In the c-band holography tests run yesterday afternoon, ea09C showed a large delay for the first three minutes of the run (48 nsec), after which its delay returned to the normal values. Any explanation?