Low distortion drivers for sub-bass in shallow sealed cabinets

[ThatSoundsGood]

The Purifi 10" is about 150mm deep so it's also not gonna work. Have you thought about a redesign on the house in order to accommodate the subwoofer? Maybe if you package it with a new back porch.......

 

[sigbergaudio]

Perhaps you need to reconsider your requirements?

What does "below 10% @ 20Hz " mean exactly? At what SPL and what distance?

 

[mcdn]

Perhaps you need to reconsider your requirements?

I can be a bit stubborn! But truly, I have considered a lot of options before settling on the shallow, wall-mounted, stereo sub solution.

What does "below 10% @ 20Hz " mean exactly? At what SPL and what distance?

Good question. Since you make very nice subs I'm sure you have looked into this a lot, so please correct any errors!

For comparing sub drivers, I think _what I want_ is to compare THD with the driver on an infinite baffle at, say 85dBSPL@1m @ 20Hz. Actually finding out that number is quite hard for most drivers though. Others may care about different numbers, but that's the one that matters to me because I am after clean sub-bass and am prepared to throw multiple drivers, DSP and watts at the problem. I also firmly believe that sub-bass distortion is very audible when the sub-bass is at reference levels, because the harmonics of 120dB of 20Hz bass are well within the window of audibility due to fletcher-munson curves.

Consider the redcatt driver. The THD at 20Hz in the graph is ~5% when summing the harmonics shown. It's taken at 90dB (presumed to be SPL/1m), but presumably that's the peak SPL across the measurement sweep, not the SPL at 20Hz. hobbyhifi.de don't publish detailed test protocols, or if they do they are only available in the print magazine. With a lot of assumptions we can say we have evidence that the driver has 5% distortion at 20Hz at a voltage that gives 90dBSPL/1m at 100Hz. Checking the datasheet that gives us 5% THD @ 20Hz @ 68dBSPL/1m. Good at first sight, but the volume is very low, we would need a lot of drivers.

For the Dayton LS10-44 we have measurements from both hobbyhifi.de and Vance Dickinson at AudioXpress, so let's compare. Vance tells us he measures near field with the mic 10cm from the dust cap, and at an SPL equivalent to 94dBSPL/1m - again that's not held constant in SPL, but is the voltage level at calibration around 100Hz (see the article here). He measures 22% THD @ 20Hz, and the datasheet shows a 12dB drop between 100Hz and 20Hz, so we have 22% THD @ 20Hz @ 82dBSPL/1m. Hobbyhifi.de gets about 10% at '90dB', which seems like it could be the same driver driven 4dB lower. So it could also be 10% THD @ 20Hz @ 78dbSPL/1m.

To take an extreme alternative example costing over $800, Purifi have measurements for the PTT10 in their datasheet, which are stated as being at 94dBSPL/1m, again clearly constant voltage ref the calibration level at 100Hz. We get 4% THD at 72dBSPL/1m at 20Hz, which is good, but it would be much better to know what the measurement is at 94dBSPL/1m @ 20Hz. Luckily hificompass.com has separate measurements at 2.83V and 16V, showing an increase from 1% (-40dB) to 8% (-22dB). This is better than the datasheet because the peak SPL was only 89dB in the 2.83V measurement vs 94dB on the datasheet. So the PTT10 does 8% THD @20Hz @ 80dBSPL/1m

The SB driver (SB26DBAC76-4), as measured by Vance Dickason again, has 4% THD @ 30Hz @ 86dBSPL/1m, but climbing rapidly as frequency drops - we can probably extrapolate 10% THD @20Hz @ 83dbSPL/1m

Normalising all of those using multiple drivers to get 120dB @ 20Hz can wait until tomorrow.

 

[sigbergaudio]

@mcdn Are you seriously suggesting you will be playing 20hz at 120dB? How many drivers are you expecting to use?

As you expect to hear 10% THD at 20hz, I am curious to hear if you have tested to play back a 20hz sine wave through your system? 20hz is barely audible even at 96dB. There's very little chance you can hear 10% distortion at 20hz when playing actual music or movie content. You can barely hear the fundamental.

 

[sigbergaudio]

Assuming 10% THD @20Hz @ 83dbSPL/1m for the SB26DBAC76-4 is correct, which doesn't seem entirely unreasonable, you will need roughly 64 drivers to reach 120dB@20hz/1m.

I suspect this is not the most space saving route to 120dB@20hz.

 

[ppataki]

I would also add that - at least according to my experience - the room will actually play a bigger role in terms of distortion than your speaker drivers....
I have seen rooms where a null-point induced distortion reached 5-10% in the range of 40-80Hz

I fully agree with @sigbergaudio that you will not likely to hear 10% distortion at 20Hz

I would grab any of the drivers you took into consideration and then spend time finding the best spot in the room where frequency response (and hence distortion) is the best

 

[mcdn]

@mcdn Are you seriously suggesting you will be playing 20hz at 120dB? How many drivers are you expecting to use?

As you expect to hear 10% THD at 20hz, I am curious to hear if you have tested to play back a 20hz sine wave through your system? 20hz is barely audible even at 96dB. There's very little chance you can hear 10% distortion at 20hz when playing actual music or movie content. You can barely hear the fundamental.

What's strange about that? THX reference level is 115dB, so roughly the same. We can call it 110dB if you want, the problem remains the same.

And yes, I have played 20Hz tones in a system in a different room (the LXsub4 woofers in the family room can't play them). What you hear is indeed distortion, hence why the fundamental needs to be up at 110-120dB.

 

[sigbergaudio]

What's strange about that? THX reference level is 115dB, so roughly the same. We can call it 110dB if you want, the problem remains the same.

And yes, I have played 20Hz tones in a system in a different room (the LXsub4 woofers in the family room can't play them). What you hear is indeed distortion, hence why the fundamental needs to be up at 110-120dB.

Well, what's strange about it is of course that you need to fill your room with drivers, and that defeats the point of shallow mounting probably.

The calculation of 64 drivers is free air assuming you didn't count on any room gain, so in reality you will probably "only" need around 20 drivers or so if we assume 10dB room gain at 20hz.

 

[mcdn]

Assuming 10% THD @20Hz @ 83dbSPL/1m for the SB26DBAC76-4 is correct, which doesn't seem entirely unreasonable, you will need roughly 64 drivers to reach 120dB@20hz/1m.

I suspect this is not the most space saving route to 120dB@20hz.

Don't we get 6dB per driver increase in SPL for correlated signals? so 120-83=37, 37/6 = ~6? And that's before room gain.

 

[mcdn]

Well, what's strange about it is of course that you need to fill your room with drivers, and that defeats the point of shallow mounting probably.

The calculation of 64 drivers is free air assuming you didn't count on any room gain, so in reality you will probably "only" need around 20 drivers or so if we assume 10dB room gain at 20hz.

See the VituixCAD sims above as well. I love the idea of a room entirely covered by bass drivers but suspect that will have to wait until the children have left home.

 

[sigbergaudio]

Don't we get 6dB per driver increase in SPL for correlated signals? so 120-83=37, 37/6 = ~6? And that's before room gain.

'
You get 6dB per doubling of drivers. 2 drivers = 6dB, 4 drivers = 12db, 8 drivers = 18dB..

 

[sigbergaudio]

And this is assuming you also double the power, otherwise you only get 3dB.

 

[mcdn]

'
You get 6dB per doubling of drivers. 2 drivers = 6dB, 4 drivers = 12db, 8 drivers = 18dB..

Doh! Yes of course. So my VituixCAD sim has 3 x SB drivers giving 100dBSPL/1m with ~500W of power (see screenshot on page 1). That's for one enclosure, so two enclosures is 106dB. Add 6-12dB of room gain and we are there, no?

 

[sigbergaudio]

Doh! Yes of course. So my VituixCAD sim has 3 x SB drivers giving 100dBSPL/1m with ~500W of power (see screenshot on page 1). That's for one enclosure, so two enclosures is 106dB. Add 6-12dB of room gain and we are there, no?

I guess three of those drivers might theoretically reach 100dB at 20hz at @1m, I suspect you won't be at 10% THD at 10mm excursion though. But ignoring your max THD requirements for a moment, a second enclosure might realistically perhaps give you 105dB. 12dB room gain (best case) will give you 117dB.

And this is of course at 1m, I'm not sure what SPL you are looking for at the listening position, and how far away that is. You theoretically loose around 6dB per doubling of distance, so if you're 3 meters away you're 9dB down. But bass works in mysterious ways in smaller rooms, so you might not loose that much in practice. So maybe 110dB at the listening position if it's not too far away, but likely not at 10% THD.

 

[sigbergaudio]

I would perhaps suggest reducing your requirements both in max SPL and also low frequency. Your shallow mount drivers (whichever you choose) will be much more comfortable and overall play at much lower distortion if they're not forced by DSP to play the lowest frequencies. Allowing them to roll off below 30hz would make for a much cleaner system, and will reduce the distortion significantly. 30hz is still very low (there is hardly any content below 30hz in most music), and 30-70hz is where most of the energy and punch is in movie effects as well.

 

[mcdn]

I guess three of those drivers might theoretically reach 100dB at 20hz at @1m, I suspect you won't be at 10% THD at 10mm excursion though. But ignoring your max THD requirements for a moment, a second enclosure might realistically perhaps give you 105dB. 12dB room gain (best case) will give you 117dB.

And this is of course at 1m, I'm not sure what SPL you are looking for at the listening position, and how far away that is. You theoretically loose around 6dB per doubling of distance, so if you're 3 meters away you're 9dB down. But bass works in mysterious ways in smaller rooms, so you might not loose that much in practice. So maybe 110dB at the listening position if it's not too far away, but likely not at 10% THD.

Everything you say is true, but we are talking about peak SPL here. If we have 115dB of peak SPL capability with higher distortion, 100dB average with lower distortion is probably achievable, and that would work well with programme average SPL around 80dB.

Apart from ruling out the LS10-44 for its por distortion @ 20Hz, and the Purifi for being too deep, I'm still none the wiser on the best choice of driver though.

 

[mcdn]

I would perhaps suggest reducing your requirements both in max SPL and also low frequency. Your shallow mount drivers (whichever you choose) will be much more comfortable and overall play at much lower distortion if they're not forced by DSP to play the lowest frequencies. Allowing them to roll off below 30hz would make for a much cleaner system, and will reduce the distortion significantly. 30hz is still very low (there is hardly any content below 30hz in most music), and 30-70hz is where most of the energy and punch is in movie effects as well.

True, but I already have a system that is rolled off with an LR4 @ 30Hz, and it just isn't satisfying [edit - satisfying enough, it's fantastic of course, but it could do more] for movies or electronica.

 

[sigbergaudio]

Everything you say is true, but we are talking about peak SPL here. If we have 115dB of peak SPL capability with higher distortion, 100dB average with lower distortion is probably achievable, and that would work well with programme average SPL around 80dB.

Apart from ruling out the LS10-44 for its por distortion @ 20Hz, and the Purifi for being too deep, I'm still none the wiser on the best choice of driver though.

No shallow mount driver is perfect, typically struggling with assymetric BL etc, but the SB acoustics one is a decent driver as long as you use enough of them. It's popular in shallow bass towers here in Norway.

Here is a video featuring yours truly assembling and tuning a bass tower with those drivers (from around 1:05).

 

[sigbergaudio]

So that system is using 12 of those SB Acoustics drivers, the room is not more than maybe 12-15m^2

 

[mcdn]

Love it! I just checked with a typical track (Tinlicker, because you move me), and my system at "quite loud" is about 100dB(C) or 80dB(A) 1m from the speakers. Clearly massively bass dominated for this track so I think the C weighting is more useful. About 95dB(C) or 75dB(A) at the listening position in a 40m^2 room. The drop off is small with listening distance because of the bass dominance on that track.