Beyma 12P80ND + TPL-150

How much of a real world issue is this mismatch in dispersion really? OK it is not ideal, on that we can all agree, but how much does it actually impact the sound negatively, if you could quantify it as a percentage of the overall quality of the presentation? Is it 5% or is it 35%? What does this problem actually sound like in the frequency band where it is a problem, i.e. from where the 12" woofer starts to beam till where the TPL-150H takes over, especially for listeners slightly off-center (e.g. in a home theater environment)? And how sensitive is our hearing at the relevant frequencies - perhaps we can't really tell there's a problem?

These two drivers seem to match each other rather well from a sound quality perspective, and it would be a shame if they shouldn't be used together; So is this gap in their optimal passbands a real deal breaker for the more discerning listener or is its importance completely overblown on forums like these, scaring people off from trying out, on purely academic grounds. what would otherwise be a fantastic listening experience?!

By the way, StigErik used the TPL-150H down to approximately 900 Hz to meet a 15P80Nd in his InConcert Miles, which design a lot of people seem to like and enjoy

There are so many designs, where is a big mismatch of dispersion between the different speakers. See all these horn speaker designs, where the bass is a direct radiator, and above comes a round tractrix or LeCleac'h horn. The bass has wide dispersion, while the horn has a much more narrow dispersion. Many do not have any issue with it......

I really became aware of this, when i changed from the big lower midrange horn in combination with Beyma TPL-150, to 12p80nd , in combination with Beyma TPL-150. The lower midrange horn has a rather narrow dispersion, which does not match with the TPL-150, which has a wide horizontal dispersion. When i switched to 12p80nd in combination with TPL-150, suddenly the dispersion became almost identical, and that has been one of the reasons the 12p80nd combines and sounds much better than the combo with the lower midrange horn. I have also no issue with clarity of the 12p80nd in the crossover range, it sounds very open, precise, and clear. I don't miss anything. And i have no problem with the TPL-150 being crossed at 1,2khz. I don't hear any harshness . It sounds very smooth. I would wonder if the Acoustic Elegance 12" speaker has such a fantastic lower midrange as the 12p80nd. If so, and also good low bass, than that could be a great match as well for a 2 way design.

Angelo
 
My rule of thumb, before polar measurement of each speaker:

10" midbass has ~90degree polar at 1100-1300Hz
12" midbass has ~60degree polar at 1200-1400Hz


There have been a few well reviewed designs using the Acoustic Elegance Lambda TD10M midbass with the Beyma TPL150H 80x30 horn crossed ~1200Hz. A 10" diameter midbass has ~90 degree polar response around 1100-1300Hz based upon cone profile, surround, phase plug, etc... For decades, JBL's 4345 monitors used the JBL 2012H 10" midrange with a 90 degree tweeter horn, plus a 18" woofer. I classify the Lambda TD10M as a "vocal range" speaker since it can cover 80-1100Hz. I own Lambda TD10Ms.

The Levinson Daniel Hertz M1 uses the Beyma 12p80nd with a compression driver on a 60x40 horn, which I "think" is crossed 18db/octave at 1600Hz.
 
My rule of thumb, before polar measurement of each speaker:

10" midbass has ~90degree polar at 1100-1300Hz
12" midbass has ~60degree polar at 1200-1400Hz


There have been a few well reviewed designs using the Acoustic Elegance Lambda TD10M midbass with the Beyma TPL150H 80x30 horn crossed ~1200Hz. A 10" diameter midbass has ~90 degree polar response around 1100-1300Hz based upon cone profile, surround, phase plug, etc...

1. At what frequency do woofers of the following sizes have a 80 degree polar for best match with the TPL-150H:
  • 6.5"
  • 8"
  • 10"
  • 12"
  • 15"
2. What is the beamwidth in degrees (-6dB) of the 12P80Nd @ 1.8kHz?

3. At what frequency does the TPL-150H start to beam and at what frequency would it be optimal, from a directivity point of view, to cross over to a RAAL 70-20X?

The Levinson Daniel Hertz M1 uses the Beyma 12p80nd with a compression driver on a 60x40 horn, which I "think" is crossed 18db/octave at 1600Hz.


To match 60 degrees @ 1.6kHz is obviously easier for the 12P80Nd than the 80 degrees of the TPL-150H @ 1.2kHzx.

4. What is the crossover frequency between the 12P80Nd and the 18P80Nd in the M1 speakers?
 
Coming to this thread late in the proceedings - but here is my contribution, to the debate.

I am using the the TPL crossed with a JBL 2482 in a midrange horn ( xo 1200Hz) changed from a round horn to a rectangular horn - to allow the horns to better integrate physically. I found an improvement in the "togetherness" of the horns. Maybe changing your midrange horns to match the dispersion of the TPL horn could be an alternative approach. I feel that CD's match the dynamics of the AMT's better than cone drivers, especially if you are considering 12". For info - I tried a pair of audax 6.5" PR170MO's in MTM configuration with the TPL and found it did not work for me.
 
Coming to this thread late in the proceedings - but here is my contribution, to the debate.

I am using the the TPL crossed with a JBL 2482 in a midrange horn ( xo 1200Hz) changed from a round horn to a rectangular horn - to allow the horns to better integrate physically. I found an improvement in the "togetherness" of the horns. Maybe changing your midrange horns to match the dispersion of the TPL horn could be an alternative approach. I feel that CD's match the dynamics of the AMT's better than cone drivers, especially if you are considering 12". For info - I tried a pair of audax 6.5" PR170MO's in MTM configuration with the TPL and found it did not work for me.

Interesting. One of the advantages i like with the 12p80nd is the wide frequency range it covers, flat from ~80hz up to 1,2khz, which alouds to cover a wide part of the human voice frequency range. In regard of dynamics, its also a great match, i like it a lot. In the next few weeks, i will try Line Magnetic 555 in my big lower midrange horn - will see, how it will match with TPL-150.

Do you have a picture of your system ?
 
Last edited:
That is a 4-octave passband! Do you really want doppler IMD in the critical voice range?

In case , that would be a problem for all fullrange, wideband, and 2 way speakers. Anyhow, if going to add a sub, it could be crossed a 100hz, to solve this issue......but i do not hear anything that disturbs me. Maibe you could clarify, how these doppler IMD are identified and perceived ?
 
1. At what frequency do woofers of the following sizes have a 80 degree polar for best match with the TPL-150H:

4. What is the crossover frequency between the 12P80Nd and the 18P80Nd in the M1 speakers?


Years back, JBL published a SIMPLE graph which estimated the polar pattern vs. frequency for THEIR speakers. Today, some manufactures plot the polar response for each driver, and this better illustrates how much cone profile, surround, phase plug, etc..., can affect the polar response at different frequencies. (figures)

Today, everyone understands "horn honk" and tries to minimize C-to-C between the midbass speaker and tweeter horn/waveguide.

Today, everyone understands controlled directivity and "has a plan" for how the polar response should be continuously and smoothly shaped for their listening environment from: 20Hz==360degrees, bafflestep==180degrees, tweeterbeam <<90degrees.

Today, because of HT with large flat panels on the wall, "the plan" OFTEN attempts to get the best sould from against-the-wall, or in-the-corner speakers.

----------------------------

4. == Daniel Hertz M1 speaker Xover is reported as 18db/octave at 80Hz.
FWIK: Mark Levinson and Dan Krell and Nelson Pass built expensive Class_A active crossovers. A single Class_A amplifier stage can execute an 18db/octave Xover, plus modest equalization curve shaping.(old Linkwitz papers a good read)
 

Attachments

  • JBL_Polar.jpg
    JBL_Polar.jpg
    83 KB · Views: 802
  • SomePolars x8.jpg
    SomePolars x8.jpg
    108.2 KB · Views: 778
  • SomePolars x10.jpg
    SomePolars x10.jpg
    112.1 KB · Views: 767
  • SomePolars x12.jpg
    SomePolars x12.jpg
    116.8 KB · Views: 761
That is a 4-octave passband! Do you really want doppler IMD in the critical voice range?

There are a few simple Doppler/IMD modeling tools that are said to provide conservative estimates of high_on_low frequency intermodulation distortion.

I attached figures from one tool for a 12" speaker covering 80 or 100Hz midbass up to 1200Hz midhigh range. As you can see, moving from 80Hz to 90Hz or 100Hz provides significant extra IMD margin. -40db is typically reported as the baseline inaudible level. A LR4 Xover at 80Hz removes ~6db of power, so the 100db SPL graph point is conservative.

With these tools, the Classic Altec Voice of the theater with a 15" woofer covering 30Hz up to 700Hz has much less(worst) IMD margin(11db Altec vs. 40db for the 12" modeled @80Hz).

I am a fan of "vocal range" midbass speakers. My HT speakers use the Lambda TD10M from 85Hz to 1200Hz with LR4 slopes. I have experimented with 1000Hz sines mixed with 100Hz sines and do not hear high frequency "flutter" at normal listening levels.
 

Attachments

  • Doppler  80Hz_100Hz.jpg
    Doppler 80Hz_100Hz.jpg
    99.6 KB · Views: 765
Today, everyone understands "horn honk" and tries to minimize C-to-C between the midbass speaker and tweeter horn/waveguide.
At last I have something I can disagree with you on. Unless I'm not part of everyone.:)
Sure I have an idea of what horn honk is, but I haven't tried to understand it.
Is it relevant with the TPL-150H?
What does C-C have to do with horn honk?
 
Demands on a cone

Hi,

As always good well researched info from line source, thanks.

I have seen on this forum a very usefull chart which shows the relationship between frequency and driver travel ie for each octave you go down the cone has to move double or tripple the distance it moves.
Can someone post that chart here, I have lost the link!
Thanks.
Derek.
 
4 times the excursion per octave

Just checked and it is actually 4 times the excursion for each octave you drop down.
I.e. : If a 12 inch driver is being driven to a given SPL, say 110dB at 200Hz, and this requires 2mm of excursion ( plus 1mm then minus 1mm) and you then drop the crossover point 1 octave to 100Hz the same driver needs to move 8mm ( plus 4mm then minus 4mm)...That is significant!
Dropping another octave to 50Hz would of course require 16mm excursion ( plus & minus 8mm).

Now looking at what exactly that means for the music signal / cone movement is interesting.
If you are choosing a two way solution with a crossover point of around 1200Hz here is what you are asking your 12 inch cone to do when playing music.

As the 50Hz bass line is asking for 50 Hz " Happenings per second " the cone needs to stay 100% pistonic and start and stop ( the stopping is most important!!) on a dime. It is pumping away doing its 16mm ( plus / minus 8mm) when simultaneously it is asked to produce some low end piano at 100Hz, then some Cello at 200Hz, then violin at 400 Hz, then some woodwind at 800 Hz and now some vocals.....
Most driver impulse and transient specs and associated claims of linear or pistonic ( the two are very different ) performance use fixed test tones assume that the driver is at REST and starts from the centre point of its XMax travel.
Imagine 1,000 Happenings per second vocal note which requires tiny fast delicate movements of less than one twentieth (!) of the distance the 50Hz bass is demanding...While the cone is pulsing in and out

Now we start to see one reason all the lovely frequency response graphs showing sweet smooth curves are 100% invalid, because they use fixed test tones.
Never mind the laws of physics i.e. you can’t be in two places at one time, how can the cone produce clean 50Hz, 400Hz, 800Hz, 1,000Hz etc. all at the same time ? Answer: IT CANT!!!

Therein lies the root of all our problems, the mass on a spring ( all cone , dome , ribbons or so called "pistonic" drivers ) are so fundamentally flawed all we can hope to do is make the best of an acoustic train wreck.

I hope I don’t sound toooo pessimistic, and despite the above being true we still get some great sounds, buts lets be 100% honest, none of us can design build of have ever heard any system that will win the
" Is it live or is it Memorex" i.e. a live orchestra sounds fab, play the same thing through a PA system or HiFi and it is instantly 100% obvious.

All the best
Derek.
 
Calculating the dispersion for a given driver at a given frequency

djk has provided a clever rule of thumb for calculating the dispersion for a given driver size at a given frequency - the "1,000,000" rule

1,000,000 / Cone size in inches / dispersion in degrees = frequency in Hz

Please note, that "cone size in inches" refer to the actual cone size of the driver, which is usually a little less than the "formal" size for, say, a 10" driver.

For instance, a 10" PHL 3451 mid driver has an actual cone size diameter = 8.27". Thus, the frequency for which the dispersion for the 10" PHL 3451 would be 80° is:

1,000,000 / 8.27" / 80° = 1511.48 Hz

So, if one were to use the PHL 3451 together with the Beyma TPL-150H, and one would like to match the horizontal dispersion around the crossover frequency, then a crossover point around 1511 Hz would be a good starting point.

Please note, that this is only a guide line. Proper measuring of the actual dispersion would be even better of course.

Best regards
Peter



1. At what frequency do woofers of the following sizes have a 80 degree polar for best match with the TPL-150H:
  • 6.5"
  • 8"
  • 10"
  • 12"
  • 15"
2. What is the beamwidth in degrees (-6dB) of the 12P80Nd @ 1.8kHz?

3. At what frequency does the TPL-150H start to beam and at what frequency would it be optimal, from a directivity point of view, to cross over to a RAAL 70-20X?




To match 60 degrees @ 1.6kHz is obviously easier for the 12P80Nd than the 80 degrees of the TPL-150H @ 1.2kHzx.

4. What is the crossover frequency between the 12P80Nd and the 18P80Nd in the M1 speakers?
 
Just checked and it is actually 4 times the excursion for each octave you drop down.
I.e. : If a 12 inch driver is being driven to a given SPL, say 110dB at 200Hz, and this requires 2mm of excursion ( plus 1mm then minus 1mm) and you then drop the crossover point 1 octave to 100Hz the same driver needs to move 8mm ( plus 4mm then minus 4mm)...That is significant!
Dropping another octave to 50Hz would of course require 16mm excursion ( plus & minus 8mm).

Now looking at what exactly that means for the music signal / cone movement is interesting.
If you are choosing a two way solution with a crossover point of around 1200Hz here is what you are asking your 12 inch cone to do when playing music.

As the 50Hz bass line is asking for 50 Hz " Happenings per second " the cone needs to stay 100% pistonic and start and stop ( the stopping is most important!!) on a dime. It is pumping away doing its 16mm ( plus / minus 8mm) when simultaneously it is asked to produce some low end piano at 100Hz, then some Cello at 200Hz, then violin at 400 Hz, then some woodwind at 800 Hz and now some vocals.....
Most driver impulse and transient specs and associated claims of linear or pistonic ( the two are very different ) performance use fixed test tones assume that the driver is at REST and starts from the centre point of its XMax travel.
Imagine 1,000 Happenings per second vocal note which requires tiny fast delicate movements of less than one twentieth (!) of the distance the 50Hz bass is demanding...While the cone is pulsing in and out

Now we start to see one reason all the lovely frequency response graphs showing sweet smooth curves are 100% invalid, because they use fixed test tones.
Never mind the laws of physics i.e. you can’t be in two places at one time, how can the cone produce clean 50Hz, 400Hz, 800Hz, 1,000Hz etc. all at the same time ? Answer: IT CANT!!!

Therein lies the root of all our problems, the mass on a spring ( all cone , dome , ribbons or so called "pistonic" drivers ) are so fundamentally flawed all we can hope to do is make the best of an acoustic train wreck.

I hope I don’t sound toooo pessimistic, and despite the above being true we still get some great sounds, buts lets be 100% honest, none of us can design build of have ever heard any system that will win the
" Is it live or is it Memorex" i.e. a live orchestra sounds fab, play the same thing through a PA system or HiFi and it is instantly 100% obvious.

All the best
Derek.

Then add the effect of the spider and surround...
For any given driver the excursion are limited due to the spider and surround - the bigger the excursion the bigger the limitation....

Due to this I have engineered my own dynamic drivers without these limitations.
 
djk has provided a clever rule of thumb for calculating the dispersion for a given driver size at a given frequency - the "1,000,000" rule

1,000,000 / Cone size in inches / dispersion in degrees = frequency in Hz


So, if one were to use the PHL 3451 together with the Beyma TPL-150H, and one would like to match the horizontal dispersion around the crossover frequency, then a crossover point around 1511 Hz would be a good starting point.

Peter,

Thanks. But where does that formula come from. Seems very arbitrary that the magical number just so happens to be a nicely rounded 1,000,000...?

So 1.5kHz would match a PHL 3451 to the 80 degree horizontal directivity of the TPL-150H. But what about matching the 30 degrees vertical directivity, is that not as important? This goes back to my question of how important a smooth and even power response really is? You would think if this is so critical that you would also want the reflections from the floor and ceiling to have the same spectral makeup as the direct sound - and that all waveguids would be made round... :confused:
 
djk will know from the formula is derived - I don't, I'm afraid.

You are correct, of course, that this does not solve the problem of vertical dispersion. However, although not irrelevant, vertical dispersion may be less important since you move less around in the vertical dimension than in the horizontal (e.g. when you have three listeners sit in a sofa, horizontal dispersion makes a clear difference whereas the horiziontal differences are are minor). Besides, given that ceiling and floor are usually closer to the listener than side walls, you are usually not interested in a wide dispersion in the vertical dimension.

Best regards
Peter



Peter,

Thanks. But where does that formula come from. Seems very arbitrary that the magical number just so happens to be a nicely rounded 1,000,000...?

So 1.5kHz would match a PHL 3451 to the 80 degree horizontal directivity of the TPL-150H. But what about matching the 30 degrees vertical directivity, is that not as important? This goes back to my question of how important a smooth and even power response really is? You would think if this is so critical that you would also want the reflections from the floor and ceiling to have the same spectral makeup as the direct sound - and that all waveguids would be made round... :confused:
 
So 1.5kHz would match a PHL 3451 to the 80 degree horizontal directivity of the TPL-150H. But what about matching the 30 degrees vertical directivity, is that not as important?
The 30 degrees mentioned on the datasheet is average between 1200Hz and 16kHz. Quite misleading, methinks.

The graph on the TPL150H datasheet shows the vertical and horizontal beamwidth is approximately equal below about 1300Hz. At 1500Hz, I would estimate 70 degrees vertical beamwidth. This is one benefit of crossing the TPL150H below 1500Hz.
 
Mass on a spring without any "bouncing"

Hi Rayctech,

At last someone who agrees with me about the elephant in the room!

I have posted a lot about this problem for over 10 years and most forum members think I am mad or stupid (or both !) when I talk about the fundamental problems of conventional cone / dome / ribbons etc.

A mass suspended on a spring will oscillate when set in motion / energised. It will have a natural resonance and settling time.
Its mass, frictional losses / suspension stiffness and motor strength will dictate just how long and how far it bounces around its " centre point" and how long it takes to settle.

Please can you tell me more about your own drivers design and how you have solved the problem?

Thanks in advance and all the best
Derek.
 
Never mind the laws of physics i.e. you can’t be in two places at one time, how can the cone produce clean 50Hz, 400Hz, 800Hz, 1,000Hz etc. all at the same time ? Answer: IT CANT!!!


Do you ever wonder how it can be that signal on your recording media is at different frequency all at same time?

Did your eardrums multiple themselves to be able to follow different tones at same time? Or mic membrane?

Cone doesnt need to be at different place at different time. Cone needs to be just on place where respective signal is. The different frequencies are superimposed one to another and your ear or mic membrane sees just positive or negative change in pressure. There is no multiple things at the same time.