1,161 reputation
411
bio website marty-green.blogspot.com
location Canada
age
visits member for 3 years, 4 months
seen 23 hours ago

Aug
4
answered oscillations of blocks connected by a spring
Aug
4
answered How is a spherical electromagnetic wave emitted from an antenna described in terms of photons?
Aug
1
answered Books that every layman should read
Aug
1
revised Does decoherence explain all instances of wave function collapse?
added 258 characters in body
Jul
31
asked Tracking photon color in Bell experiments
Jul
28
comment Why is the Ritz combination principle incompatible with Classical Mechanics?
I'm glad you liked it. Of course, it's only my interpretation of what Dirac might have meant; I don't know what he was really thinking.
Jul
28
comment Trying to understand the EPR paradox
OK, then I think you must be wrong. You seem to be saying that entangled particles are prepared with spin oriented in the y direction. Instead of |+-> - |-+> (singlet state)you are analyzing the case of |+-> + |-+> (triplet state)? Is that right?
Jul
28
answered Why is the Ritz combination principle incompatible with Classical Mechanics?
Jul
28
answered Trying to understand the EPR paradox
Jul
26
revised Hydrogen transition and photon behavior
added 1326 characters in body
Jul
24
answered Hydrogen transition and photon behavior
Jul
17
comment Coincidence detectors in Bell tests: How close is close enough?
OK, I mixed up micro and milliseconds, so my reading of the graph gives 10 microseconds average between events, within a factor of 2 give or take with what your revised data shows. This may be explainable by normal bunching (non-Poisson distribution of thermal etc. light). I still think the discrepancy (factor of six) in the first set of points is a little large, especially considering how much data you gave me to work with.
Jul
17
comment Coincidence detectors in Bell tests: How close is close enough?
Peter, your new data provides a tantalizing near-confirmation of my analysis of your earlier data. You have a zone 2-3 nsec wide of "entangled" events, and a second zone on the order of 100 times wider of "dark count" coincidences. Ignoring the entangled band, the "dark count" events are, on average about 10 msec apart. But you are missing the one piece of data that would tie it all together: the total elapsed time! 8000 events 10 msec apart should suggest a total elapsed time of 80 sec. This is the calculation which failed to line up in your ealier data set. Do you have the elapsed time?
Jul
17
comment Coincidence detectors in Bell tests: How close is close enough?
Let's disregard the bizarre echo for now as an experimental artifact. I "assume" Poisson statistics because there is no need to explain Poisson behavior...it just "is". The clear deviation from Poisson, which your new data brings to light, means there is some unexplained physics going on. If we extrapolate: 15,000 events within the 3 nsec band, 20,000 events within the 100 nsec band...we get all 400,000 events within an 8000 nsec band, for an average spacing between events of 4 microsec. We should therefore "expect" around 2.5 million detection events...not the 400,000 reported. What gives?
Jul
17
comment Coincidence detectors in Bell tests: How close is close enough?
to clarify my problem: to ballpark the statistics, I take the reported 380,000 Alice events (approx. 40,000 per second) and find them to be spaced on average 25000 nseconds apart. Let's call them dark counts for sake of argument. Then I ask: how often will two adjacent dark counts occur 2.5 nseconds apart? You can see there is a factor of 10,000 here, which leads to an estimate of around 4 dark count coincidences per second. But from the graph I need to justify close to 200 coincidences outside the 1-nanosecond band. So what are they?
Jul
17
comment Coincidence detectors in Bell tests: How close is close enough?
@bjorn well that's my problem. I'm trying to reconcile the total number of Alice events (388860 over 10 seconds) with the observed frequency of counts within the 3-nsec window. If I assume everything outside the 1 n-sec band is due to dark counts, then assuming Poisson there should be close to 4,000,000 Alice events over the 10-second interval. (By my ballpark estimate: I count about 200 events per second between the 1 nsec and 3 nsec band.) So I have a discrepancy of a factor of 10 in understanding the data as presented.
Jul
17
comment Coincidence detectors in Bell tests: How close is close enough?
Point well taken. But you do see what I mean about two overlapping data sets, I hope? The density of events outside the 1-nanosecond band is just too steady to be the tail of a pure Poisson. If not ordinary photons, can I speculate that the entangled photons come in two varieties: singlet and triplet state? Perhaps the triplet state photons have narrower transition bandwidth so they are coherent over a longer time frame?
Jul
17
comment Coincidence detectors in Bell tests: How close is close enough?
Thanks, Peter. Excellent answer, and I'm especially gratified that you weren't put off by my zero "acceptance rate", which is of course now no longer zero. However, the data raises a new question for me: one would like to understand the scatter of points as the overlay of two types of data: entangled and ordinary photons, with each subset of photons governed by its own Poisson statistics. However, a rough count of events within the 3 nanosecond band seems to show about 200 "ordinary" photons per second, and I can't seem to line this up with the expected Poisson numbers...
Jul
16
awarded  Scholar
Jul
16
accepted Coincidence detectors in Bell tests: How close is close enough?