I checked the infrared beam at different position: before the telescope, before the dichoric mirror and before the cavity.
The former two positions, I just put a mirror to reflect the beam and look at it at far field, since the space before the cavity is narrow, so I used the filter to reflect the beam.
For now, the infrared beam looks normal even in far field, but I cannot manage to look at with the CCD.
And I changed the BS mirror after the cavity to another dichoric mirror(CVI 1 inch one) as the picture shows, which reflect infrared and transmit green.
According to Raffaele's advice, I used the signal generator to send a sin wave at modulation frequency but with 100mV amplitude to the oscilloscope, and with a T shape lemo connector plug in and out the RF in port of the demodulation board to see the change of the amplitude. When the RF in port did not connect, the amplitude showed on the oscilloscope is A, when the RF in port plug in, it changes to nearly a quarter of A.
And Manuel found these two PD while clean-up, it can be used in infrared.