Triple Mid-Woofer design questions?

[mmoeller]

I have been investigating the design of the Revel Performa F52 tower

http://www.revelspeakers.com/products/product-overview.asp?ID=25

speaker. They use 3 6.5" mid-woofer drivers, ported to the rear. These drivers seem to be able to extend into the 30's, but being smaller drivers also would allow them to extend upward further than most larger woofers. The tower also includeds a 5.25" midrange. My question deals with the meshing of these. Revel publishes that the speaker is a 3-way with a low crossover at about 200hz.

Is that crossover point higher than larger woofers would require?

Could, with a well behaved 5.25" driver, a 2.5-way setup be a good use of the capabilities of the 6.5" mid-woofers?

I understand that the solution would depend on numerous factors, but I am curious about the seemingly prevalent use of multiple mid-woofer drivers in tower type speakers. What are the real benifits? Size? Small baffle? Upper end woofer range?

[Kal Rubinson]

Is that crossover point higher than larger woofers would require?

Depends on the 'larger' woofer. You can run 8" that high, 10" mebbe.

Could, with a well behaved 5.25" driver, a 2.5-way setup be a good use of the capabilities of the 6.5" mid-woofers?

Why? You would have to deal with the interaction of the drivers at midrange frequencies.

I understand that the solution would depend on numerous factors, but I am curious about the seemingly prevalent use of multiple mid-woofer drivers in tower type speakers. What are the real benifits? Size? Small baffle? Upper end woofer range?

Narrow baffle.

Kal

[mmoeller]

Why? You would have to deal with the interaction of the drivers at midrange frequencies.

The Revel lists the upper crossover at about 2.3kHz. Would it be an issue for the 5.25" midrange to span the gap from 200Hz to the 2.3kHz?

Would it be wise to attempt to use a midrange that would have a higher crossover point, allowing that the tweeter require that? Meaning that a tweeter that extended well past 20k may well need a higher xover than 2.3k. In this type of set up would a 2.5-way be more manageable? or Is the 2.5 way just a way to save a couple of bucks on xover parts? I'm just being hypothetical. I'm not sure it even makes sense.

[Kal Rubinson]

The Revel lists the upper crossover at about 2.3kHz. Would it be an issue for the 5.25" midrange to span the gap from 200Hz to the 2.3kHz?

No.

Would it be wise to attempt to use a midrange that would have a higher crossover point, allowing that the tweeter require that?

I generally prefer the lowest possible crossovers for all drivers as long as there is adequate power handling. My reason is that driving beaming increases with frequency.

Meaning that a tweeter that extended well past 20k may well need a higher xover than 2.3k.

Not necessarily since we are speaking in generalities.

In this type of set up would a 2.5-way be more manageable? or Is the 2.5 way just a way to save a couple of bucks on xover parts? I'm just being hypothetical. I'm not sure it even makes sense.

IMHO, 2.5way speakers are not to save on xover parts but on drivers. It is a way to get adequate power handling in the bass by using both drivers and to minimize driver interaction in the mids by rolling-off the lower mid-woofer while not actually paying for 2 woofers plus a midrange.

Kal

[Rudy Jakubin]

Reading their pdf file I see a problem.
Let's break it down.
9 1/2" baffle width = 480 hz & 2855 hz slope to be corrected.
High order crossover's?
They'd have to overlap the woofers and underlap the mid for crossover point of 200 hz.
Same for the high end - 2.3 khz.
13560 / 5.5" = 2465 hz.
13560 / 6.5" = 2086 hz.

They would have had to listen quite a bit to be satisfied by the 'human voice' being transferred and blended in the critical fundamental frequency range.

My current project resembles theirs, only it's 2.5 way sealed. Almost the same size except mine is 24" deep. Using 7" Dayton Aluminum woofers and North D25 tweeter (F3 about 48 hz);

[Dennis H]

Speaking in generalities, not about the Revel, it's hard to wire up 3 drivers effectively. Wire the 3 in parallel and the impedance dips quite low. Wire them in series and the sensitivity is low, about the same as using a single driver. I suspect Revel had some custom 16 ohm drivers built so they could wire them in parallel. Quality 16 ohm drivers are hard for the DIY type to come by so we're stuck with 2 drivers or 4 drivers as better options in most cases.

[mmoeller]

Speaking in generalities, not about the Revel, it's hard to wire up 3 drivers effectively. Wire the 3 in parallel and the impedance dips quite low. Wire them in series and the sensitivity is low, about the same as using a single driver. I suspect Revel had some custom 16 ohm drivers built so they could wire them in parallel. Quality 16 ohm drivers are hard for the DIY type to come by so we're stuck with 2 drivers or 4 drivers as better options in most cases.

Would you consider anything under 4 ohms as too low. 3? 2?

[mmoeller]

[QUOTE=Rudy Jakubin]
Same for the high end - 2.3 khz.
13560 / 5.5" = 2465 hz.
13560 / 6.5" = 2086 hz.
QUOTE]

Would you consider these distances even with counter sunk drivers?

[dlr]

Curious about the overlap suggestion

They'd have to overlap the woofers and underlap the mid for crossover point of 200 hz.

Even if the acoustic centers of the woofer and mid were 4" off (seems unlikely given the sizes, even with the vertical offset), 200Hz would have little need for overlap/underlap to work. 4" is still only about 21 degrees of rotation. In all liklihood, the rotation will be less than this for most drivers.

I suspect that their expected listening distance will tend to further minimize the vertical offset component.

[Dennis H]

Would you consider anything under 4 ohms as too low.

Yup although it depends on the amp. Most nominal 4-ohm speakers will dip to 3 ohms somewhere. A lot of receivers don't like a nominal 4-ohm load but most good-quality separates can handle it fine. Below that, it's a crap-shoot. Some amps will handle it and some will choke although no amp is really "happy" running a 2-ohm load. So, short version, keeping it at or above 4 ohms nominal is a good idea.

[mmoeller]

Yup although it depends on the amp. Most nominal 4-ohm speakers will dip to 3 ohms somewhere. A lot of receivers don't like a nominal 4-ohm load but most good-quality separates can handle it fine. Below that, it's a crap-shoot. Some amps will handle it and some will choke although no amp is really "happy" running a 2-ohm load. So, short version, keeping it at or above 4 ohms nominal is a good idea.

So 3 8ohms in parallel not so good. In general terms.

Would there be a way to correct for this type of setup? Let's say the following speaker used in a 3 woofer setup.

http://www.madisound.com/cgi-bin/index.cgi?cart_id=4550443.16380&pid=1958

[mmoeller]

One more question.

I have been simulating the above posted drivers in Winisd. Initial results show that each driver would need about .5 cu ft, but the low end extension is not good reaching -3db at 70hz. Increasing the volume to 1 cu. ft per driver gives more extension, and depending on the port can give a reasonably flat response to about 40Hz.

My question is this. How does increasing the box volume effect the tone of the speakers? What is the trade off for the increase?

To what frequency should I worry about cone excursion? It seems dangerous to try and extend these woofers too far. Past the port resonance the drivers seem to quickly go out of control.Would a high pass filter set at a frequency below port tune ever be used to help control cone excursion? WinISD tells me that past the port frequency the excursion basically becomes out of control.

I have attached a pic of the cone excursion w/ and without a 2nd order butterworth highpass at approx. 20hz. Reasonable?