as much of the background and cosmic ray produced counts as possible. In the case of cosmic rays, most of them were of such high energy that they were above the upper window level for both gamma and liquid scintillation counting.
To give some idea of the difference having 2 PMT's in coincident counting mode made, if I counted pulses from individual tubes I would get about 10K-20K counts per minute, and with the same tubes in coincidence counting it would drop to 15 to 20 counts per minute.
Carl is right in that photon energies can not be determined accurately (by a PMT) if they are only energetic enough to cause emission of a few electron from the photo cathode. The more electrons you get initially the more accurate the result. Ain't quantum effects and statistics wonderful.
One book I keep next to "The Art of Electronics" is "Building Electro-Optical Systems, Making it All Work" by Philip Hobbes. It has been a constant resource, although I am using other components and don't have personal experience with PMTs. The book is written in a very practical style, although it does throw around a lot of terminology that often leaves me with as many questions as answers. The chapter on detectors of course includes PMTs operating in photo counting and analog modes, and also limitations and aging factors.
To give some idea of the difference having 2 PMT's in coincident counting mode made, if I counted pulses from individual tubes I would get about 10K-20K counts per minute, and with the same tubes in coincidence counting it would drop to 15 to 20 counts per minute.
Thank you, kwinn. Yes, I can see the wisdom in using the coincidence mode, and my earlier prototype used exactly that and it worked pretty well. Problem is, my ultimate application is bioluminescence and I'm told the photons will probably NOT be emitted in bursts but will simply "rain down faster and faster." The photons won't necessarily be correlated, so coincidence circuitry might actually miss what I'm looking for. So I think finding a good threshold value to discern signal from noise is going to take me a lot of trial and error.
Tracy Allen said...
...."Building Electro-Optical Systems, Making it All Work" by Philip Hobbes. It has been a constant resource...
Tracy, thanks for the tip. I'll see if my library can hunt this down.
Creating single-atom and single-photon sources from entangled atomic ensembles
M. Saffman and T. G. Walker
Department of Physics, University of Wisconsin, 1150 University Avenue, Madison, Wisconsin 53706
~Received 14 March 2002; published 16 December 2002!
We discuss the application of dipole blockade techniques for the preparation of single-atom and singlephoton
sources. A deterministic protocol is given for loading a single atom in an optical trap, as well as ejecting
a controlled number of atoms in a desired direction. A single-photon source with an optically controlled
beamlike emission pattern is described.
Source Quote
We turn now to the creation of a phased array single photon
source with a diffraction-limited emission pattern.
When the optical fields propagate through the sample, atoms
will be excited with position dependent relative phases as in
Eq. ~1!. This results in an entangled state with a phase structure
that mimics the phase of the exciting beam and can be
used to create a single-photon source with well-defined directionality.
While previous work has achieved single-photon
sources with a controlled emission direction by coupling to
microcavities @12# our approach results in a source emission
pattern that is reconfigurable and is defined by the structure
of the preparation optical fields. Referring to Fig. 3, fields at
v1 and v2 with Rabi frequencies V1 and V2 drive a twophoton
transition ua&!ue&!ur& to the Rydberg level ur& in
an N-atom ensemble. The effective Rabi frequency for the
two-photon process acting on atom j is uVueif11if2
5(uV1ueif1(rj)uV2ueif2(rj))/2De , where f1 and f2 are the
phases of fields v1 , v2 at the atomic position rj and De
5v12vea . As long as Pdouble given by Eq. ~3! is small only
transitions to states with a single excited atom are energetically
allowed and we have an effective dipole blockade. Under
these conditions an ensemble of atoms in ua& subjected to
a p-pulse applied on ua&!ur& produces the entangled symmetric
superposition state uc&5(2i/AN)(jeı(f1 j1f2 j)ur j&.
The phases fmj5km•rj are simply the phase of the mth laser
field (;eı(km•r2vmt)) at the position rj of the jth atom.
Creating single-atom and single-photon sources from entangled atomic ensembles.....
Thanks Humanoido,
I think I happened across that article (or one like it) back when I was trying to figure out how to prove my photon counters were actually working the way I wanted them to. Of course all that single photon generation stuff is way over my head, so I just went with something dumb and simple.
I am going back to the original question part one . How fast can you modulate a LED .. Well Fiber optics TX LED based modules are up to 600Mhz . If that helps
Ironical... ElectricAye I was museing on the idea of a gieger counter this week .
there EASY to make .
Those would be perfect for the long nights when you are mystified by quantum mechanics,
can't sleep, and just have to do that double slit experiment one more time...just to be certain. :-)
It seems like you could generate a single photon by using a series of baffles, mirrors and pinhole appertures. You might be able to use a prism to select the wavelength that you want. Pulse the LED for as short a time as possible. Let's say it generates a trillion photons. Run that through a couple of pinhole colimators to get it down to a few billion photons. Pass that through a prism to get the desired wavelength, and then pass that through another pinhole apperture, so its down to a million photons. If you use mirrors that scatter the light over a wide angle you should be able to get down to a single photon after a few iterations of mirrors and pinholes.
... you should be able to get down to a single photon after a few iterations of mirrors and pinholes.
Or no photons at all. How do you know when you've filtered too much? And how do you find just the right edge between one photon and none? This is beginning to sound like quantum homeopathy!
idQuantique manage it with their single photon sources by attenuating the light from a laser. When you get it right you are guaranteed to get single photons, it's an inherent property of quantum mechanics. I think they measure the photon production rate and adjust the laser intensity until they get the required rate.
I can see Phil's point, and it's something I've always wondered about any sort of experiment where they claim to get down to a single photon. How could you ever know you got only a single photon? How can you know the efficiency of your counting device - the apparent vs. the real number of photons hitting it? If you count a photon, then you've interacted with it and so it can't be counted and then used because, well, you've just used it up when you counted it. Maybe you can count lots of them, then dilute them with darkness, but then you'd only be presuming you are getting down to a statistical single photon with the proper dilutions... But how would you really know that some sort of weird property of physics doesn't kick in and tinker with the statistics at the single photon level?
By simple observation:
How fast can you blink an LED?
You can blink it so fast that it appears to stay on and not blink.
How dim can it get?
You can make it so dim that it will appear to be off.
BTW, darkness, as nothing, is not an element for quantification.
Even in the most primordial state, there is a froth.
By definition, one can only create the absence of light.
It is light that is quantifiable.
How do you know when you get single photons? You look at the output of the photomultiplier on a scope, and when the pulses have gaps between them you are getting single events. I found one method on the internet that uses two polarizers to attenuate the light, plus very small pinholes as I described earlier. Their setup didn't even use mirrors or baffles -- just the polarizers and the pinhole apperature.
Comments
Thanks, Carl, I think I'm finally beginning to understand why that is definitely the case. Thank you, too, for your illustration of "several million throws of the dice". That sums up the process quite well.
cheers,
Mark
A beta particle (free electron) has a much lower energy and can only penetrate a few millimeters of air, so to measure them the beta emitting material is mixed with a liquid that emits light when struck by an electron and placed in a glass vial. This flash of light is so faint that the signal produced by the PMT is in the same range as that produced by thermionic emissions. To detect it a PMT is placed on each side of the vial and only those pulses that appear on both PMT's at the same time are analyzed. Almost all of the thermionic noise is eliminated since those pulses occur at random times, and only rarely happen at the same time in both PMT's.
Both the sodium iodide detector and the liquid scintillation vial and PMT's were placed inside lead shielding to remove as much of the background and cosmic ray produced counts as possible. In the case of cosmic rays, most of them were of such high energy that they were above the upper window level for both gamma and liquid scintillation counting.
To give some idea of the difference having 2 PMT's in coincident counting mode made, if I counted pulses from individual tubes I would get about 10K-20K counts per minute, and with the same tubes in coincidence counting it would drop to 15 to 20 counts per minute.
Carl is right in that photon energies can not be determined accurately (by a PMT) if they are only energetic enough to cause emission of a few electron from the photo cathode. The more electrons you get initially the more accurate the result. Ain't quantum effects and statistics wonderful.
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
Tracy Allen
www.emesystems.com
Thank you, kwinn. Yes, I can see the wisdom in using the coincidence mode, and my earlier prototype used exactly that and it worked pretty well. Problem is, my ultimate application is bioluminescence and I'm told the photons will probably NOT be emitted in bursts but will simply "rain down faster and faster." The photons won't necessarily be correlated, so coincidence circuitry might actually miss what I'm looking for. So I think finding a good threshold value to discern signal from noise is going to take me a lot of trial and error.
Tracy, thanks for the tip. I'll see if my library can hunt this down.
thanks again,
Mark
http://www.google.com/url?sa=t&source=web&cd=7&sqi=2&ved=0CDwQFjAG&url=http%3A%2F%2Fwww-atoms.physics.wisc.edu%2Fpapers%2F062.singlephoton.pdf&rct=j&q=how%20to%20create%20a%20single%20photon&ei=21BZTcvzLY3ZcYas_NAM&usg=AFQjCNG7yerA3qiVgN4dKKyzIEKaXGBPkg&cad=rja
Creating single-atom and single-photon sources from entangled atomic ensembles
M. Saffman and T. G. Walker
Department of Physics, University of Wisconsin, 1150 University Avenue, Madison, Wisconsin 53706
~Received 14 March 2002; published 16 December 2002!
We discuss the application of dipole blockade techniques for the preparation of single-atom and singlephoton
sources. A deterministic protocol is given for loading a single atom in an optical trap, as well as ejecting
a controlled number of atoms in a desired direction. A single-photon source with an optically controlled
beamlike emission pattern is described.
Source Quote
We turn now to the creation of a phased array single photon
source with a diffraction-limited emission pattern.
When the optical fields propagate through the sample, atoms
will be excited with position dependent relative phases as in
Eq. ~1!. This results in an entangled state with a phase structure
that mimics the phase of the exciting beam and can be
used to create a single-photon source with well-defined directionality.
While previous work has achieved single-photon
sources with a controlled emission direction by coupling to
microcavities @12# our approach results in a source emission
pattern that is reconfigurable and is defined by the structure
of the preparation optical fields. Referring to Fig. 3, fields at
v1 and v2 with Rabi frequencies V1 and V2 drive a twophoton
transition ua&!ue&!ur& to the Rydberg level ur& in
an N-atom ensemble. The effective Rabi frequency for the
two-photon process acting on atom j is uVueif11if2
5(uV1ueif1(rj)uV2ueif2(rj))/2De , where f1 and f2 are the
phases of fields v1 , v2 at the atomic position rj and De
5v12vea . As long as Pdouble given by Eq. ~3! is small only
transitions to states with a single excited atom are energetically
allowed and we have an effective dipole blockade. Under
these conditions an ensemble of atoms in ua& subjected to
a p-pulse applied on ua&!ur& produces the entangled symmetric
superposition state uc&5(2i/AN)(jeı(f1 j1f2 j)ur j&.
The phases fmj5km•rj are simply the phase of the mth laser
field (;eı(km•r2vmt)) at the position rj of the jth atom.
Thanks Humanoido,
I think I happened across that article (or one like it) back when I was trying to figure out how to prove my photon counters were actually working the way I wanted them to. Of course all that single photon generation stuff is way over my head, so I just went with something dumb and simple.
Pretty cool stuff, though.
Ironical... ElectricAye I was museing on the idea of a gieger counter this week .
there EASY to make .
Those would be perfect for the long nights when you are mystified by quantum mechanics,
can't sleep, and just have to do that double slit experiment one more time...just to be certain. :-)
-Phil
-Phil
This may or may not help, but, here is a link to nanosec.. ir, visible, and laser diode avalanche drivers using an ordinary 2N2369A.
http://www.jensign.com/electronics.html
Rick
How fast can you blink an LED?
You can blink it so fast that it appears to stay on and not blink.
How dim can it get?
You can make it so dim that it will appear to be off.
BTW, darkness, as nothing, is not an element for quantification.
Even in the most primordial state, there is a froth.
By definition, one can only create the absence of light.
It is light that is quantifiable.