|
95Honda
Guest
|
 |
« Reply #5 on: November 28, 2007, 11:26:31 am » |
|
Err, no and no..
There is no "They" before the filter (crossover), only "The" input signal (it's one thing)....
After the filter, "The" becomes "They" because you have now split the full range signal in to 2 diiferent ranges, what those ranges are is kind of meaningless.
What is meaningful is that "They" or the two ranges now have a phase relationship to each other of 180 degress, or essentially opposite electrical polarity.
The difference between the two signals is 180 degrees, not 180 each. You could look at it like each signal went 90 degrees in another direction......
Basically, after the input signal is divided up by the filter, the relationship of both filters is 180 degrees twards each other. If you look at the O'scope picture of both sines (again, this was taken right in the middle of the crossover region, that is why both sines are the exact same amplitude) are almost perfectly 180 degrees out of phase with each other. This is why you would switch the polarity of one of the drivers to not have a cancellation in this region (two equal waveforms 180 degrees out of phase essentially cancel each other when they are combined). If you didn't reverse the polarity of one driver, at the mid most point you would have a large suckout due to the cancellations, but the farther you move above or below that point, the less effect this has because one of the sines will eventually get bigger than the other and oover power it, a few octaves above or below the crossover region won't really be effected by the phase relationship.
I think you guys may be thinking into this to hard. It is simpler than you are making it. Just think of it as dependant upon filter topology, the different outputs of the filters will not be in phase with each other at the crossover region, sometimes as much as 180 degrees as is the case with 2nd order filters. Now when you start getting differences of 90 and 270 degrees, you may not need to switch polarity of one driver, sometimes you do, but most of the time you don't. The slope of filters and Q can also change the dip or peak in the crossover region, much like a phase differential would.
It all goes back to how voltage and current is transfered through the reactive components in a filter network, in this case capacitors and inductors (resistors are not reactive and don't mess with phase). Without getting too technical, current will lead voltage through a capacitor by 90 degrees, the capacitor resists changes in voltage in essesence (that is why they make good power filters) and voltage leads current by 90 degrees through an inductor becasue inductors resist changes to current (They resist because the current is what makes the magnetic energy or field in the coil, and when it changes, it has to collapse and rebuild the field)... So, you put these charactoristics of these different components together and all kinds of wierd things happen with phase, and it really doesn't just stop with a 2nd order is 180 degrees, etc... There can be all kinds of goof phase relationships, some crossovers have a bagillion parts in them with all kinds of notch filters and eq networks etc, and end up have outputs that are 234.678509 degrees out of phase with each other, and different relationships at different point in the crossover region. So don't get hung up with one simple relationship for each filter. Just remember, when AC current is flowing through filter networks the outputs of the networks may have different phase relationships, and in the case of the classic 2nd order 2-way, as shown above, the relationship at the crossover point with be 180 degrees between the 2 outputs of the filter.....
hope this helps.
|
