Posts: 1291 Location: Oregon Joined: December 18, 2012
What's important is that the drivers sum flat through the crossover region(s) on the design/listening axis.
With some crossover types, especially LR2 (12dB/octave, Q=0.5, both drivers -6dB at crossover frequency), the drivers MUST be connected in opposite polarity in order to sum flat. Having them in the same polarity will result in a deep response null at that frequency, due to phase cancellation.
With odd-order Butterworth slopes (6 or 18dB/octave, Q=0.7, both drivers -3dB at crossover) or LR4 (24dB/octave, Q=0.5, both drivers -6dB at crossover), they will sum the flattest with both in the same polarity. Reverse polarity in these instances will lead to a deep null at crossover (LR4) or a broad, shallow dip an octave either side of it (B1 or B3).
And yeah, you can definitely hear the differences. When the polarity is "wrong" for the particular crossover type, the phase error will be heard as a hollowness or disconnectedness in the response.
RE: When multiple drivers are connected in positive acoustic polarity, can you tell..., posted on November 3, 2022 at 06:15:18
You can certainly tell the difference with an impulse excitation measurement. Adjacent drivers are reversed in relative polarity sometimes to support the crossover design used.
You need to elaborate further on your query.
Dave.
RE: When multiple drivers are connected in positive acoustic polarity, can you tell..., posted on November 3, 2022 at 10:08:42
Think you are referring to driver electrical polarity, as crossover topologies can necessitate an electrical inverted polarity to maintain Acoustic polarity through the crossover region of the shared bandpass of two drivers.
The ability to "tell" is highly reliant on the implementation of the crossover and the person. Fellow home constructor acquaintance was very accurate in detecting reversed electrical polarity than myself and other constructors.
I am not particularly good at it, and have tried many different setups and topologies to see if the "textbook requirements" of inverting driver electrical polarity are always the best result. If the null is very narrow and steep the textbook may be less correct, rare though.