so how low?

[thinking]

I've been browsing through all the finer points here, and I thought I had seen it once in the past, but I just can't seem to place it now. How low will it go? I think I found one reference to an IB response being essentially flat down to the driver's F3, and then rolling off at 6dB per octave. Is this correct?
I know there are always a host of issues involved, extending to even the room itself, but what I'm getting at is once you have selected a specific driver, what do you expect the bottom end to look like.

[ThomasW]

An IB just a very big sealed box, like all sealed boxes they ultimately roll off at a rate of -12dB/octave below the F3, and that's in anechoic space.

I posted this in the EQ forum below, but I'll repost it here. These are actual measurements of my small IB (4-12" drivers with a Fs of 16Hz). The blue plot is nearfield. The magenta plot is ~4m away. This is an example of the effects of room gain.

Note there is certainly output lower than 16Hz. I didn't measure it because I was posting this plot for people using the BFD for EQ and it's lowest filter is 20Hz.

[thinking]

Thanks, that's exactly the sort of info I was looking for. So, if your driver has a F3 of, say 20, that means that 1/2 half octave down it will, by the numbers, drop 6 dB at 10 Hz, which really isn't performance to be ashamed of.

[ThomasW]

So, if your driver has a F3 of, say 20, that means that 1/2 half octave down it will, by the numbers, drop 6 dB at 10 Hz,

From 20Hz to 10Hz is a full octave, so it will be down -12dB @ 10Hz.

[thinking]

Oops, I knew that, should have said down 6 dB at 15 Hz

[tweeker]

Technical nitpick, would that not be 6dB down at 14.1hz? *.707 for 1/2 octave? More pointless pedandtry, I think it would also be somewhat contingent on driver q.

[dewardh]

Thomas:

How much actual amplifier output power does it take to produce the spl shown in your response plot (nominal 85 db at listening position)?

I ask because I'm in the midst of installing 2x2 12 in. drivers (forward firing line) with 4 60 watt (nominal at 8 ohm, 90 watt at 4 ohm) amps and am a bit worried about possible overexcursion, especially below resonance.

dewardh

[ThomasW]

How much actual amplifier output power does it take to produce the spl shown in your response plot (nominal 85 db at listening position)?

Beats me. The only accurate way to determine that is by taking voltage measurements at the back of the power amp while the test tones are being played.

You can 'sim' this using a box modeling program, those are of course anechioc.

Unless you have a VERY small room, 2-12"s are going to quickly run out of gas...

[dewardh]

Thomas:

Yes, that was the question . . . what is the *actual* power required to produce 85db nominal at your listening position from 20-40 Hz. The various sims produce wildly divergent results, especially when "room gain" is factored in, which is why a reality check appeals to me. At least one sim suggests that 16 watts will take my drivers (unloaded) to Xmax below resonance.

4 (2x2) 12 in. drivers have about 20% *more* total SD than 2 15 in. drivers, and thus correspondingly more volume displacement for the same excursion. I'm not looking to bring down the plaster here <g>. In addition the more extended (rather than point) source(s) should (in theory) reduce the effect of room modes . . .

In any case I already had drivers (NHT 1259, from a previous sealed box project) and face significant constraints regarding IB mountng location (radiant heat in the floor, existing built-in bookcases etc.) and have a desire for seamless integration with the also-under-construction ORION mains. The system will be used as much as or more for music than for HT, so the design considerations are somewhat more stringent than were the subs for HT LFE alone.

All of which is pretty far from the question "how much power" <g>, but may help explain some of my other-than-ordinary design choices . . .

dewardh

[ThomasW]

This is all a moot point given your drivers. Those are specifically designed to run in a smaller sealed boxes where the driver's suspension can utilize the 'air spring'. They're not intended for use in any other alignment. This is info per Ken Kantor.

[dewardh]

Thomas:

"Intended for" and "suitable for" are, of course, two entirely different things. If you compare, for example, the TS parameters of the NHT 1259 and the Adire DPL 12 you will find them surprisingly similar:

. Fs . Qts . Vas . Sd . Xmax . SPL
16.3, .400, 149, 481, 14.3, 84 (1,1)
16.5, .445, 238, 491, 13.0, 89 (2.8,1)

Otherwise they differ modestly in Qes and significantly in Qms, where the NHT has *much* better damping (lower Q), which should be better in an IB. What NHT parameters would make it less suitable for IB than the DPL (or the similar Tempest)? The "Tempest White Paper" at the Adire site describes both design intent and resulting parameters almost identical to those for the NHT driver (see also the Tempest parameters in the "Driver Comparison Chart" on the FAQ page of this site) . . . is the Tempest series also unsuited for IB use?

dewardh

[ThomasW]

Sorry I'm not one to get into semantic debates.

When the man that designed the driver says it shouldn't be used in any alignment other than a small sealed box. I pay attention.

The Qms, Vas, Cms, are the tell-tale parameters.

Good luck with your project

[dewardh]

Thomas:

Thanks for the kind wishes. I hope, of course, to improve my "luck" with careful consideration and design <g>. Since all the published parameters suggest that the NHT 1259 will work just fine in an IB (as well as or better than many of the commonly used drivers from Adire like the DPL-12, or the IB15, for examples), and no conflicting evidence has been forthcoming here, I'll proceed confident on that front.

If you happen to measure the actual power input to your DPL-12 array in the next few weeks or month I'd be interested in hearing the numbers . . . in any case I'll report the results for mine in a month or two when, hopefully, with luck, and time permitting, everything will be working.

dewardh

[ThomasW]

[e-mail]

[dewardh]

Thomas:

Thank you for the referral. On your advice I e-mailed Ken Kantor, and he graciously took time from an obviously *very* busy schedule to reply. I outlined our discussion here, and my planned application. The core of my question was:

“I recognize the danger of overexcursion to any driver in an IB, and that the
1259 may be more vulnerable because of its unusually high compliance, and
will protect against that with both high pass filtering and peak limiting.
Would you be willing to comment on any *other* potential problem(s) with the
1259 in an IB? [One point made was that] the 1259 was not *designed for* IB use . . . but that, of course, applies to *all* the LF drivers that folks are today using in IB arrays, and I cannot find any particular fault in the 1259 specifications that makes it uniquely unsuited for such use.”

His reply, in full:

“You are correct. If you properly take into account the possibility of over-excursion, you are fine with an IB. Yes, the 1259 was "designed" for a sealed box. As such, a decent sealed woofer can be made without requiring limiting or eq. I wouldn't use it in an IB as a purely passive system. But, there is no magic here. The 1259 motor is pretty linear up to its limits, and any alignment that respects these limits will perform as predicted.

Let me know how your design works out!

-k”

I hope (with luck and some spare time <g>) to be able to present some results both here and to Mr. Kantor in the (relatively <g>) near future . . .

Deward

[ThomasW]

I wouldn't use it in an IB as a purely passive system. But, there is no magic here. The 1259 motor is pretty linear up to its limits, and any alignment that respects these limits will perform as predicted.

Read this part carefully since I think it's a bit of a disclaimer.....He seems to be saying use a subsonic filter (maybe something even higher) and have enough drivers to safely cover the SPL you want without stressing the driver.

Use Unibox to model up your drivers and see at what power levels and SPL/frequency you hit the excursion limits, and then adjust your expectations accordingly.

My suggestion is this, make a manifold with the drivers mounted in removable panels (like Nousaine's design). If you run out of excursion before you get the SPL you want, you can drop in new inserts with different woofers. (I'd make the removable panels large enough to accommodate 15" drivers)