Center-to-Center considerations in TMWW

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
As I progress with my TMWW design I want to give proper consideration to center-to-center distances. I realize best practice is to keep c-to-c under 1/4 wavelength at the crossover point, and 1/2 wavelength is also considered acceptable.

My tweeter is a Beyma TPL-150H, which is 23cm high and then makes it imposible to comply with the above. I plan to cross it over somewhere in the 2-2.5kHz range.

Midrange is TBD, and I've drawn a B&W FST (6.5"). I might use an 8" driver though. Lower xo point somewhere in the 350-500Hz range.

Midbass (starting at 80Hz) to be covered with 2 Beyma 10G40. Subwoofers below them.

I've drawn a MTWW and TMWW, with the former looking better and the latter simulated to measure better (I think).

The TPL is kept centered at 95cm from the floor, which is ear level. What's the effect to be expected from having c-to-c in the TPL-to-midrange spaced 20cm when 1/2 the wavelength at 2kHz is 8.6cm? When using an AMT, does it even make sense to think in terms of c-to-c?

Keeping the midrange below the TPL keeps it closer to the midbasses. Am I right in understanding in this case the c-to-c I'm concerned with is the midrange-to-lower midbass? On B) this is 52cm, while 1/2 wavelength even at 350Hz is 49cm...close enough? What is to be expected if I cross-over at 500Hz (1/2 wavelength=34cm)?

FWIW, I used Bagby's Diffraction Simulator to simulate the response from each midbass unit in two baffles with different height. Looks to me B) is better. Would you agree?

I'll go get some MDF tomorrow and start experimenting with the 10G40 boxes. It will be helpful to have a good starting point for baffle height.

Looking forward to your c omments!
 

Attachments

  • Slide1.JPG
    Slide1.JPG
    62.8 KB · Views: 272
  • Slide2.JPG
    Slide2.JPG
    125.7 KB · Views: 270
diyAudio Moderator
Joined 2008
Paid Member
What is to be expected if I cross-over at 500Hz (1/2 wavelength=34cm)?
Nulls appear near the floor and ceiling, and the main lobe narrows as distance increases. Side lobes appear behind the nulls. You have some directivity to your advantage.
What's the effect to be expected from having c-to-c in the TPL-to-midrange spaced 20cm when 1/2 the wavelength at 2kHz is 8.6cm?
Maybe you could low-pass sooner on the lower woofer?
 
Last edited:
--TM-WW has a better appearance. I have seen a few TM-WW using the TPL-150, but I have not seen a MT-WW.
--TM-WW has a shorter/better mid to woofers C-to-C. Don't forget it is desirable to have 1"-2" absorption material on the inside top of the W-cabinet, and this will also increase separation distances.

--I would put each W in a separate volume with only a modest separation distance for wood+absorption material. This should generate the best sound and also provide the greatest flexibility. Separate W volumes prevents interaction if you find it desirable in your experiments to send different signals to the top_W than the bottom_W. Keeping the W-W gap small improves the bass wavefront launch.(B&W research)

--Give some consideration to edge diffraction, especially on the TM. Large radius router or angled edge cuts. (R=1.5" fits 0.5" router)
--Because the TPL150 horn is ~ 9" rectangle, a rectangular top box might perform better than tapered for constant directivity. 2,000Hz has 6.8" wavelength.
 

Attachments

  • Midranges 6_12.jpg
    Midranges 6_12.jpg
    76.4 KB · Views: 228
diyAudio Moderator
Joined 2008
Paid Member
Either. Assuming they are in phase on axis, when they are half wavelength apart on the baffle they will cancel directly above and below. This matters little when the directivity prevents them having much energy at 90 degrees. At greater spacing the angle of the null is smaller.
 
diyAudio Moderator
Joined 2008
Paid Member
Separate W volumes prevents interaction if you find it desirable in your experiments to send different signals to the top_W than the bottom_W
Yes, it makes no difference where the signals are the same. The difference might only be at frequencies where the stuffing will take care of it. In fact, it might by coincidence be a fair choice to use this like a conventional 2.5 way.
 
Either. Assuming they are in phase on axis, when they are half wavelength apart on the baffle they will cancel directly above and below. This matters little when the directivity prevents them having much energy at 90 degrees. At greater spacing the angle of the null is smaller.

In my B) drawing, c-to-c mid-to-1st woofer is 25cm. Let's assume for a second the design had only one woofer. At 690Hz the 1/2 wavelength would be 25cm. So I'm understanding crossing over anywhere under 690Hz should be fine. Correct? IF it had only one woofer.

Now let's keep with the TMW idea, but now M-to-W is measured from the midrange center to the lower woofer center, so 52cm. 52cm corresponds to 1/2 wavelength of 330Hz. So crossing over anywhere above 330Hz would generate cancellations between the lower woofer and the midrange.

What would be the net effect when using TMWW, and crossing over at 500Hz (1/2 wavelength=34.5cm)? The midrange-to-top-woofer isn't causing lobes as they are 25cm apart. But the midrange-to-bottom-woofer is generating lobes. What are ways to estimate the severity of this effect?

Am I worrying too much about a non-issue?
 
Member
Joined 2008
Paid Member
Actually, that was part of the plan for using 2 drivers. Mr. Lewinski had
already calculated that the baffle step would occur in the mid bass section.
I'm pretty sure I also recommended individual enclosures for each driver, as well.

A spot of tea today, for sure :)
 
diyAudio Moderator
Joined 2008
Paid Member
What would be the net effect
If you run the 10s together you'll have close to 50cm of diaphragm height. Above the crossover, they themselves will be lobing (ie narrowing vertical directivity). The mid would be dispersing widely so the W-M lobing at some point would miss the mark and not cause the null, but not for the right reasons as the dispersion of each way is different.

You could draw up a chart of the directivity/nulls vs frequency for each of the W-W and the M-W cases and see where they go and how they overlap.

Also, I'm thinking the baffle widening is going to happen on or higher than the lower crossover.
 
Last edited:
diyAudio Moderator
Joined 2008
Paid Member
Am I worrying too much about a non-issue?
Maybe.

At the upper crossover I'd possibly be more concerned about avoiding the ceiling reflection and less about the total energy involved, for example making options like an absorber possible. Priorities might be different at the lower crossover. The reflection isn't the same deal, the total room energy can be adjusted, the way it gets into the room is a little less critical.
 
Actually, that was part of the plan for using 2 drivers. Mr. Lewinski had
already calculated that the baffle step would occur in the mid bass section.
I'm pretty sure I also recommended individual enclosures for each driver, as well.

A spot of tea today, for sure :)

Hi,

Actually not the case given the earlier questions.
There is no point separating 0.5 way drivers,
but it can be done as an effective brace.

Actually I think you're being somewhat tedious.

rgds, sreten.
 
Maybe.

At the upper crossover I'd possibly be more concerned about avoiding the ceiling reflection and less about the total energy involved, for example making options like an absorber possible. Priorities might be different at the lower crossover. The reflection isn't the same deal, the total room energy can be adjusted, the way it gets into the room is a little less critical.


Making sure I'm tracking you as you mean to: For the T-M crossover the concern is more about the differences in dispersion the two drivers exhibit rather than lobbing? And treating the ceiling to absorb the reflection would be a solution. Correct?

BTW, T-M dispersion at 2 to 2.5kHz xo:
  • The TPL-150H is 80° H x 30° V.
  • a standard (whatever that means!) 6.5" midrange would be about 88° at 2kHz and 70° at 2.5kHz
  • a std 8" would be 68° at 2kHz and 54° at 2.5kHz.
I guess we can only do so much with a tall tweeter like the TPL. Probably everybody using it has to live with a tradeoff regarding c-to-c.
 
If you run the 10s together you'll have close to 50cm of diaphragm height.

OK, let's focus on the M-WW.

Above the crossover, they themselves will be lobing (ie narrowing vertical directivity).
But, by definition, the 10s wouldn't operate above the xo...Maybe I'm misunderstanding?

If the 10s are 27cm c-to-c, then even if the chosen xo point was 500Hz then the 1/2 wavelength would be 34cm so seems all frequencies they reproduce below 500Hz would be fine. Right?

The mid would be dispersing widely so the W-M lobing at some point would miss the mark and not cause the null, but not for the right reasons as the dispersion of each way is different.

You could draw up a chart of the directivity/nulls vs frequency for each of the W-W and the M-W cases and see where they go and how they overlap.
How should I set up that chart?

Also, I'm thinking the baffle widening is going to happen on or higher than the lower crossover.
Baffle widening? Do you mean the baffle step? I'm working with a 40cm wide baffle for the WW, which implies an F3 of 290Hz for the baffle step. In fact the idea behind dual woofers here was to better allow to compensate for the baffle step.
 
diyAudio Moderator
Joined 2008
Paid Member
I plan to cross it over somewhere in the 2-2.5kHz range.
BTW, T-M dispersion at 2 to 2.5kHz xo:
  • The TPL-150H is 80° H x 30° V.
  • a standard (whatever that means!) 6.5" midrange would be about 88° at 2kHz and 70° at 2.5kHz
  • a std 8" would be 68° at 2kHz and 54° at 2.5kHz.
......I guess we can only do so much with a tall tweeter like the TPL. Probably everybody using it has to live with a tradeoff regarding c-to-c.

You're probably right. A couple here sound as if they've discovered a good compromise for a TPL. The shape of the driver is difficult to match in practice but less so where the patterns meet in the ~1200Hz region.
Making sure I'm tracking you as you mean to: For the T-M crossover the concern is more about the differences in dispersion the two drivers exhibit rather than lobbing? And treating the ceiling to absorb the reflection would be a solution. Correct?
I wasn't suggesting you need to absorb the ceiling but making the point that the specific reflections can be somewhat more important at the higher frequencies, and room power can be somewhat more important at lower frequencies. My :2c: based on you using a 6.5" driver that matches the horizontal dispersion better is that a ceiling absorber may get you the result you seek.
 

Attachments

  • Untitled.png
    Untitled.png
    51.8 KB · Views: 116
diyAudio Moderator
Joined 2008
Paid Member
OK, let's focus on the M-WW.

....
How should I set up that chart?
What I can see you trading off is three things. Firstly the 10s, when you run them in parallel may behave like a 20cm x 50cm diaphragm which could give 90 degree vertical directivity around 500Hz which you are trying to match to the 6.5". Crossing an octave below this may be minimally sufficient. Altering the height of the woofer array by cutting one of them out at some frequency could also help.

So the first thing I might want to look at is the directivity of the W and M sections where the mid is likely to be 180 degrees across the region.

The third item is the baffle compensation where you would want to cut one woofer at some frequency related to the baffle. So the question might be, is it possible to cut one woofer at one frequency that helps both of these causes? If so this could be quite a clever design.

One thing I'd want to be able to compromise on, considering that you already have drivers, is the crossover frequency where necessary. Then, if you cannot find a suitable compromise I'd prioritise the first issue.

The reason I'd put that first is because the only benefit I can see to the .5 way concept (apart from spatially) is the elegant simplicity, and the same results can be achieved via the crossover instead. I think I'd just be looking for that good transition frequency at the moment.
 
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.