by Din » Tue Jul 21, 2015 12:44 pm
Joe Farina wrote:When optics are set up on a table, I suppose that the redirecting mirrors and other components can move or twist in any way during a long exposure(up/down, side to side, or any variation in three-dimensional space). John felt that up/down twisting or warping was most likely with his table. In that case, I would assume that horizontal fringes (made by displacing the two interfering beams slightly in the vertical direction) would show more movement than vertical fringes. Does that make any sense?
It really depends on the vibrational modes of your table.Assuming that all the optics and meccanno are laid out uniformly (that is, no large variation of density of components), then the table has primarily a transverse mode and a longitudinal mode. Which particular mode(s) is (are) activated depends on the disturbance, it's frequency and, of course, it's relative dimensions. There are higher modes, but they generally don't occur unless the frequency of the disturbance is quite high, which is usually unlikely. If your compenents do have a variable density, then you could model the system as if there were two driving forces, the external force and an "internal" force dependant on the weight of the non-uniform distribution of the components.
I've never used a fringe locker, so no personal experience. But, I suspect that Bob is right that the non-uniformity of circular fringes would lend itself to better detection. It is possible to track the fringe motion relative to table motion, since the position of the pth ring in a circular fringe system is proportional to the square root of p. Hence, you can always calculate the phase difference that would make one fringe move by delta(x) relative to the other fringe. Or, you can simply thump the table and watch the fringe motion.
[quote="Joe Farina"]When optics are set up on a table, I suppose that the redirecting mirrors and other components can move or twist in any way during a long exposure(up/down, side to side, or any variation in three-dimensional space). John felt that up/down twisting or warping was most likely with his table. In that case, I would assume that horizontal fringes (made by displacing the two interfering beams slightly in the vertical direction) would show more movement than vertical fringes. Does that make any sense?[/quote]
It really depends on the vibrational modes of your table.Assuming that all the optics and meccanno are laid out uniformly (that is, no large variation of density of components), then the table has primarily a transverse mode and a longitudinal mode. Which particular mode(s) is (are) activated depends on the disturbance, it's frequency and, of course, it's relative dimensions. There are higher modes, but they generally don't occur unless the frequency of the disturbance is quite high, which is usually unlikely. If your compenents do have a variable density, then you could model the system as if there were two driving forces, the external force and an "internal" force dependant on the weight of the non-uniform distribution of the components.
I've never used a fringe locker, so no personal experience. But, I suspect that Bob is right that the non-uniformity of circular fringes would lend itself to better detection. It is possible to track the fringe motion relative to table motion, since the position of the pth ring in a circular fringe system is proportional to the square root of p. Hence, you can always calculate the phase difference that would make one fringe move by delta(x) relative to the other fringe. Or, you can simply thump the table and watch the fringe motion.