SY said:
Do you mean it produces a more-spherical radiation pattern, or does it have other advantages over the 3rd order Butterworth?
Also, does "4th order L-R" mean a 4th-order electrical crossover, or a network which produces a 4th order acoustic rolloff? Those terms have me a bit confused, as I see them both in project galleries.
Sorry, need a bit of help with terms (kind of a newbie):
Thanks!
That sounds like it makes the pattern more directive rather than spreading it out evenly in all directions. Do I have that right?
Produces fewer peaks and dips in the polar pattern, right?
I think I lost you there. Does that mean less sensitive to the phase differences between the drivers?
Thanks!
d88, in order:
Yes, that's exactly it, specifically in the vertical plane.
Yes, again. Some people like that sound, others don't. For me, it makes the image much "tighter" and more defined, but it does narrow the sweet spot. And as salas implied, it makes the speaker less room-critical.
"Phase" is a bit tricky. "Acoustic center" is a different way to think about it. There are some nice illustrations of this in Dickason's Loudspeaker Design Cookbook.
Yes, that's exactly it, specifically in the vertical plane.
Yes, again. Some people like that sound, others don't. For me, it makes the image much "tighter" and more defined, but it does narrow the sweet spot. And as salas implied, it makes the speaker less room-critical.
"Phase" is a bit tricky. "Acoustic center" is a different way to think about it. There are some nice illustrations of this in Dickason's Loudspeaker Design Cookbook.
diesel_88 said:
John K's music and design site has some interesting data on MT and MTM polar and power response which seems to show some advantages for odd order Xover acoustic slopes for MTM.
http://www.musicanddesign.com/Power.html
"Here we make a brief comparison of several crossover when used for MTM systems. The first two figure compare a 2K Hz 3rd order Butterworth MTM and MT speaker. The MTM speaker has significantly reduced power response below the crossover frequency due to cancleation between the M drivers. The polar response, at the crossover frequency, is much smoother for the MTM, as was originally demonstrated by D'Appolito. however, we can expect symmetric dips in the polar response below the crossover frequency, again, due to the interference between the M drivers."
The Synergy horns which are very popular here are very similar to a five driver d'appolito, except symmetrical in both the Y AND the X axis.
MTM L-R vs d'Appolito
I use the d'Appolito odd-order MTM for my surrounds, because they are hung from the ceiling and if I was to point them at the listening area instead of square against the wall, my wife would totally rebel.
Problem: they are 1st order, using tweeters doped with ferrofluid to damp the resonance to an almost-flat impedance. I've already burned out one set. The tweeters themselves do not radiate spherically and I tried adding more highs which roasted them. The replacements are aim-able. Still have to adjust the rear tweeters because they aren't in phase at all over a few octaves. 1st order x-overs are trouble.
3rd order or 18db/octave would be best out of the box and as long as the drivers are close enough together you would get that seamless transition anywhere in front. The absolute phase, though, would be off from low to high. Opinions differ on what that does to sound quality.
L-R is Linkwitz-Reilly. In phase. There are nulls, but they fix them so they are where your ears wouldn't be. Same nulls up as down.
I use the d'Appolito odd-order MTM for my surrounds, because they are hung from the ceiling and if I was to point them at the listening area instead of square against the wall, my wife would totally rebel.
Problem: they are 1st order, using tweeters doped with ferrofluid to damp the resonance to an almost-flat impedance. I've already burned out one set. The tweeters themselves do not radiate spherically and I tried adding more highs which roasted them. The replacements are aim-able. Still have to adjust the rear tweeters because they aren't in phase at all over a few octaves. 1st order x-overs are trouble.
3rd order or 18db/octave would be best out of the box and as long as the drivers are close enough together you would get that seamless transition anywhere in front. The absolute phase, though, would be off from low to high. Opinions differ on what that does to sound quality.
L-R is Linkwitz-Reilly. In phase. There are nulls, but they fix them so they are where your ears wouldn't be. Same nulls up as down.
Last edited:
Hi,
One thing to recognise is by definition a MTM's vertical dispersion
will be symmetrical on the tweeter axis, and all vertical lobing
will be symmetrical.
Consequently something that doesn't work well for a MT,
due to asymmetrical lobing, can work in a MTM, and a
pukka D'Appolito has 90 degree phase difference at x/o,
utilising 3rd order Butterworth acoustic slopes.
A more prosaic viewpoint would be MTM's were bound
to take off due to the use of MTM centre channels and
then having matching left and rights. D'Appolito makes
a pretty grim centre channel, but decent left and right.
Whatever, the term "D'Appolito" generically means
a MTM layout in general parlance, and the general
problems of MTM's are less apparent when used
for AV L+R than for discerning music listening.
rgds, sreten.
One thing to recognise is by definition a MTM's vertical dispersion
will be symmetrical on the tweeter axis, and all vertical lobing
will be symmetrical.
Consequently something that doesn't work well for a MT,
due to asymmetrical lobing, can work in a MTM, and a
pukka D'Appolito has 90 degree phase difference at x/o,
utilising 3rd order Butterworth acoustic slopes.
A more prosaic viewpoint would be MTM's were bound
to take off due to the use of MTM centre channels and
then having matching left and rights. D'Appolito makes
a pretty grim centre channel, but decent left and right.
Whatever, the term "D'Appolito" generically means
a MTM layout in general parlance, and the general
problems of MTM's are less apparent when used
for AV L+R than for discerning music listening.
rgds, sreten.
Overlooked Advantages of MTM
A given volume displacement requirement for bass reproduction can be provided using smaller diameter bass drivers. This permits a higher c/o to be used while sustaining matching horzontal as well as vertical radiation patterns in the c/o frequency overlap region. Also by doing this, more system headroom is provided by operating the displacement limited tweeter at a higher high-pass frequency.
Regards,
WHG
A given volume displacement requirement for bass reproduction can be provided using smaller diameter bass drivers. This permits a higher c/o to be used while sustaining matching horzontal as well as vertical radiation patterns in the c/o frequency overlap region. Also by doing this, more system headroom is provided by operating the displacement limited tweeter at a higher high-pass frequency.
Regards,
WHG
I often read these sort of discussions with a complete sense of frustration that diyaudio contributors are shooting in the dark.
There is a lot to be said for the PA-type approach of doubling drivers, but D'Appolito was a flawed idea.
It works much better with MTTM (Two tweeters!) on theoretical grounds. And theoretical is part of what I do.
There are two obvious improvements which can be made to D'Appolito. The first is wiring drivers in series rather than parallel. This avoids the inter-driver resonance that has the drivers in a common box resonating in anti-phase. Think of it this way. Wired in series, both woofers see the same current. The second is doubling the tweeters to give a common 1/R fall-off of power with distance, rather than 1/R squared.
BW3 lobing and D'Appolito combing are overrated as being unpleasant to the ear. They are not.
BW3 is a very nice solution to constant power delivery. Which is something you CAN hear.
Below is a typical BW3 versus LR4 dispersion graph. You can see that BW3 does a much better job near axis with a two-way. It is theoretically even better with an MTTM.
There is a lot to be said for the PA-type approach of doubling drivers, but D'Appolito was a flawed idea.
It works much better with MTTM (Two tweeters!) on theoretical grounds. And theoretical is part of what I do.
There are two obvious improvements which can be made to D'Appolito. The first is wiring drivers in series rather than parallel. This avoids the inter-driver resonance that has the drivers in a common box resonating in anti-phase. Think of it this way. Wired in series, both woofers see the same current. The second is doubling the tweeters to give a common 1/R fall-off of power with distance, rather than 1/R squared.
BW3 lobing and D'Appolito combing are overrated as being unpleasant to the ear. They are not.
BW3 is a very nice solution to constant power delivery. Which is something you CAN hear.
Below is a typical BW3 versus LR4 dispersion graph. You can see that BW3 does a much better job near axis with a two-way. It is theoretically even better with an MTTM.
Attachments
I'm sorry, what?
If you put the drivers in parallel, both drivers will still see the same current, if they're reasonably well-matched.
What do you mean by "resonating in anti-phase"?
Chris
No. They see the same voltage. Individual current to each woofer is undefined! The TOTAL current is, of course constant. Do you get it?
Slightly separately, how BW3 works really, is it gives far more overlap between the drivers. Which means output falls off far less in all directions at crossover, including sideways. BW3 sounds much better than LR4 off-axis.
I'm not sure why you would advocate MTTM as you will typically get bad lobing through the tweeter range. I'm also not seeing a great distinction between your 3rd and 4th order examples.
Hard to generalize between 3rd and 4th order. We shouldn't make any assumptions about phase angle at the crossover point as it becomes quite arbitrary when unknown driver delay is factored in. Also, the rules that apply regarding phase blending at the crossover point don't apply when symmetrical arrays are considered. This was the primary finding of DiAppolito, that a 90 degree phase shift at crossover gave less severe lobes at crossover point.
Hard to generalize between 3rd and 4th order. We shouldn't make any assumptions about phase angle at the crossover point as it becomes quite arbitrary when unknown driver delay is factored in. Also, the rules that apply regarding phase blending at the crossover point don't apply when symmetrical arrays are considered. This was the primary finding of DiAppolito, that a 90 degree phase shift at crossover gave less severe lobes at crossover point.
How can you get any lobing whatsoever with a symmettrical driver layout? MTM or MTTM? Frankly, speaker dave, you're confused! And I am thinking clearly.
Lobing is where the upwards and downwards axis response is different. You KNOW I am right.
Steve, I'm somewhat confused. You appear to be stating that two tweeters in an MTTM configuration are exempt from the issues that they would have in any other situation, that their outputs will sum throughout their passband to behave exactly as if they were a single unit, without any kind of interference pattern (i.e. 'output lobing' / 'comb-filter' or whatever you like to call it) being created per Huygens, Fresnel that varies depending on phase angle from each unit.
Last edited:
How can you get any lobing whatsoever with a symmettrical driver layout? MTM or MTTM? Frankly, speaker dave, you're confused! And I am thinking clearly.
Lobing is where the upwards and downwards axis response is different. You KNOW I am right.
Lobing can be both symmetrical and non-symmetrical. Whenever driver outputs are out of phase at some angles, lobing will occur. Dual tweeters will definitely lobe.
I may be confused but I ain't wrong!
Regards,
David
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.