On the mysterious passive radiator...

[boomzilla]

When using a passive-radiator subwoofer, in what direction should the passive radiator (henceforth referred to as the PR) face?

As I understand it, the PR and the driver work in tandem to produce a smooth output curve. At some frequencies, the driver and the PR work in tandem; at some frequencies, the PR moves while the driver remains stationary; at some frequencies, the PR is stationary while the driver moves. So depending on the frequency, one or the other (or both) may be moving.

Now since the PR is in close proximity to the driver (less than a wavelength for all credible frequencies), the two should blend totally for a smooth combined response at the listening position.

It goes without saying that the PR can't be obstructed by close proximity to walls or furniture, and that if the PR is on the bottom of the sub, the manufacturer-supplied feet give adequate clearance for the PR to function per design.

My specific question pertains to a design with a bottom-mounted driver and a side-wall mounted PR. Options for placement include:

1. PR facing the listening position
2. PR facing a side wall
3. PR facing a back wall with some clearance

I can pull out my SPL meter and determine the smoothest response empirically, but would enjoy knowing the theoretical best position (and why).

Another question pertaining to the sub in general:

The curve is a smooth one with about 3dB roll-off at 20 and 100 Hz. Since I'd like to cross over closer to the 100 Hz. mark, I have two options for flattening the sub's response:

a) use a relatively narrow-band DSP boost to bring the 100 Hz. response up to the peak of the curve or
b) use a broad-band DSP cut to bring the remainder of the sub's response down to the 100 Hz. level

Which is superior, and why?

And finally, is a PR sub effectively identical to a ported sub in terms of equalizing for low frequency response? The "rule of thumb" that I've always heard is that a ported sub should not be equalized lower than ½ octave below its natural roll-off point to avoid losing port loading and potentially damaging the driver. I have no documentation of this, and don't know if it's true or not. Using that rule, if a ported sub had an anechoic -3dB point of 20 Hz, it would be "safe" to boost the 15 Hz. level by 3dB (but no more) using a very narrow-band boost. True, false, old-wives tale, or not so simple?

Thanks - Boomzilla

[DYohn]

You ask many questions in one post.

First of all, a PR works using different principles than a vented system, although they both are intended to produce the same result - utilizing the back wave of the woofer to increase system efficiency and extend response at the lowest frequencies. Vented or ported systems use the Helmholtz effect (the energy moving inside the enclosure to activate the resonance of the port produces sound.) PR systems utilize the resonance of the enclosure to activate the radiator, producing sound through sympathetic vibration. Both the port and the PR produce sound over a relatively narrow band, and the designer will create the system so that band adds to the woofer as it naturally rolls off to extend frequency response downward. Efficiency of the system increases over that of the raw woofer since the system uses both the front and the rear wave coming off the woofer.

In general the "direction" of the PR does not matter as the frequency being produced is non-directional, but the sound will add an additional wave to the room so careful placement in relation to walls is important as it will make room equalization more difficult. Also, all subwoofers create what is known as "group delay" (google if you don't know what this is.) Sealed systems create the least, then vented and bandpass, and the worst group delay is found in PR systems. This can create ringing in the room and what is known as bass overhang, and is often described by users as a "slower" sound although fast and slow in regards to subwoofers is not really applicable. It's the group delay they are hearing and describing.

What do you mean by -3db roll off? Do you mean the -3db FR point? No system rolls off at that rate so I'm not sure what you're looking at. And if it rolls off at the top of the frequency band it may be a bandpass system?

As far as using equalization to correct sub output response in room, cutting is always preferable to boosting if you'd like to be sure you avoid amplifier clipping.

And again, I am not sure I understand you final question. Are you talking about using a subsonic filter on a woofer below its natural roll-off? I never do that unless there is some sort of system interaction happening in the room. Let the woofer woof. And again, boosting by +3db at 15Hz is generally a quick way to overdrive an amp since that +3db will cause the amplifier to output twice the power it normally would at that frequency. But if you are certain you have the 2X headroom, sure go for it if you need it.

[boomzilla]

So if I understand you correctly, any subwoofer with a passive radiator should be used with as low a crossover point as feasible to prevent the group delay from interfering with the audible range?

Put another way, the best use of a subwoofer with a passive radiator would be to augment the truly low bass (40 Hz. on down, for example) outside of the satellite speakers' range?

Finally, is it fair to say that subs with PRs are best placed away from ANY room boundaries?

[DYohn]

Not at all. You cannot avoid the group delay of a subwoofer system, it occurs mainly at the PR tuning frequency, which is generally at the -3db point (or so) of the system. The low-pass Xover has very little to do with it. It should be set depending on what is needed for system integration. For a THX-style home theater, 80Hz. For a Dolby style home theater, 120Hz. For a 2-channel system, whatever works best with the mains.

The system should be placed where it will best enhance the in-room goals. Put it in a corner if maximum room gain is desired, for example, or as far away from any reflective surface as possible if maximum flatness is desired. It all depends on the intended use, you cannot make blanket statements.

[boomzilla]

When calculating Q for a filter should the bandwidth be measured from the -3dB points from the peak, or the -3dB down points from the baseline you want to achieve?

[DYohn]

The Q of a filter defines the slope of the envelope around the center frequency, with a higher Q factor meaning the filter is active across a narrower bandwidth. So again, you use what you need. In general when trying to tame a room mode with a PEQ, you want to measure the room and use a Q that closely matches the slope of the peak (generally, I use a Q of 1 for most room EQ, but sometimes a broader effect (lower Q) sounds better, and for stage use to tame feedback a high-Q filter is often best.) If you are trying to boost the low end to compensate for subwoofer roll-off, you'd likely be better served by moving the sub closer into a corner of the room and picking up some room gain than using EQ. NEVER use manufacturers published specs to try and decide what to do unless you're listening in an anechoic chamber. Always set up the system the way you want to use it and measure the room. And remember, boosting the low end is the fastest road to amplifier clipping, as I mentioned before.

[Perpendicular]

Oh, great! Now I have to dump my Goldenear SuperSub and purchase a sealed unit.