I know it may be a property of the diagram rather than the experiment, the top half looks a little like the usual EPR set-up, while the bottom half is a reference path to compare against.

 

The bottom half provides a second entangled beam which can interfere with the first. So long as an observer at A doesn't turn on the "WPD", the probability waves passing through the device from B to C should interfere with each other, creating an interference pattern. When the "WPD" is switched on, the pattern should be destroyed.

The bottom half provides a second entangled beam which can interfere with the first. So long as an observer at A doesn't turn on the "WPD", the probability waves passing through the device from B to C should interfere with each other, creating an interference pattern. When the "WPD" is switched on, the pattern should be destroyed.

 

Er, yes I understood that already (pictures!), but - oh, never mind...

 

I drew the diagram to put things in the 'traditional' context of a diffraction experiment - thought it would clarify the question of whether an interference pattern would form or not. I stared at it and it looks okay (edit: but what would I know?), so you may have a (very worrying) winner.

 

Most of the time, however, I'm aware that everyone is wrong about this kind of thing [QM without supporting math], so I think someone will have to do the math... with matrices and everything.

 

Volenteers?

I'm aware that everyone is wrong about this kind of thing [QM without supporting math'], so I think someone will have to do the math... with matrices and everything.

 

I'm fairly convinced this can't work for whatever reason, and what I'm really looking for is why...

I'm fairly convinced this can't[/i'] work for whatever reason, and what I'm really looking for is why...

 

...and doing the math won't show that?

 

Oddly, right now I don't know if I really care if it works or not. I like to think that people are exploring these kinds of puzzles, but I imagine that - mostly - they're not.

 

I don't like pictures of devices, in general.

 

The diagram I drew removes the down converters, leaving only one photon to detect and an interference pattern at both ends.

 

To me, your experiment is like simultaniously blocking two adjacent slits in my drawing, which obviously kills interference on both sides.

 

So, we're left only with the devices: the down converters...

 

Does it make a difference that you know, even without measuring, that two photons can never arrive simultaniously at one side? Does this prevent interference? I didn't think it did, but...

Someone on another forum suggested that spontaneous collapse of the probability wave would render the device unusable. I say this is false.

 

For example, consider the delayed choice quantum eraser experiment where the "which path" information is erased 50% of the time. The photon stream traces out the combination of the presence and absence of an interference pattern:

 

 

Compare to when which path information is always taken and the interference pattern is completely destroyed:

 

 

The consequence of this is that you have to look at statistical trends of millions or perhaps billions of photons to reliably discern between the two, so countless photons are required to send just a single "bit". However this approach does provide redundancy and reliable transmission even in the wake of spontaneous probability wave collapse.

I assume you're talking about the Scully and Drühl Experiment - haven't read it fully yet, but it seems to be pretty similar to your proposal...

 

Hey, there's an awful lot to read about this from Google! (duh)

 

Thanks for the pointer...