photodetector coincidence or anticoincidence ? [Archive]

Patrick Van Esch

Oct12-04, 10:49 AM

<jabberwocky><div class="vbmenu_control"><a href="jabberwocky:;" onClick="newWindow=window.open('','usenetCode','toolbar=no, location=no,scrollbars=yes,resizable=yes,status=no ,width=650,height=400'); newWindow.document.write('<HTML><HEAD><TITLE>Usenet ASCII</TITLE></HEAD><BODY topmargin=0 leftmargin=0 BGCOLOR=#F1F1F1><table border=0 width=625><td bgcolor=midnightblue><font color=#F1F1F1>This Usenet message\'s original ASCII form: </font></td></tr><tr><td width=449><br><br><font face=courier><UL><PRE>\n\nHello,\n\nI\'m far from an expert in quantum optics, and recently I had a\ndiscussion with someone who claims that all optical EPR kinds of\nresults are reproducible with classical EM (no need for photons) - in\nthat we never need a "photon".\nHe claims moreover that photodetectors such as photomultipliers,\ngenerate clicks (photoelectrons) as a function of the incident\nintensity.\n\nI proposed the following setup to discriminate between his views and\nquantum theory:\nTake a parametric down converted photon pair (a red photon and a blue\nphoton, say), and send the red one directly on a PM. Send the blue\nphoton on an arborescence of beam splitters so that we get N equally\nintense split beams with the blue one, and place a PM at each end.\nNow my claim is that quantum theory predicts that in the limit of low\nenough intensities (so that the probability of having 2 pairs of\nphotons in the setup during an observation time is as low as one\nwishes), only ONE (or 0, because of low efficiency) of the blue PMs\nwill click at a time in each short coincidence window which is opened\nby the detection of a red one, while his view will give drawings which\nare a combinatorial distribution with equal probabilities (1/N x\nefficiency) for all detectors at once.\nThis is of course a rather naive prediction of quantum theory, where\nthe photon is a marble and has to choose his detection channel.\n\nBut the guy claims that quantum theory makes exactly the same\nprediction as he, namely that we will have independent detection\nevents as a function of the incident intensity which has been divided\nby N, and that the above experiment doesn\'t distinguish quantum\npredictions from classical ones (and hence that we do not need the\nconcept of photon). He then goes on saying that I\'m confusing Glauber\ncorrelations with detection correlations and so on, and that this has\nbeen worked out long ago in quantum optics, by people like Santos (he\ndoes have some papers on the Arxiv). I\'m at loss there.\n\nSo my questions to the learned lot here:\n1) Does quantum theory (when taking into account all possible effects\nof an actual detection) predict, or not, that in the low intensity\nlimit we will have strict anticorrelation between the N blue detectors\n?\n\n2) Has any such experiment been performed ?\n\n\nthanks a lot,\n\nPatrick.\n</UL></PRE></font></td></tr></table></BODY><HTML>');"> <IMG SRC=/images/buttons/ip.gif BORDER=0 ALIGN=CENTER ALT="View this Usenet post in original ASCII form">  View this Usenet post in original ASCII form </a></div><P></jabberwocky>Hello,

I'm far from an expert in quantum optics, and recently I had a
discussion with someone who claims that all optical EPR kinds of
results are reproducible with classical EM (no need for photons) - in
that we never need a "photon".
He claims moreover that photodetectors such as photomultipliers,
generate clicks (photoelectrons) as a function of the incident
intensity.

I proposed the following setup to discriminate between his views and
quantum theory:
Take a parametric down converted photon pair (a red photon and a blue
photon, say), and send the red one directly on a PM. Send the blue
photon on an arborescence of beam splitters so that we get N equally
intense split beams with the blue one, and place a PM at each end.
Now my claim is that quantum theory predicts that in the limit of low
enough intensities (so that the probability of having 2 pairs of
photons in the setup during an observation time is as low as one
wishes), only ONE (or 0, because of low efficiency) of the blue PMs
will click at a time in each short coincidence window which is opened
by the detection of a red one, while his view will give drawings which
are a combinatorial distribution with equal probabilities (1/N x
efficiency) for all detectors at once.
This is of course a rather naive prediction of quantum theory, where
the photon is a marble and has to choose his detection channel.

But the guy claims that quantum theory makes exactly the same
prediction as he, namely that we will have independent detection
events as a function of the incident intensity which has been divided
by N, and that the above experiment doesn't distinguish quantum
predictions from classical ones (and hence that we do not need the
concept of photon). He then goes on saying that I'm confusing Glauber
correlations with detection correlations and so on, and that this has
been worked out long ago in quantum optics, by people like Santos (he
does have some papers on the Arxiv). I'm at loss there.

So my questions to the learned lot here:
1) Does quantum theory (when taking into account all possible effects
of an actual detection) predict, or not, that in the low intensity
limit we will have strict anticorrelation between the N blue detectors
?

2) Has any such experiment been performed ?

thanks a lot,

Patrick.

carlip-nospam@physics.ucdavis.edu

Oct12-04, 01:22 PM

<jabberwocky><div class="vbmenu_control"><a href="jabberwocky:;" onClick="newWindow=window.open('','usenetCode','toolbar=no, location=no,scrollbars=yes,resizable=yes,status=no ,width=650,height=400'); newWindow.document.write('<HTML><HEAD><TITLE>Usenet ASCII</TITLE></HEAD><BODY topmargin=0 leftmargin=0 BGCOLOR=#F1F1F1><table border=0 width=625><td bgcolor=midnightblue><font color=#F1F1F1>This Usenet message\'s original ASCII form: </font></td></tr><tr><td width=449><br><br><font face=courier><UL><PRE>\n\nPatrick Van Esch <vanesch@ill.fr> wrote:\n\n[...]\n> I\'m far from an expert in quantum optics, and recently I had a\n> discussion with someone who claims that all optical EPR kinds of\n> results are reproducible with classical EM (no need for photons)\n> - in that we never need a "photon".\n\n> He claims moreover that photodetectors such as photomultipliers,\n> generate clicks (photoelectrons) as a function of the incident\n> intensity.\n\nI am also not an expert on quantum optics. But for this, look at\nDavis and Mandel, in _Coherence and Quantum Optics_ (ed. Mandel and\nWolf, 1973). This is a very careful observation of ``prompt\'\'\nelectrons in the photoelectric effect, that is, electrons emitted\nbefore there would be time, in a wave model, for enough energy to\nbuild up (even over the entire cathode!) to overcome the potential\nbarrier. The experiment shows that the energy is *not* delivered\ncontinuously to a photodetector, and that this cannot be explained\nsolely by detector properties unless one is prepared to give up\nenergy conservation.\n\n> I proposed the following setup to discriminate between his views and\n> quantum theory:\n\n> Take a parametric down converted photon pair (a red photon and a blue\n> photon, say), and send the red one directly on a PM. Send the blue\n> photon on an arborescence of beam splitters so that we get N equally\n> intense split beams with the blue one, and place a PM at each end.\n> Now my claim is that quantum theory predicts that in the limit of low\n> enough intensities (so that the probability of having 2 pairs of\n> photons in the setup during an observation time is as low as one\n> wishes), only ONE (or 0, because of low efficiency) of the blue PMs\n> will click at a time in each short coincidence window which is opened\n> by the detection of a red one, while his view will give drawings which\n> are a combinatorial distribution with equal probabilities (1/N x\n> efficiency) for all detectors at once.\n\nThis is the experiment of Grangier et al., Europhys. Lett. 1 (1986) 173.\nYour prediction is verified.\n\n[...]\n> 1) Does quantum theory (when taking into account all possible effects\n> of an actual detection) predict, or not, that in the low intensity\n> limit we will have strict anticorrelation between the N blue detectors?\n\nIt does. Of course, if the intensity is higher, there will be some\naccidental coincidences, but these will decrease in a predictable\nmanner as the intensity decreases.\n\n> 2) Has any such experiment been performed ?\n\nIt has. Grangier et al. used an atomic cascade as a trigger rather than\na parametric down converted photon pair, but it\'s been redone your way;\nsee Thorne et al., Am. J. Phys. 72 (September 2004) 1210. There\'s a\nreprint on the Web, along with a description of how to do the experiment\nyourself, at http://marcus.whitman.edu/~beckmk/QM/grangier/grangier.html.\n\nSteve Carlip\n</UL></PRE></font></td></tr></table></BODY><HTML>');"> <IMG SRC=/images/buttons/ip.gif BORDER=0 ALIGN=CENTER ALT="View this Usenet post in original ASCII form">  View this Usenet post in original ASCII form </a></div><P></jabberwocky>Patrick Van Esch <vanesch@ill.fr> wrote:

[...]
> I'm far from an expert in quantum optics, and recently I had a
> discussion with someone who claims that all optical EPR kinds of
> results are reproducible with classical EM (no need for photons)
> - in that we never need a "photon".

> He claims moreover that photodetectors such as photomultipliers,
> generate clicks (photoelectrons) as a function of the incident
> intensity.

I am also not an expert on quantum optics. But for this, look at
Davis and Mandel, in _Coherence and Quantum Optics_ (ed. Mandel and
Wolf, 1973). This is a very careful observation of ``prompt''
electrons in the photoelectric effect, that is, electrons emitted
before there would be time, in a wave model, for enough energy to
build up (even over the entire cathode!) to overcome the potential
barrier. The experiment shows that the energy is *not* delivered
continuously to a photodetector, and that this cannot be explained
solely by detector properties unless one is prepared to give up
energy conservation.

> I proposed the following setup to discriminate between his views and
> quantum theory:

> Take a parametric down converted photon pair (a red photon and a blue
> photon, say), and send the red one directly on a PM. Send the blue
> photon on an arborescence of beam splitters so that we get N equally
> intense split beams with the blue one, and place a PM at each end.
> Now my claim is that quantum theory predicts that in the limit of low
> enough intensities (so that the probability of having 2 pairs of
> photons in the setup during an observation time is as low as one
> wishes), only ONE (or 0, because of low efficiency) of the blue PMs
> will click at a time in each short coincidence window which is opened
> by the detection of a red one, while his view will give drawings which
> are a combinatorial distribution with equal probabilities (1/N x
> efficiency) for all detectors at once.

This is the experiment of Grangier et al., Europhys. Lett. 1 (1986) 173.
Your prediction is verified.

[...]
> 1) Does quantum theory (when taking into account all possible effects
> of an actual detection) predict, or not, that in the low intensity
> limit we will have strict anticorrelation between the N blue detectors?

It does. Of course, if the intensity is higher, there will be some
accidental coincidences, but these will decrease in a predictable
manner as the intensity decreases.

> 2) Has any such experiment been performed ?

It has. Grangier et al. used an atomic cascade as a trigger rather than
a parametric down converted photon pair, but it's been redone your way;
see Thorne et al., Am. J. Phys. 72 (September 2004) 1210. There's a
reprint on the Web, along with a description of how to do the experiment
yourself, at http://marcus.whitman.edu/~beckmk/QM/grangier/grangier.html.

Steve Carlip

Joe Rongen

Oct13-04, 02:13 PM

<jabberwocky><div class="vbmenu_control"><a href="jabberwocky:;" onClick="newWindow=window.open('','usenetCode','toolbar=no, location=no,scrollbars=yes,resizable=yes,status=no ,width=650,height=400'); newWindow.document.write('<HTML><HEAD><TITLE>Usenet ASCII</TITLE></HEAD><BODY topmargin=0 leftmargin=0 BGCOLOR=#F1F1F1><table border=0 width=625><td bgcolor=midnightblue><font color=#F1F1F1>This Usenet message\'s original ASCII form: </font></td></tr><tr><td width=449><br><br><font face=courier><UL><PRE>\n"Patrick Van Esch" <vanesch@ill.fr> wrote in message\nnews:c23e597b.0410122300.35380958@posting .google.com...\n>\n> carlip-nospam@physics.ucdavis.edu wrote in message\nnews:<ckh6k0\\$8e8\\$2@skeeter.ucdavis.ed u>...\n>\n> > It has. Grangier et al. used an atomic cascade as a trigger rather than\n> > a parametric down converted photon pair, but it\'s been redone your way;\n> > see Thorne et al., Am. J. Phys. 72 (September 2004) 1210. There\'s a\n> > reprint on the Web, along with a description of how to do the experiment\n> > yourself, at\nhttp://marcus.whitman.edu/~beckmk/QM/grangier/grangier.html.\n>\n>\n> Thanks a lot. From September 2004 !! I\'m in sync with what happens\n> at the frontlines of current research in quantum optics :-))\n> The paper addresses indeed exactly what I was aiming at. A pity one\n> has to be subscribed (my institute is) to read it.\n\n\nOne can download another project for free here (1.2 Mb):\n\nhttp://departments.colgate.edu/physics/research/Photon/ajpphtwf.pdf\n\n"Photon Quantum Mechanics\nThis is a project funded by the National Science Foundation to establish\na set of undergraduate laboratories to study the fundamentals of quantum\nmechanics. The experiments are laboratory exercises on topics of quantum\nmechanics that are otherwise theoretical, abstract or even unintuitive. The\ncentral issue of the experiments is quantum superposition: the ability of a\nquantum to be in two places at the same time or to be in to be in a\ncorrelated superposition of states with other quanta."\n\nRegarding their detectors:\n\nhttp://departments.colgate.edu/physics/research/Photon/detectors.htm\n\n"The photon detectors need to have high efficiency because coincidence\ndetection depends on the product of the efficiencies on the two detectors.\nPhotomultipliers are not good enough because their efficiencies are low\nin the near infrared. For this reason everybody uses avalanche photodiodes."\n\nStated under Fig. 1\n"Single photon detectors in its enclosure.\nWe have about 300 s-1 dark counts."\n\nIs that not a very high dark count to start with?\n\nRegards Joe\n\n\n" ...not only are the drops of rain mere appearances, but even their\nround shape, and even the space in which they fall, are nothing in\nthemselves, but merely modifications of fundamental forms of our\nsensible intuition, and the transcendental object remains unknown to\nus"\n\nKant, I. Critique of Pure Reason.\n\n\n---\nOutgoing mail is certified Virus Free.\nChecked by AVG anti-virus system (http://www.grisoft.com).\nVersion: 6.0.776 / Virus Database: 523 - Release Date: 10/12/04\n</UL></PRE></font></td></tr></table></BODY><HTML>');"> <IMG SRC=/images/buttons/ip.gif BORDER=0 ALIGN=CENTER ALT="View this Usenet post in original ASCII form">  View this Usenet post in original ASCII form </a></div><P></jabberwocky>"Patrick Van Esch" <vanesch@ill.fr> wrote in message
news:c23e597b.0410122300.35380958@posting.google.c om...
>
> carlip-nospam@physics.ucdavis.edu wrote in message
news:<ckh6k0$8e8$2@skeeter.ucdavis.edu>...
>
> > It has. Grangier et al. used an atomic cascade as a trigger rather than
> > a parametric down converted photon pair, but it's been redone your way;
> > see Thorne et al., Am. J. Phys. 72 (September 2004) 1210. There's a
> > reprint on the Web, along with a description of how to do the experiment
> > yourself, at
http://marcus.whitman.edu/~beckmk/QM/grangier/grangier.html.
>
>
> Thanks a lot. From September 2004 !! I'm in sync with what happens
> at the frontlines of current research in quantum optics :-))
> The paper addresses indeed exactly what I was aiming at. A pity one
> has to be subscribed (my institute is) to read it.

One can download another project for free here (1.2 Mb):

http://departments.colgate.edu/physics/research/Photon/ajpphtwf.pdf

"Photon Quantum Mechanics
This is a project funded by the National Science Foundation to establish
a set of undergraduate laboratories to study the fundamentals of quantum
mechanics. The experiments are laboratory exercises on topics of quantum
mechanics that are otherwise theoretical, abstract or even unintuitive. The
central issue of the experiments is quantum superposition: the ability of a
quantum to be in two places at the same time or to be in to be in a
correlated superposition of states with other quanta."

Regarding their detectors:

http://departments.colgate.edu/physics/research/Photon/detectors.htm

"The photon detectors need to have high efficiency because coincidence
detection depends on the product of the efficiencies on the two detectors.
Photomultipliers are not good enough because their efficiencies are low
in the near infrared. For this reason everybody uses avalanche photodiodes."

Stated under Fig. 1
"Single photon detectors in its enclosure.
We have about 300 s-1 dark counts."

Is that not a very high dark count to start with?

Regards Joe

" ...not only are the drops of rain mere appearances, but even their
round shape, and even the space in which they fall, are nothing in
themselves, but merely modifications of fundamental forms of our
sensible intuition, and the transcendental object remains unknown to
us"

Kant, I. Critique of Pure Reason.

---
Outgoing mail is certified Virus Free.
Checked by AVG anti-virus system (http://www.grisoft.com).
Version: 6..776 / Virus Database: 523 - Release Date: 10/12/04

Patrick Van Esch

Oct14-04, 11:07 AM

<jabberwocky><div class="vbmenu_control"><a href="jabberwocky:;" onClick="newWindow=window.open('','usenetCode','toolbar=no, location=no,scrollbars=yes,resizable=yes,status=no ,width=650,height=400'); newWindow.document.write('<HTML><HEAD><TITLE>Usenet ASCII</TITLE></HEAD><BODY topmargin=0 leftmargin=0 BGCOLOR=#F1F1F1><table border=0 width=625><td bgcolor=midnightblue><font color=#F1F1F1>This Usenet message\'s original ASCII form: </font></td></tr><tr><td width=449><br><br><font face=courier><UL><PRE>\n"Joe Rongen" <joe@alpha.to> wrote in message news:<001b01c4b12d\\$a7d4c3c0\\$2723fea9@joerongen >...\n\n> Stated under Fig. 1\n> "Single photon detectors in its enclosure.\n> We have about 300 s-1 dark counts."\n>\n> Is that not a very high dark count to start with?\n\nCompared to a photomultiplier, it is. However, it is no issue if you\ndo coincidence counting, because the probability of generating\ncoincidences of dark counts is extremely low (in a time window of\n10ns, you have a probability of 3.0 10^(-6) to have another spurious\nhit), so as long as you\'re not looking at precisions of that order,\nyou\'re fine...\n\ncheers,\nPatrick.\n</UL></PRE></font></td></tr></table></BODY><HTML>');"> <IMG SRC=/images/buttons/ip.gif BORDER=0 ALIGN=CENTER ALT="View this Usenet post in original ASCII form">  View this Usenet post in original ASCII form </a></div><P></jabberwocky>"Joe Rongen" <joe@\alpha.to> wrote in message news:<001b01c4b12d$a7d4c3c0$2723fea9@joerongen>...

> Stated under Fig. 1
> "Single photon detectors in its enclosure.
> We have about 300 s-1 dark counts."
>
> Is that not a very high dark count to start with?

Compared to a photomultiplier, it is. However, it is no issue if you
do coincidence counting, because the probability of generating
coincidences of dark counts is extremely low (in a time window of
10ns, you have a probability of 3. 10^(-6) to have another spurious
hit), so as long as you're not looking at precisions of that order,
you're fine...

cheers,
Patrick.

Oz

Oct17-04, 01:46 PM

<jabberwocky><div class="vbmenu_control"><a href="jabberwocky:;" onClick="newWindow=window.open('','usenetCode','toolbar=no, location=no,scrollbars=yes,resizable=yes,status=no ,width=650,height=400'); newWindow.document.write('<HTML><HEAD><TITLE>Usenet ASCII</TITLE></HEAD><BODY topmargin=0 leftmargin=0 BGCOLOR=#F1F1F1><table border=0 width=625><td bgcolor=midnightblue><font color=#F1F1F1>This Usenet message\'s original ASCII form: </font></td></tr><tr><td width=449><br><br><font face=courier><UL><PRE>\n\n[NB I reply to Patrick\'s post at the end]\n\ncarlip-nospam@physics.ucdavis.edu writes\n>I am also not an expert on quantum optics. But for this, look at\n>Davis and Mandel, in _Coherence and Quantum Optics_ (ed. Mandel and\n>Wolf, 1973). This is a very careful observation of ``prompt\'\'\n>electrons in the photoelectric effect, that is, electrons emitted\n>before there would be time, in a wave model, for enough energy to\n>build up (even over the entire cathode!) to overcome the potential\n>barrier. The experiment shows that the energy is *not* delivered\n>continuously to a photodetector, and that this cannot be explained\n>solely by detector properties unless one is prepared to give up\n>energy conservation.\n\nOh no!\n<visual of oz in small plane trailing flames & heading for ground...>\n\nI\'m not concerned that emissions are not continuous (in the sense that\nelectrons are emitted randomly) that fits my model.\n\nI do have a problem with photons emitted before enough energy had been\ndumped into the detector.\n\nLet me think about this for a minute....\n\nOK. Consider this:\n\n1) A very low amplitude light beam. Lets say it is equivalent to one\nemitted electron/minute.\n\n2) The photodetector operates a shutter that cuts off the light beam as\nsoon as an electron is detected.\n\nIn this scenario we can have two options:\n\na) No electrons are detected until about enough energy is deposited in\nthe detector.\n\nb) Electrons are emitted even though not enough energy appears to have\nbeen transferred to the detector.\n\nI do have a horrible feeling that (b) is not unreasonable but would\ndepend on the *source*. The entire system is in fact\n\nsource -> em wave -> detector\n\nThe source will have a characteristic emission time.\nThat is each atom will take some number of nanosecs to emit a packet of\nem wave and so the attenuated wave is in fact a series of packets of\nlength (in time) characterised by the source. In this case its easy to\nsee that electron emission will not follow the apparent average\nintensity, but depend on the packet size. So that\'s dealt with.\n\nNow, if this also happens with a long-coherence-length laser (and it may\nwell do so) then the above argument will in part fail. I assume that it\nis known that the electric field of such a laser, when attenuated to\nthese very low levels, is in fact of constant amplitude and does not in\npractice fluctuate significantly. I suspect that it might because its\ngenerated by a large number of atoms in the laser, each of which\ndeposits a wavepacket in phase and the total averages to a laser beam of\nvery highly constant amplitude. This could of course be measured by\nseeing if the diffraction pattern of such a laser degrades slightly at\nextreme attenuation. I imagine this work was done decades ago. I have a\nnasty feeling that its been found that it doesn\'t. In that case I have\nto take my argument further (this post is too long, if need be I will do\nso later).\n\n========Patrick posts========\nCompared to a photomultiplier, it is. However, it is no issue if you\ndo coincidence counting, because the probability of generating\ncoincidences of dark counts is extremely low (in a time window of\n10ns, you have a probability of 3.0 10^(-6) to have another spurious\nhit), so as long as you\'re not looking at precisions of that order,\nyou\'re fine...\n============================\n\nIn my model of a detector this level of dark counts is a worry.\nI model a detector as a large number of resonant systems that have an\nenergy gap and a detection occurs when an electron (in this case) jumps\nthe gap due to energy from the incident em wave. The em wave and all the\natoms in the detector are taken to become entangled. The high dark\ncurrent implies that these systems are already rather excited to the\nextent that a number jump the gap spontaneously. One would thus expect,\npurely from maxwell, that an incident em field at the right frequency\nwould cause a small increase in this spontaneous jumping with the energy\nlargely provided from the circuit. That is it would be similar to\nheating the device, which does much the same thing (but given the\nfrequency range, less efficiently).\n\n--\nOz\nThis post is worth absolutely nothing and is probably fallacious.\n\nUse oz@farmeroz.port995.com [ozacoohdb@despammed.com functions].\nBTOPENWORLD address has ceased. DEMON address has ceased.\n</UL></PRE></font></td></tr></table></BODY><HTML>');"> <IMG SRC=/images/buttons/ip.gif BORDER=0 ALIGN=CENTER ALT="View this Usenet post in original ASCII form">  View this Usenet post in original ASCII form </a></div><P></jabberwocky>[NB I reply to Patrick's post at the end]

carlip-nospam@physics.ucdavis.edu writes
>I am also not an expert on quantum optics. But for this, look at
>Davis and Mandel, in _Coherence and Quantum Optics_ (ed. Mandel and
>Wolf, 1973). This is a very careful observation of ``prompt''
>electrons in the photoelectric effect, that is, electrons emitted
>before there would be time, in a wave model, for enough energy to
>build up (even over the entire cathode!) to overcome the potential
>barrier. The experiment shows that the energy is *not* delivered
>continuously to a photodetector, and that this cannot be explained
>solely by detector properties unless one is prepared to give up
>energy conservation.

Oh no!
<visual of oz in small plane trailing flames & heading for ground...>

I'm not concerned that emissions are not continuous (in the sense that
electrons are emitted randomly) that fits my model.

I do have a problem with photons emitted before enough energy had been
dumped into the detector.

Let me think about this for a minute....

OK. Consider this:

1) A very low amplitude light beam. Lets say it is equivalent to one
emitted electron/minute.

2) The photodetector operates a shutter that cuts off the light beam as
soon as an electron is detected.

In this scenario we can have two options:

a) No electrons are detected until about enough energy is deposited in
the detector.

b) Electrons are emitted even though not enough energy appears to have
been transferred to the detector.

I do have a horrible feeling that (b) is not unreasonable but would
depend on the *source*. The entire system is in fact

source -> em wave -> detector

The source will have a characteristic emission time.
That is each atom will take some number of nanosecs to emit a packet of
em wave and so the attenuated wave is in fact a series of packets of
length (in time) characterised by the source. In this case its easy to
see that electron emission will not follow the apparent average
intensity, but depend on the packet size. So that's dealt with.

Now, if this also happens with a long-coherence-length laser (and it may
well do so) then the above argument will in part fail. I assume that it
is known that the electric field of such a laser, when attenuated to
these very low levels, is in fact of constant amplitude and does not in
practice fluctuate significantly. I suspect that it might because its
generated by a large number of atoms in the laser, each of which
deposits a wavepacket in phase and the total averages to a laser beam of
very highly constant amplitude. This could of course be measured by
seeing if the diffraction pattern of such a laser degrades slightly at
extreme attenuation. I imagine this work was done decades ago. I have a
nasty feeling that its been found that it doesn't. In that case I have
to take my argument further (this post is too long, if need be I will do
so later).

========Patrick posts========
Compared to a photomultiplier, it is. However, it is no issue if you
do coincidence counting, because the probability of generating
coincidences of dark counts is extremely low (in a time window of
10ns, you have a probability of 3. 10^(-6) to have another spurious
hit), so as long as you're not looking at precisions of that order,
you're fine...
============================

In my model of a detector this level of dark counts is a worry.
I model a detector as a large number of resonant systems that have an
energy gap and a detection occurs when an electron (in this case) jumps
the gap due to energy from the incident em wave. The em wave and all the
atoms in the detector are taken to become entangled. The high dark
current implies that these systems are already rather excited to the
extent that a number jump the gap spontaneously. One would thus expect,
purely from maxwell, that an incident em field at the right frequency
would cause a small increase in this spontaneous jumping with the energy
largely provided from the circuit. That is it would be similar to
heating the device, which does much the same thing (but given the
frequency range, less efficiently).

--
Oz
This post is worth absolutely nothing and is probably fallacious.

Use oz@farmeroz.port995.com [ozacoohdb@despammed.com functions].
BTOPENWORLD address has ceased. DEMON address has ceased.

Thomas Trotter

Oct20-04, 03:18 AM

<jabberwocky><div class="vbmenu_control"><a href="jabberwocky:;" onClick="newWindow=window.open('','usenetCode','toolbar=no, location=no,scrollbars=yes,resizable=yes,status=no ,width=650,height=400'); newWindow.document.write('<HTML><HEAD><TITLE>Usenet ASCII</TITLE></HEAD><BODY topmargin=0 leftmargin=0 BGCOLOR=#F1F1F1><table border=0 width=625><td bgcolor=midnightblue><font color=#F1F1F1>This Usenet message\'s original ASCII form: </font></td></tr><tr><td width=449><br><br><font face=courier><UL><PRE>\n\nvanesch@ill.fr (Patrick Van Esch) wrote in message news:<c23e597b.0410111952.1f797b23@posting.google. com>...\n[snip]\n\nPatrick Van Esch (PVE) wrote:\nHe claims moreover that photodetectors such as photomultipliers,\ngenerate clicks (photoelectrons) as a function of the incident\nintensity.\n\nTT:\nThey do, don\'t they?\n\nPVE:\nTake a parametric down converted photon pair (a red photon and a blue\nphoton, say), and send the red one directly on a PM. Send the blue\nphoton on an arborescence of beam splitters so that we get N equally\nintense split beams with the blue one, and place a PM at each end.\n\nTT:\nThe split beams aren\'t equally intense.\n\nPVE:\nNow my claim is that quantum theory predicts that in the limit of low\nenough intensities (so that the probability of having 2 pairs of\nphotons in the setup during an observation time is as low as one\nwishes), only ONE (or 0, because of low efficiency) of the blue PMs\nwill click at a time in each short coincidence window which is opened\nby the detection of a red one, while his view will give drawings which\nare a combinatorial distribution with equal probabilities (1/N x\nefficiency) for all detectors at once.\n\nTT:\nYour views don\'t seem to be at odds. In the low intensity limit, if\none of the split beams is intense enough to generate a detection,\nthen the other won\'t be.\n\nThe probability for either one or the other to register is 1/N x efficiency.\n\nPVE:\nThis is of course a rather naive prediction of quantum theory, where\nthe photon is a marble and has to choose his detection channel.\n\nTT:\nThe photon is a detection probability distribution, isn\'t it?\n\nI have a question: can these considerations be related to\nthe conundrum of coincidental detection in Bell tests?\n\nThat is, for some coincidence or correlation window triggered\nby a detection at A, then the probability of detection at B is\ncos^2 theta (where theta is the angle between the transmission\naxes of the linear polarizers at A and B).\n\nBut what if the common polarization axis (Lambda) of the opposite\nmoving light beams was not originally along the transmission\naxis of the polarizer at A for the coincidence window being\nconsidered?\n\nHow can probability of detection at B be given as\ncos^2 |b-Lambda| (where b is the transmission axis of the linear\npolarizer at B, and Lambda is the common polarization axis\nof the beams incident on the polarizers at A and B), and also\nbe given as cos^2 |b-a|, following a detection at A, when a (the\ntransmission axis of the polarizer at A) is different from Lambda?\n</UL></PRE></font></td></tr></table></BODY><HTML>');"> <IMG SRC=/images/buttons/ip.gif BORDER=0 ALIGN=CENTER ALT="View this Usenet post in original ASCII form">  View this Usenet post in original ASCII form </a></div><P></jabberwocky>vanesch@ill.fr (Patrick Van Esch) wrote in message news:<c23e597b.0410111952.1f797b23@posting.google.com>...
[snip]

Patrick Van Esch (PVE) wrote:
He claims moreover that photodetectors such as photomultipliers,
generate clicks (photoelectrons) as a function of the incident
intensity.

TT:
They do, don't they?

PVE:
Take a parametric down converted photon pair (a red photon and a blue
photon, say), and send the red one directly on a PM. Send the blue
photon on an arborescence of beam splitters so that we get N equally
intense split beams with the blue one, and place a PM at each end.

TT:
The split beams aren't equally intense.

PVE:
Now my claim is that quantum theory predicts that in the limit of low
enough intensities (so that the probability of having 2 pairs of
photons in the setup during an observation time is as low as one
wishes), only ONE (or 0, because of low efficiency) of the blue PMs
will click at a time in each short coincidence window which is opened
by the detection of a red one, while his view will give drawings which
are a combinatorial distribution with equal probabilities (1/N x
efficiency) for all detectors at once.

TT:
Your views don't seem to be at odds. In the low intensity limit, if
one of the split beams is intense enough to generate a detection,
then the other won't be.

The probability for either one or the other to register is 1/N x efficiency.

PVE:
This is of course a rather naive prediction of quantum theory, where
the photon is a marble and has to choose his detection channel.

TT:
The photon is a detection probability distribution, isn't it?

I have a question: can these considerations be related to
the conundrum of coincidental detection in Bell tests?

That is, for some coincidence or correlation window triggered
by a detection at A, then the probability of detection at B is
cos^2 \theta (where \theta is the angle between the transmission
axes of the linear polarizers at A and B).

But what if the common polarization axis (\Lambda) of the opposite
moving light beams was not originally along the transmission
axis of the polarizer at A for the coincidence window being
considered?

How can probability of detection at B be given as
cos^2 |b-\Lambda| (where b is the transmission axis of the linear
polarizer at B, and \Lambda is the common polarization axis
of the beams incident on the polarizers at A and B), and also
be given as cos^2 |b-a|, following a detection at A, when a (the
transmission axis of the polarizer at A) is different from \Lambda?

Patrick Van Esch

Oct20-04, 10:48 AM

<jabberwocky><div class="vbmenu_control"><a href="jabberwocky:;" onClick="newWindow=window.open('','usenetCode','toolbar=no, location=no,scrollbars=yes,resizable=yes,status=no ,width=650,height=400'); newWindow.document.write('<HTML><HEAD><TITLE>Usenet ASCII</TITLE></HEAD><BODY topmargin=0 leftmargin=0 BGCOLOR=#F1F1F1><table border=0 width=625><td bgcolor=midnightblue><font color=#F1F1F1>This Usenet message\'s original ASCII form: </font></td></tr><tr><td width=449><br><br><font face=courier><UL><PRE>\n\nthomastrotter2005@juno.com (Thomas Trotter) wrote in message news:<21970122.0410191020.7704e36@posting.google.c om>...\n\n> PVE:\n> Take a parametric down converted photon pair (a red photon and a blue\n> photon, say), and send the red one directly on a PM. Send the blue\n> photon on an arborescence of beam splitters so that we get N equally\n> intense split beams with the blue one, and place a PM at each end.\n>\n> TT:\n> The split beams aren\'t equally intense.\n\nWell, that\'s not how a beam splitter (half-silvered mirror) is\nsupposed to be working, no ? You assume that the beams ARE equally\nintense in a classic sense. Otherwise how would you explain\ninterference when you recombine beams which were split with a beam\nsplitter ?\n\ncheers,\nPatrick.\n</UL></PRE></font></td></tr></table></BODY><HTML>');"> <IMG SRC=/images/buttons/ip.gif BORDER=0 ALIGN=CENTER ALT="View this Usenet post in original ASCII form">  View this Usenet post in original ASCII form </a></div><P></jabberwocky>thomastrotter2005@juno.com (Thomas Trotter) wrote in message news:<21970122.0410191020.7704e36@posting.google.com>...

> PVE:
> Take a parametric down converted photon pair (a red photon and a blue
> photon, say), and send the red one directly on a PM. Send the blue
> photon on an arborescence of beam splitters so that we get N equally
> intense split beams with the blue one, and place a PM at each end.
>
> TT:
> The split beams aren't equally intense.

Well, that's not how a beam splitter (half-silvered mirror) is
supposed to be working, no ? You assume that the beams ARE equally
intense in a classic sense. Otherwise how would you explain
interference when you recombine beams which were split with a beam
splitter ?

cheers,
Patrick.

Thomas Trotter

Oct21-04, 02:47 AM

<jabberwocky><div class="vbmenu_control"><a href="jabberwocky:;" onClick="newWindow=window.open('','usenetCode','toolbar=no, location=no,scrollbars=yes,resizable=yes,status=no ,width=650,height=400'); newWindow.document.write('<HTML><HEAD><TITLE>Usenet ASCII</TITLE></HEAD><BODY topmargin=0 leftmargin=0 BGCOLOR=#F1F1F1><table border=0 width=625><td bgcolor=midnightblue><font color=#F1F1F1>This Usenet message\'s original ASCII form: </font></td></tr><tr><td width=449><br><br><font face=courier><UL><PRE>\n\nvanesch@ill.fr (Patrick Van Esch) wrote in message news:<c23e597b.0410200630.5fcfa006@posting.google. com>...\n> thomastrotter2005@juno.com (Thomas Trotter) wrote in message news:<21970122.0410191020.7704e36@posting.google.c om>...\n>\n> > PVE:\n> > Take a parametric down converted photon pair (a red photon and a blue\n> > photon, say), and send the red one directly on a PM. Send the blue\n> > photon on an arborescence of beam splitters so that we get N equally\n> > intense split beams with the blue one, and place a PM at each end.\n> >\n> > TT:\n> > The split beams aren\'t equally intense.\n>\n> Well, that\'s not how a beam splitter (half-silvered mirror) is\n> supposed to be working, no ? You assume that the beams ARE equally\n> intense in a classic sense. Otherwise how would you explain\n> interference when you recombine beams which were split with a beam\n> splitter ?\n\nThe recombined beams are from the same source.\n</UL></PRE></font></td></tr></table></BODY><HTML>');"> <IMG SRC=/images/buttons/ip.gif BORDER=0 ALIGN=CENTER ALT="View this Usenet post in original ASCII form">  View this Usenet post in original ASCII form </a></div><P></jabberwocky>vanesch@ill.fr (Patrick Van Esch) wrote in message news:<c23e597b.0410200630.5fcfa006@posting.google.com>...
> thomastrotter2005@juno.com (Thomas Trotter) wrote in message news:<21970122.0410191020.7704e36@posting.google.com>...
>
> > PVE:
> > Take a parametric down converted photon pair (a red photon and a blue
> > photon, say), and send the red one directly on a PM. Send the blue
> > photon on an arborescence of beam splitters so that we get N equally
> > intense split beams with the blue one, and place a PM at each end.
> >
> > TT:
> > The split beams aren't equally intense.
>
> Well, that's not how a beam splitter (half-silvered mirror) is
> supposed to be working, no ? You assume that the beams ARE equally
> intense in a classic sense. Otherwise how would you explain
> interference when you recombine beams which were split with a beam
> splitter ?

The recombined beams are from the same source.

Patrick Van Esch

Oct22-04, 12:01 PM

<jabberwocky><div class="vbmenu_control"><a href="jabberwocky:;" onClick="newWindow=window.open('','usenetCode','toolbar=no, location=no,scrollbars=yes,resizable=yes,status=no ,width=650,height=400'); newWindow.document.write('<HTML><HEAD><TITLE>Usenet ASCII</TITLE></HEAD><BODY topmargin=0 leftmargin=0 BGCOLOR=#F1F1F1><table border=0 width=625><td bgcolor=midnightblue><font color=#F1F1F1>This Usenet message\'s original ASCII form: </font></td></tr><tr><td width=449><br><br><font face=courier><UL><PRE>\nthomastrotter2005@juno.com (Thomas Trotter) wrote in message news:<21970122.0410202108.158ae885@posting.google. com>...\n> vanesch@ill.fr (Patrick Van Esch) wrote in message news:<c23e597b.0410200630.5fcfa006@posting.google. com>...\n> > thomastrotter2005@juno.com (Thomas Trotter) wrote in message news:<21970122.0410191020.7704e36@posting.google.c om>...\n> >\n> > > PVE:\n> > > Take a parametric down converted photon pair (a red photon and a blue\n> > > photon, say), and send the red one directly on a PM. Send the blue\n> > > photon on an arborescence of beam splitters so that we get N equally\n> > > intense split beams with the blue one, and place a PM at each end.\n> > >\n> > > TT:\n> > > The split beams aren\'t equally intense.\n> >\n> > Well, that\'s not how a beam splitter (half-silvered mirror) is\n> > supposed to be working, no ? You assume that the beams ARE equally\n> > intense in a classic sense. Otherwise how would you explain\n> > interference when you recombine beams which were split with a beam\n> > splitter ?\n>\n> The recombined beams are from the same source.\n\nWell, I\'m also talking about the same beam (the "blue" beam from the\ndown conversion).\nAnd, BTW, the red and the blue beam ARE equally intense, in that they\ncorrespond to a 2-photon state. But that\'s not a necessary\nassumption for this experiment ; in fact it follows from the result of\nthe experiment (well, if you perform it twice, when interchanging the\nred and the blue beams).\n\ncheers,\npatrick.\n</UL></PRE></font></td></tr></table></BODY><HTML>');"> <IMG SRC=/images/buttons/ip.gif BORDER=0 ALIGN=CENTER ALT="View this Usenet post in original ASCII form">  View this Usenet post in original ASCII form </a></div><P></jabberwocky>thomastrotter2005@juno.com (Thomas Trotter) wrote in message news:<21970122.0410202108.158ae885@posting.google.com>...
> vanesch@ill.fr (Patrick Van Esch) wrote in message news:<c23e597b.0410200630.5fcfa006@posting.google.com>...
> > thomastrotter2005@juno.com (Thomas Trotter) wrote in message news:<21970122.0410191020.7704e36@posting.google.com>...
> >
> > > PVE:
> > > Take a parametric down converted photon pair (a red photon and a blue
> > > photon, say), and send the red one directly on a PM. Send the blue
> > > photon on an arborescence of beam splitters so that we get N equally
> > > intense split beams with the blue one, and place a PM at each end.
> > >
> > > TT:
> > > The split beams aren't equally intense.
> >
> > Well, that's not how a beam splitter (half-silvered mirror) is
> > supposed to be working, no ? You assume that the beams ARE equally
> > intense in a classic sense. Otherwise how would you explain
> > interference when you recombine beams which were split with a beam
> > splitter ?
>
> The recombined beams are from the same source.

Well, I'm also talking about the same beam (the "blue" beam from the
down conversion).
And, BTW, the red and the blue beam ARE equally intense, in that they
correspond to a 2-photon state. But that's not a necessary
assumption for this experiment ; in fact it follows from the result of
the experiment (well, if you perform it twice, when interchanging the
red and the blue beams).

cheers,
patrick.

FrediFizzx

Oct24-04, 09:04 AM

<jabberwocky><div class="vbmenu_control"><a href="jabberwocky:;" onClick="newWindow=window.open('','usenetCode','toolbar=no, location=no,scrollbars=yes,resizable=yes,status=no ,width=650,height=400'); newWindow.document.write('<HTML><HEAD><TITLE>Usenet ASCII</TITLE></HEAD><BODY topmargin=0 leftmargin=0 BGCOLOR=#F1F1F1><table border=0 width=625><td bgcolor=midnightblue><font color=#F1F1F1>This Usenet message\'s original ASCII form: </font></td></tr><tr><td width=449><br><br><font face=courier><UL><PRE>\n\n<carlip-nospam@physics.ucdavis.edu> wrote in message\nnews:ckh6k0\\$8e8\\$2@skeeter.ucdavis.edu ...\n|\n|\n| Patrick Van Esch <vanesch@ill.fr> wrote:\n\n[snip]\n| > 2) Has any such experiment been performed ?\n|\n| It has. Grangier et al. used an atomic cascade as a trigger rather than\n| a parametric down converted photon pair, but it\'s been redone your way;\n| see Thorne et al., Am. J. Phys. 72 (September 2004) 1210. There\'s a\n| reprint on the Web, along with a description of how to do the experiment\n| yourself, at http://marcus.whitman.edu/~beckmk/QM/grangier/grangier.html.\n\nI bought this referenced article for \\$18 online and it has a model for\nphotodetection but none for the beam splitter (BS). I am wondering how the\nBS works without destroying the original photons? Does anyone here know or\nhave a good reference for a model (both classical and quantum) of the BS?\n\nFrediFizzx\n\n</UL></PRE></font></td></tr></table></BODY><HTML>');"> <IMG SRC=/images/buttons/ip.gif BORDER=0 ALIGN=CENTER ALT="View this Usenet post in original ASCII form">  View this Usenet post in original ASCII form </a></div><P></jabberwocky><carlip-nospam@physics.ucdavis.edu> wrote in message
news:ckh6k0$8e8$2@skeeter.ucdavis.edu...
|
|
| Patrick Van Esch <vanesch@ill.fr> wrote:

[snip]
| > 2) Has any such experiment been performed ?
|
| It has. Grangier et al. used an atomic cascade as a trigger rather than
| a parametric down converted photon pair, but it's been redone your way;
| see Thorne et al., Am. J. Phys. 72 (September 2004) 1210. There's a
| reprint on the Web, along with a description of how to do the experiment
| yourself, at http://marcus.whitman.edu/~beckmk/QM/grangier/grangier.html.

I bought this referenced article for $18 online and it has a model for
photodetection but none for the beam splitter (BS). I am wondering how the
BS works without destroying the original photons? Does anyone here know or
have a good reference for a model (both classical and quantum) of the BS?

FrediFizzx

carlip-nospam@physics.ucdavis.edu

Oct27-04, 10:55 AM

<jabberwocky><div class="vbmenu_control"><a href="jabberwocky:;" onClick="newWindow=window.open('','usenetCode','toolbar=no, location=no,scrollbars=yes,resizable=yes,status=no ,width=650,height=400'); newWindow.document.write('<HTML><HEAD><TITLE>Usenet ASCII</TITLE></HEAD><BODY topmargin=0 leftmargin=0 BGCOLOR=#F1F1F1><table border=0 width=625><td bgcolor=midnightblue><font color=#F1F1F1>This Usenet message\'s original ASCII form: </font></td></tr><tr><td width=449><br><br><font face=courier><UL><PRE>\n\nThomas Trotter <thomastrotter2005@juno.com> wrote:\n\n\n> vanesch@ill.fr (Patrick Van Esch) wrote in message\n> news:<c23e597b.0410111952.1f797b23@posting.google. com>...\n\n[...]\n\n> PVE:\n> Take a parametric down converted photon pair (a red photon and a blue\n> photon, say), and send the red one directly on a PM. Send the blue\n> photon on an arborescence of beam splitters so that we get N equally\n> intense split beams with the blue one, and place a PM at each end.\n\n> TT:\n> The split beams aren\'t equally intense.\n\n> [...] In the low intensity limit, if\n> one of the split beams is intense enough to generate a detection,\n> then the other won\'t be.\n\nThis is testable, by recombining the beams and looking at the contrast\nin the resulting interference pattern.\n\nNote first that in the Grangier experiment, this interference takes\nplace "one photon at a time." If the terminology is begging the\nquestion, you can translate it into classical language -- the shutter\nis open a very short time, and does not reopen until well after the\nlight-travel time of a pulse of light through the apparatus, so the\ninterference takes place one pulse at a time, where each "pulse" leads\nto a single "photon" detection.\n\nIf the a given pulse/photon is unevenly divided at the beam-splitter,\nthen the interference upon recombination won\'t be complete. At an\ninterference minimum, for instance, the weaker beam will be able to\ncancel only part of the stronger beam, and will leave part uncanceled.\nAs a result, the "fringe visibility" -- the difference between count\nrates at the maxima and minima, normalized to the total count rate --\nwill be less than 100%.\n\nGrangier et al. check this, by building a Mach-Zehnder interferometer\naround their beam-splitter. They find a fringe contrast of 98.7%,\nthat is, very nearly complete destructive interference at the minima.\nThat means that the two beams *can\'t* be very different in intensity:\nat most, they differ by a few percent.\n\nNow, you could argue that even a very small difference is enough.\nBut then you are asking us to swallow an amazing coincidence, that\nthe photomultipliers used by Grangier et al. just happen to have a\nsharp threshold of almost exactly half the intensity of a pulse\nemitted by a single atomic transition in their light source. And\nsince the experiment has been repeated with different light sources\nand different detectors, you would have to ask us to believe that\n*those* detectors happen to have a sharp threshold of just half the\nintensity of a "one-photon" event from *those* light sources.\n\nThis seems to be clutching at straws...\n\nSteve Carlip\n</UL></PRE></font></td></tr></table></BODY><HTML>');"> <IMG SRC=/images/buttons/ip.gif BORDER=0 ALIGN=CENTER ALT="View this Usenet post in original ASCII form">  View this Usenet post in original ASCII form </a></div><P></jabberwocky>Thomas Trotter <thomastrotter2005@juno.com> wrote:

> vanesch@ill.fr (Patrick Van Esch) wrote in message
> news:<c23e597b.0410111952.1f797b23@posting.google.com>...

[...]

> PVE:
> Take a parametric down converted photon pair (a red photon and a blue
> photon, say), and send the red one directly on a PM. Send the blue
> photon on an arborescence of beam splitters so that we get N equally
> intense split beams with the blue one, and place a PM at each end.

> TT:
> The split beams aren't equally intense.

> [...] In the low intensity limit, if
> one of the split beams is intense enough to generate a detection,
> then the other won't be.

This is testable, by recombining the beams and looking at the contrast
in the resulting interference pattern.

Note first that in the Grangier experiment, this interference takes
place "one photon at a time." If the terminology is begging the
question, you can translate it into classical language -- the shutter
is open a very short time, and does not reopen until well after the
light-travel time of a pulse of light through the apparatus, so the
interference takes place one pulse at a time, where each "pulse" leads
to a single "photon" detection.

If the a given pulse/photon is unevenly divided at the beam-splitter,
then the interference upon recombination won't be complete. At an
interference minimum, for instance, the weaker beam will be able to
cancel only part of the stronger beam, and will leave part uncanceled.
As a result, the "fringe visibility" -- the difference between count
rates at the maxima and minima, normalized to the total count rate --
will be less than 100%.

Grangier et al. check this, by building a Mach-Zehnder interferometer
around their beam-splitter. They find a fringe contrast of 98.7%,
that is, very nearly complete destructive interference at the minima.
That means that the two beams *can't* be very different in intensity:
at most, they differ by a few percent.

Now, you could argue that even a very small difference is enough.
But then you are asking us to swallow an amazing coincidence, that
the photomultipliers used by Grangier et al. just happen to have a
sharp threshold of almost exactly half the intensity of a pulse
emitted by a single atomic transition in their light source. And
since the experiment has been repeated with different light sources
and different detectors, you would have to ask us to believe that
*those* detectors happen to have a sharp threshold of just half the
intensity of a "one-photon" event from *those* light sources.

This seems to be clutching at straws...

Steve Carlip