[nilak (OP)]

I was thinking of a way to send instant signals using entanglement.
Introduction:
Some say the measurement changes the system state. For example I make a measurement on a photon which has an unknown spin, then subsequent measurements for the same axis will show the same spin. We would conclude the first measurement changed the quantum state of the particle. If you do that to entangled particles, you change the state of the whole system. A measurement on the system using this time the other pair, will give the expected result of oposite spin. The system is not in a definite state before measurement which was demonstrated by the absence of hidden variables. Therefore we are not revealing the state of the system, we are interacting with it, yet they say there is no instant cause and effect.
The experiment:
We setup a laser, and using lenses and a BBO crystal we send a pair of photons vertically polarized in opposite directions. The entangled photons will have half of the frequency of the original beam so we can distinguish between entangled photons and not entangled ones. They reach two locations preferably  very far from each other ar different distance from the laser, location A will be closer and B farther. At each site we use a linearly polarized  filter in front of each beam. If A places the filter at 90 degrees it blocks all the radiation and it should change the entangled system so that a following measurement at 0 degrees will result in no detection.
I wonder what happens at location B. Before we place the filter at A all photons are vertically polarized and will pass. Will the filter at A influence what happens at B? I guess not but what would be the answer given by QM. I expect it would give exactly the result we get from the experiment.

[Kryptid]

If you have set the experiment up so that both beams are definitely known to be vertically-polarized, I'm not sure why you think putting one of the beams through a filter would change that fact for either beam. It seems more likely that the filter would simply break the entanglement. You can't force entangled particles to be in one state or another.

[nilak (OP)]

If you have set the experiment up so that both beams are definitely known to be vertically-polarized, I'm not sure why you think putting one of the beams through a filter would change that fact for either beam. It seems more likely that the filter would simply break the entanglement. You can't force entangled particles to be in one state or another.

Yes, you are right. I thought that if you start with a polarized beam, let it go through the BBO crystal the entangled particles will have the initial polarization. But entangled particles always have opposite spin upon measurement. I don't see how this can work.

[Kryptid]

Yes, you are right. I thought that if you start with a polarized beam, let it go through the BBO crystal the entangled particles will have the initial polarization. But entangled particles always have opposite spin upon measurement. I don't see how this can work.

Quantum spin isn't the same as light polarization though, so I don't see how this is a problem.

[nilak (OP)]

"Quantum spin isn't the same as light polarization though, so I don't see how this is a problem."
No it is not, you are right, but the way photons behave when passing through  filters is related to quantum spin. By definition spin 1 corresponds to a left hand circularly polarized photon and -1 for right , (when h/2pi=1 ), and then you can write |H> and |V>  using |L> and |R>. However you can also talk about  photon polarization.

But if you use a linear filter and then send photons through the BBO, I don't think the two pairs will have same polarization between successive pairs. So here must be the problem. 

Apparently a BBO produces SPDC of two types and you can have type I correlations where in fact the polarization will be the same for both photons and type II where they will have perpendicular polarization.