So closed box speakers are pretty much a dead?

[617]

Good post, @617.
For reference, here is what may be possible at normal listening levels (< 85dB SPL) using a so-called Linkwitz Transform (afaict, he did not invent this) on a 10" Dayton RSS265-HF in a 30L (1 cu.ft) sealed enclosure. I moved the -3dB cutoff down to 20Hz and shifted the Q to 1/sqrt(2).

Spoiler: Source Code

import numpy as np
import scipy.signal as sig
import matplotlib.pyplot as plt


# Exploring a so-called Linkwitz Transform filter for subwoofer response tailoring
fs = 44100
fn = fs//2 # Nyquist
fc = 36.95 # corner freq of closed-box system in Hz
Q = 0.803 # Q of close-box system (aka Q_tc)

b = [1,0,0] # zeros
a = [1, (2*np.pi*fc/Q), (2*np.pi*fc)**2] # poles

w = np.linspace(5, 2*np.pi*fn, num=fn+1, endpoint=True)

w, Ho = sig.freqs(b, a, worN=w)
plt.figure()
plt.semilogx(w/(2*np.pi), 20*np.log10(np.abs(Ho)), label='Original System')
#plt.xlim([1, 150])

fc1 = 20 # desired corner frequency of close-box system in Hz
Q1 = 1/np.sqrt(2) # desired Q of closed-box system # np.euler_gamma #
lt_b = a
lt_a = [1, (2*np.pi*fc1/Q1), (2*np.pi*fc1)**2]
w1, H1 = sig.freqs(lt_b, lt_a, worN=w)
plt.semilogx(w1/(2*np.pi), 20*np.log10(np.abs(H1)), label='LT Filter')

bt = np.convolve(b, lt_b)
at = np.convolve(a, lt_a)
wt, Ht = sig.freqs(bt, at, worN=w)
plt.semilogx(wt/(2*np.pi), 20*np.log10(np.abs(Ht)), label='LT + Original')

plt.axhline(-3, color='grey', linestyle='--')
plt.axvline(20, color='grey', linestyle='--')
plt.xlabel('Freq [Hz]')

plt.legend(loc='lower right')
plt.ylim([-30, np.max(20*np.log10(np.abs(H1)))+2])
plt.xlim([5, 1200])
plt.grid(b=True, which='both', axis='both')

View attachment 38704

I can never get my LT to work right, so I normally just combine a peaking PEQ filter with a steep high pass at 20-30hz. How does the LT affect group delay? Surely just the same as a sealed enclosure with the same roll off?

 

[dc655321]

I can never get my LT to work right, so I normally just combine a peaking PEQ filter with a steep high pass at 20-30hz. How does the LT affect group delay? Surely just the same as a sealed enclosure with the same roll off?

I calc this as follows (leaving off from my previous code posting). Someone should check this.

Spoiler: Source Code

gd_orig = -np.diff(np.unwrap(np.angle(Ho))) / np.diff(w)
gd_lt = -np.diff(np.unwrap(np.angle(Ht))) / np.diff(w)

plt.figure()
plt.semilogx(w[1:], gd_orig, 'b-', label='Original System')
plt.semilogx(wt[1:], gd_lt, 'g-', label='LT + Original')
plt.legend()
plt.grid(b=True, which='both', axis='both')
plt.xlabel('Freq [Hz]')
plt.ylabel('Group Delay [s]')
plt.xlim([10, 1200])

EDIT: In general this transform modifies the system resonant frequency and its width (Q), so magnitude/phase/group delay changes are the norm: a Linkwitz transform nullifies the system's original poles, adding its own according to design.

 

[aarons915]

I used to think sealed mains were a better design for the lower group delay, easier integration with subs, etc but I've changed my opinion over the years. First of all, there doesn't seem to be any evidence that group delay is audible under 100Hz at the levels ported speakers produce, so I'm not sure that matters. It is true that ported mains are harder to integrate with subs since they don't rolloff like sealed speakers but I just found a processor (Emotiva UMC 200) that can do a 4th order high pass on the mains and problem solved. This basically allows you to take advantage of the low distortion bass from the port while not having any of the drawbacks of trying to play a lot of bass under the port tuning frequency since the 4th order slope rolls off the bass very sharply.

 

[jhaider]

Turns out they are not dead, just very expensive .

Not always. Aside from the NHTs (C3 auditioned, measured and recommended; C4 not auditioned but presumptively recommended based on similarities to C3), there's Monoprice (not auditioned)!

https://www.monoprice.com/product?c_id=304&cp_id=30404&cs_id=3040409&p_id=34188&seq=1&format=2

Bass is sealed but I must say since they use DSP correction which can't work properly to lower bass in a vented enclosure...

I think what you meant is boost below vent resonance is useless. That is true but DSP correction can be used to with resonant boxes. The trick is to tune the port or PR as low as you want the cutoff to be and use DSP to fill in the range between the woofer's rolloff and resonant output.

 

[Frank Dernie]

The trick is to tune the port or PR as low as you want the cutoff to be and use DSP to fill in the range between the woofer's rolloff and resonant output.

Interesting. I struggle to get my head around that, as a vibration engineer used to "mass dampers", which is effectively what a traditional reflex port is, but I'll take your word for it!

 

[Wombat]

Interesting. I struggle to get my head around that, as a vibration engineer used to "mass dampers", which is effectively what a traditional reflex port is, but I'll take your word for it!

The 'unloaded' region is below the useful loaded-mass port resonant response bandwidth which boosts/extends the natural driver frequency roll-off. Boost applied to this 'unloaded' region will cause excessive cone travel and likely damage the driver. Basically uncontrolled resonance.

In fact, normal signal is problematic for the unloaded region. For my BR speakers, tuned to 32Hz, I use a steep 20Hz Hi-pass filter for safety even though there is little content so low.

Berenak See page 387, Section 7.14 Diaphragm Displacement.

 

[Blumlein 88]

Interesting. I struggle to get my head around that, as a vibration engineer used to "mass dampers", which is effectively what a traditional reflex port is, but I'll take your word for it!

I don't think you should take his word for it.

It doesn't all add up. You'll overstress drivers doing what he described.

 

[Juhazi]

I think what you meant is boost below vent resonance is useless. That is true but DSP correction can be used to with resonant boxes. The trick is to tune the port or PR as low as you want the cutoff to be and use DSP to fill in the range between the woofer's rolloff and resonant output.

Large and Low Tuned (LLT) subwoofers or floorstanding speakers (B&W most models) use that. EQ with some way is not always necessary because of boundary reinforcement. Same applies to large sealed boxes/drivers.

I wouldn't use LLT with speakers for music, but it's ok for HT if you don't/can't use subwoofers.

Sonotube LLT subwoofers are popular diy projects. https://www.hometheatershack.com/fo...eral-discussion/29579-maelstrom-21-llt-7.html

 

[Wombat]

I don't think you should take his word for it.

It doesn't all add up. You'll overstress drivers doing what he described.

Only at the unloaded region below the port resonance band. There is still suspension stiffness and damping factor to ameliorate the condition. The designer needs to understand the BR behaviour in this region.

In real life we don't see a pyre of self destructed BRs even though the unwary owners subject them to material containing LF they weren't meant to handle. How many members operate smaller BRs, full range at loud levels without safety high pass filters to no destructive effect?

 

[Tks]

Maybe I am misunderstanding. But can a speaker with a passive radiator be considered sealed? If so, then there are a few I see coming out more and more these past few months.

I got Vanatoo T1 Encores in my living room. I love these things.

 

[jhaider]

I don't think you should take his word for it.

It doesn't all add up. You'll overstress drivers doing what he described.

Possible, but it depends on the driver. It's no different from EQ'ing a closed box driver below its rolloff.

Such assisted reflex alignments long predate DSP. Some of the earlier subwoofers - JBL B380 and B460 - had dedicated passive EQ boxes that do what I describe.

 

[617]

Maybe I am misunderstanding. But can a speaker with a passive radiator be considered sealed? If so, then there are a few I see coming out more and more these past few months.

I got Vanatoo T1 Encores in my living room. I love these things.

A passive radiator and a vent are almost identical in function. They both use a resonating mass to extend bass response. In a vented system, it's the mass of air inside the port tube pushing and pulling against the air in the box, in a passive radiator it is a speaker cone pushing and pulling against the suspension of the radiator. The Passive radiator is just a clever way of tuning a box very low without huge port tubes, and eliminating noise from the port.

 

[restorer-john]

How many members operate smaller BRs, full range at loud levels without safety high pass filters to no destructive effect?

Me.

No high pass filters unless I happen to notice the woofer moving slowly with large excursions (A/D converter near-DC issues on some CDs). The CD player, the preamp and the power amp are flat to <2Hz. Or on the rare times I play records on a TT.

There was a CD I played the other week which had some spurious very low frequency content (unrelated to the music) which appeared to be around 0.5Hz (2 second or so cyclic movement of about 8mm), At first I figured I had a component developing an issue and shut everything down to track down the source. Ruled out all the gear and it was on the actual CD. Wish I could remember which one it was now.

 

[Tks]

A passive radiator and a vent are almost identical in function. They both use a resonating mass to extend bass response. In a vented system, it's the mass of air inside the port tube pushing and pulling against the air in the box, in a passive radiator it is a speaker cone pushing and pulling against the suspension of the radiator. The Passive radiator is just a clever way of tuning a box very low without huge port tubes, and eliminating noise from the port.

Does it still qualify as something OP was asking for though? That's the part I am bit unclear about.

Also, just wondering in general. Are ported speakers simply a paradigm that's been around forever, and thats why they still rule when you take a look at the amount of offerings, or are passive radiators simply more complex to get right? I am a bit confused since all I ever see is people talk about superior bass performance from passive radiator speakers. Well if that's the benefit, what's the benefit of ported designs that dissuade manufacturers from moving on to this sealed design? Is simply because making huge ports is that big of a problem or something?

You say yourself is a clever tactic. Am I to take that as an implication that most makers aren't clever enough to make use of passive radiators?

 

[Juhazi]

Bass reflex became popular during '80s, because it is an efficient and cheap way to make cheap and small/midsize speakers have louder bass at 50-100Hz- and that's what people want. Then also vinyl players started to disapper from the majority of homes, and rumble was not making woofers' voice coils melt or suspensions break so easily any more.

To make a high end hifi speaker's bass reflex working well - tuned around 30Hz and with minimal port noise is difficult! Loud listening levels and sophisticated/educated listening habits easily detect chuffing from port and sitll many hardcore hifists have vinyl players with "high-end" preamps that often don't have rumble filters! Asking for trouble!

Using subwoofers helps the main speaker to minimize the side effects of port tuning, except the midrange leakge which seems to be a real problem - in measurements. Most people don't even know that it happens and thus don't hear it!

Passive radiator reflex doesn't chuff, but it may clonk with high output and also leaks midrange in a 2-way speaker! It is good practise to have doubled cone area in passive units, to prevent overexcursion and mechanical noise. There aren't many commercial subwoofers doing that, more in diy scene.

https://www.css-audio.com/single-post/2018/01/02/When-Passive-Radiators-Are-Better-Than-Ports
https://www.css-audio.com/online-store/SDX12-and-Dual-APR12-Passive-Radiator-Kit-p112666755

 

[jhaider]

I am a bit confused since all I ever see is people talk about superior bass performance from passive radiator speakers.

They are usually imagining things. Seeing an extra cone can cause that.

Is simply because making huge ports is that big of a problem or something?

Possibly, because pipe resonances.

Am I to take that as an implication that most makers aren't clever enough to make use of passive radiators?

The biggest drawback of PRs is that a magnetless woofer costs a lot more than a tube.

I discuss PRs a little bit in my Amphion Argon 3S review. Find at link in signature.

 

[JohnBooty]

Is simply because making huge ports is that big of a problem or something?

There can of course be chuffing issues.

But the even bigger problem is how much dang space a big port takes up inside the cabinet. You have to make the enclosure bigger just to fit the giant port.

That's one reason why pretty much 100% of portable Bluetooth speakers use passive radiators. They are trying to squeeze some semblance of bass out of a tiny enclosure. There's no room for a big port inside. Of course, water/weather resistance is the other benefit to passive radiators for those speakers. Can't very well be waterproof if you've gone with a bass reflex design an a big ol' open port.

 

[JohnBooty]

Consequently, most drivers today are really optimized for vented enclosures, with very few (Scanspeak 8535, Dayton RS270) really doing well in a sealed box.

Not sure if KEF does it with other models or the newer Q150/Q350/etc, but my Q100s shipped with a set of foam plugs so the speakers could be used in sealed mode as well. I would assume the drivers were at least decently suited to functioning in a sealed enclosure? I can't really speak to their sealed performance though. I preferred to run them ported, with a steep crossover slope to the subwoofers around 60hz.

I'm curious, what makes a driver well-suited (or not well-suited) for a sealed enclosure?

 

[617]

Not sure if KEF does it with other models or the newer Q150/Q350/etc, but my Q100s shipped with a set of foam plugs so the speakers could be used in sealed mode as well. I would assume the drivers were at least decently suited to functioning in a sealed enclosure? I can't really speak to their sealed performance though. I preferred to run them ported, with a steep crossover slope to the subwoofers around 60hz.

I'm curious, what makes a driver well-suited (or not well-suited) for a sealed enclosure?

The efficiency bandwidth product, or EBP of a woofer, is the ratio of the driver resonance frequency and it's electrical Q.

http://redspade-audio.blogspot.com/2011/03/efficiency-bandwidth-product-ebp.html Good article outlining the issue.

EBP < 50 - use only for a sealed box
EBP 50 - 100 - can be used in either
EBP > 100 - vented box only

Most drivers are in the middle range - drivers which are around 50 or below are pretty unusual. In the middle range, you get the tradeoffs seen in my graphs earlier - vented box is big but not ludicrous, sealed box is less efficient at producing bass but is half as big. For your average midbass, plugging the port will result in an essentially sealed, slightly oversized enclosure, which works fine.

The SS 8535 has a Qes of .46 and a Fs of 25hz, resulting in an EBP of 54, close to the low end of normal.
For comparison, the SS 8545, essentially the same driver, has a Qes of .22 and a Fs of 25hz, resulting in an EBP of 113, strongly suggesting a vented box.

The 8535 (sealed friendly) is a bit less efficient. That tends to be the tradeoff, often called Hoffman's iron law, as of yet unviolated, that between high efficiency, small enclosure size and low bass extension you can only have 2 of 3.

The way this shakes out is that the 8535 requires a sealed cabinet of 1 cubic foot, which is quite a bit for a driver of this size, but it goes all the way down to 47hz! The 8545 on the other hand will give you a flat sealed alignment in a tiny box (.2 cubic feet) but only goes down to about 100hz.

 

[Juhazi]

The way this shakes out is that the 8535 requires a sealed cabinet of 1 cubic foot, which is quite a bit for a driver of this size, but it goes all the way down to 47hz! The 8545 on the other hand will give you a flat sealed alignment in a tiny box (.2 cubic feet) but only goes down to about 100hz.

I guess you intended to say, that they go flat (-3dB?) to 47/100Hz. Below that typical roll-off happens. And this is just a 6,5" midwoofer, typical for bookshelf two-way speakers.