Andrew,
- The self-damping characteristics of a speaker do not determine the level of detail or transparency that can be achieved. Low self-damping causes high Q breakup nodes. Different cone materials have different distortion characteristics. Your Seas L12 driver has low distortion below the breakup node, which results in accurate reproduction and a high level of detail. But a paper or polymer driver can sound equally detailed.
No, you're mistaken, the properties of the loudspeaker cone material do have an effect on reporduction, simply because the amount of flex and absorption of the cone will impact reproduction. A breakup node is simply the point where a cone starts to no longer offer a piston like motion, but parts of the cone start to move individually rather than as a whole. Different materials do indeed offer different breakup characteristics, but this is not solely determined by cone material, that would be a rather shortsighted explanation of what's actually happening. Cone size, shape, thickness, mass, suspension, etc. all contribute to whether it'll have breakup nodes and where they'll show up in the frequency curve.
- Cone breakup does not have to be, and sometimes cannot be kept out of the driver’s passband. Notch filters are used in these situations.
I disagree, the cone breakup node is clearly audible, esp. with cones with low self-damping characteristics due to the high Q of the node, as you pointed out above. Omitting to address this breakup node will give the loudspeaker a specific tonal character. If you cannot get the breakup node of a specific driver in a design out of the passband of interest then you'd best be selecting a different driver. Notch filters can indeed be used if the breakup node is still causing problems, but to my opinion not IN the passband of interest, if that's the solution needed to get it to behave, the wrong driver was chosen for the application.
- If a 3-way speaker is designed correctly there will be no obvious phase shift between drivers.
I never make that statement so explicitly as you put it, I just mention that it is easier to make a well behaved 2-way than a 3-way loudspeaker, it is just a matter of the number of variables you have to work with for a given design.
- Again if a 3-way speaker is correctly designed, there’s no reason to believe that frequency cancellation / amplification will occur more so than with a 2-way design with multiple speakers in a room.
No, for a 5.1-channel setup, as that's what I'm referring to, the least amount of drivers per box is ideal. Actually you'd like to have a full range driver rather than two or more drivers on a baffle. Why? Simple because their voice coils will not line up, causing multiple loudspeakers pointing towards the same spot in a room to exhibit a reproduction that acts much like a comb filter. esp. with 3-way loudspeakers where the bottom woofer is not filtered steeply, hence is allowed to intermodulate with the (out-of-phase) output of the other loudspeakers. I've actually done a fair amount of measurements regarding this, and the effect is quite clear.
- If a subwoofer is integrated correctly with main speakers, there’s no reason to believe the main speakers will sound thin if the crossover is higher than 60Hz. If a subwoofer has a 2nd order crossover, it may be possible to localise low frequencies if the crossover is above about 80Hz, but that is not what you are describing.
I'm sorry? Your point being? This design is meant to also work well without a subwoofer, hence it makes sense to offer enough low-end reproduction to facilitate the use of these loudspeakers without one.
- D’Appolito speakers improve horizontal dispersion when the speaker is positioned vertically. If the speaker is placed horizontally (on top of a TV), this creates narrow horizontal dispersion directly in front of the speaker – very undesirable.
I don't think I mentioned it needed to be placed horizontally, and even then, the polar plots (not posted in the article) show it to offer good off axis response when used horizontally as well due to the small baffle and the fact that the individual drivers are grouped together as closely as I could get then without hacking up the tweeter faceplate.
- It is not possible to achieve the same -3dB point with a sealed box of any size compared to a bass reflex box. A related point is that an optimal sealed box is approximately half the size of an optimal bass reflex box. Note the -3dB point of an optimal sealed box is approximately twice the Fs of the driver.
I suggest you abandon the idea that the -3dB point is the sole determining factor of a good and accurate bass reproduction. The total Q is much more important as is the driver's excursion at low frequencies. Many people wrongly assume that a closed box should have a total Q of 0.7, when if you simulate and measure a Q of 0.5 fares much better in terms of low end extension. Especially when taking the room gain into account a Q of 0.5 will give you far better mileage than 0.7, as the latter will usually end up sounding booming or colored, whereas the 0.5 will in many cases level out to a virtually flat curve.
- Your comment that parallel crossovers are a short-sighted approach is at odds with the speaker design community and seems to be unfounded. Crossovers shape the frequency response of a driver so that the phase and frequency response may integrate well with other drivers. Both series and parallel crossovers perform this function.
They both perform this function yes, that's correct. However I find the end result with a series crossover usually more pleasing and more coherent than a parallel one. Both drivers are part of the other's filter network and hence they can be made to compliment each other much better than with a parallel filter. At least that's my experience, your mileage may vary.
- Series crossovers do not have steeper slopes relative to parallel crossovers of the same order. A first, second, third, or fourth order series crossover gives exactly the same slope as a parallel crossover of the same order.
I think you need to rethink that comment and factor in the acoustical response of both drivers in a 2-way series filter, remember that both drivers are part of the other's filter network.
- A zobel network has no effect on a woofer’s frequency response peak due to cone breakup.
Nonsense, it does just that in this particular application. Simply because the impedance of the L12RCY driver used here rises with frequency, and hence the crossover does not see a fixed impedance, the zobel compensates for this and hence the net result is that the breakup node is also attenuated more than when the zobel would not have been there. It is just one way to address a problem, I could also have used a notch, but find this more elegant as it solves two problems in one fell sweep, it compensates for the rising impedance, and further attenuates the breakup node.
Best regards,
RSS Updates