Its a question about entangled pairs and double slit experiment. I have seen and read many such on the stackexchange but I can not really understand what is happening! So please don't stop the question as a duplicate.
I would like to get an answer about an extremely simple scheme which is close to that of Birgit Dopfer one 1999.
One has a pairs of path-impulse connected photons.
Photons A are headed on to a double slit DS and later on to a CCD (A) to register the pattern.
Photons B are headed on to another CCD (B) and are registered just exactly when photons A passed the DS (or pass).
What I expect is that because the positions of photons B would be well known hence so are the positions of photons A and one knows which way the photons went through the DS. This makes the interference pattern disappear. In this case CCD (A) doesn't show interference picture.
Then one switches CCD B off. Now there is no which way information anymore. So there must emerge an interference pattern.
And most importantly its not necessary to use a coincidence circuit for the photons A and B (as it is in Dopfer experiment). The flux of photons A now is just a common flux of photons which are no more entangled so it would give simply an interference pattern. If this is not the case I would like to understand which of the basic rules of QM prevents it?