That it in a nutshell. It's helping to align the acoustic centers of the drivers. In general the smaller the driver the less far behind the front edge of the frame is the acoustic center. Keep in mind that the crossover networks themselves also create phase shift and delay that are not necessarily complementary at the crossover point, so tilting the baffle is really is of limited use.
Only if you tilt the baffle to point at the ceiling...
it is to avoid a stepped baffle which could have diffraction issues
either is needed to make simple crossovers possible
why would it be fiction
what do you mean by stepped baffle?
offsetting drivers
but google explains it much better than me
and found this thread from here http://www.diyaudio.com/forums/multi-way/237632-hypothetical-stepped-baffles.html
and ofcourse Troels have a done a study on this http://www.troelsgravesen.dk/stepped-baffle.htm
I just don't get you. You start threads and talk/act like you know so much about designing speakers. And how it's such an easy thing to do. Yet you don't know how to use the search feature or Google what a stepped baffle is?
And you tend to state things as facts without knowing a thing about what you are typing. What gives? Google got you blocked or something?
Last edited:
I don't know about slanted baffle, but one obvious side effect is the fact that at least the front and rear are not parallel, which helps avoid standing waves inside the enclosure. Now at the frequencies we are talking about, this is dubious! I am not to sure if it has any effect on this aspect.
Luc
I am not to sure if it has any effect on this aspect.Luc
Crossover can make sound coming faster from any of the driver. Slanting basically solve this issue. It assumes the standard height of stand and a standard height of ears/sofa.
But it doesn't help phase issue in crossover. Good speaker doesn't have limited listening position. In other words it must have good dispersion from the start.
But as has been pointed, slanted baffle is not rectangular baffle, so help from the standing wave etc.
But if your drivers are more than 2, may be it is more critical as you cannot just toe in or out the speaker because what you need is a relative position of the drivers to get the desired dispersion (Look at multi-speaker designs such as the Wilson Alexandria).
The concept of a loudspeaker's acoustic center is a little vaque. The physical acoustic center of a transducer with some diameter, at d/2 distance =r) from the surface of the membrane. Well, most membranes are cones, some domes which makes this a littel difficult. AES PNW Meeting Report - The Acoustic Centre
Often it is said that tweeter has it acoustic center behind the membrane. But this is just a rule of thumb with the prejudice that a woofer's center is at it's membrane level.
Often it is said that tweeter has it acoustic center behind the membrane. But this is just a rule of thumb with the prejudice that a woofer's center is at it's membrane level.
Absolute acoustic centre of a driver is difficult perhaps impossible to measure with absolute certainty, and may not necessarily be constant with frequency at frequencies above the piston range of the driver when the driver is becoming large relative to the wavelength and going through breakup modes.
Relative acoustic centre offset of two drivers which are being crossed over is easy to measure though - measure the "excess group delay" on the design axis of the speaker at a sufficient distance.
The difference in minimum arrival time within each drivers pass band (well away from the crossover points, where the crossover itself will add some additional delay) can be read off directly in fractions of a millisecond and converted to a relative distance offset.
This must be a windowed reflection free measurement with a long enough window time for the phase response within the low frequency drivers passband to be sufficiently accurate.
Last edited:
Unfortunately non-parallel walls reducing or preventing standing waves is a bit of an old wives tale - it doesn't prevent or reduce standing wave formation - it just alters the frequencies that they occur at somewhat.I don't know about slanted baffle, but one obvious side effect is the fact that at least the front and rear are not parallel, which helps avoid standing waves inside the enclosure. Now at the frequencies we are talking about, this is dubious!
Luc
If you have an enclosed space like a cabinet and not much damping you will have standing waves no matter what shape it is, even curved panels. (Same applies to listening rooms)
"vauge", "often it is said". Not really good for engineering practice. This is why we MEASURE the AC of the drivers. You don't get an absolute value, but a delta between two drivers. One could have some reference driver I guess like some planer mid-range.
I still use my old Sound Colony pulser. One of these days I will figure out a pure PC based system.
I still use my old Sound Colony pulser. One of these days I will figure out a pure PC based system.
dsp crossover makes xo phase match/time delay/time alingment "easy".
However this "fine-tuned" alignment applies only to one spesific point in space! My speaker has M/T xo at 3000Hz. Delay is so critical that phasing turns 180¤ with only 0.06ms, which means 55,7mm distance.
"Design axis" is usually between mid and tweeter height and distance 1,5 - 2,5 meters. Spl is usually metered or normalized to 1m. A vertical shift must be quite big to make the phase match roll over, but small deviations come easily. These are masked by more severe phenomenoms like diffractions etc.
3000Hz wavelength is 114,3333...mm (4,5"), it means 360¤ phase rotation.
However this "fine-tuned" alignment applies only to one spesific point in space! My speaker has M/T xo at 3000Hz. Delay is so critical that phasing turns 180¤ with only 0.06ms, which means 55,7mm distance.
"Design axis" is usually between mid and tweeter height and distance 1,5 - 2,5 meters. Spl is usually metered or normalized to 1m. A vertical shift must be quite big to make the phase match roll over, but small deviations come easily. These are masked by more severe phenomenoms like diffractions etc.
3000Hz wavelength is 114,3333...mm (4,5"), it means 360¤ phase rotation.
Last edited:
Digital delay for driver acoustic centre offset will only be correct at one point in space - as you go off the horizontal axis your digital delay is now incorrect. Inverse square law falloff between the two drivers will not track either, if the offset is large enough. (For example a woofer and horn system with a large acoustic centre offset)
Only physical alignment of acoustic centres will keep the two drivers time aligned as you go off the horizontal axis. Of course there will still be an error in the vertical off axis response but maintaining a consistent response on the horizontal axis through the listening window is far more important (especially for imaging) than the vertical off axis response.
Last edited:
"Only physical alignment of acoustic centres will keep the two drivers time aligned as you go off the horizontal axis."
Yes, I can confirm that. But the change is small until we go more than 45¤ off. I measure my dipoles 0-180¤ and phase shift can be seen. At 90¤ the physical distance difference is 0 and dsp-delay turns phase 180¤ -->reversing the phase at 3000Hz! Even between W/M xo a 900Hz suffers but only ~30¤ shift!
It is actually rather difficult to make a 3-way speaker with perfect physical time-alignment, because of the wave length of lowest xo. At least I have not seen one. This phenomenom is the only strong point of single fullrange louspeaker!
Yes, I can confirm that. But the change is small until we go more than 45¤ off. I measure my dipoles 0-180¤ and phase shift can be seen. At 90¤ the physical distance difference is 0 and dsp-delay turns phase 180¤ -->reversing the phase at 3000Hz! Even between W/M xo a 900Hz suffers but only ~30¤ shift!
It is actually rather difficult to make a 3-way speaker with perfect physical time-alignment, because of the wave length of lowest xo. At least I have not seen one. This phenomenom is the only strong point of single fullrange louspeaker!
Attachments
can you provide further support to your statements?
non parallel walls are used in cabinets and by studio engineers to design studios. If it didn't work why go to all the trouble?
- Status
- Not open for further replies.