"Orchid XP v3" <voi### [at] devnull> wrote in message 
news:46dee812$1@news.povray.org...
>>>> Interesting, how do they do that?
>>>
>>> You think *I* know?
>>
>> It used to be that more efficient speakers in general had more unwanted 
>> resonances and lumps in their frequency responses: IOW poorer fidelity. 
>> I don't particularly know why but that's how it was.  Obviously the 
>> designers would have overcome that if they could have, but some chose low 
>> efficiency in return for better sound quality.  I don't know how well 
>> they are doing now with that trade-off, it's been some years since I last 
>> read any reviews.
>
> That sounds about right.

Actually now I think of it, I remember a very common example of two 
different speaker designs that speaker manufacturers often choose tradeoffs 
with:  Infinite Baffle, vs. Tuned Port.

Theoretically, an Infinite Baffle speaker is simply a driver mounted in a 
wall (or baffle) that extends to infinity in all directions of its plane. 
No sound energy that radiates backwards from the driver (and is inverted) 
can get around the baffle where it would interfere with the 
forward-travelling sound wave.  In practice however, an IB speaker is a 
driver in a sealed box with a limited internal volume.

Good points of IB are smooth frequency curve, linear or slow-changing phase 
over frequency, good damping, low colouration -- in general high-quality 
sound with good 3-D imaging on the virtual sound stage.  Bad points are bass 
response starts to tail off slowly at a relatively high frequency due to the 
limited cabinet volume and can really lack guts on low notes like the bottom 
E1 of a bass guitar (~41 Hz).

Tuned Port speakers, on the other hand, have a vent in the cabinet (often in 
the front).  Its entire purpose is to extend the bass response of the 
speaker below the frequency limit usually imposed by the cabinet's internal 
air volume.  The port connects to the internal cavity by a tube of a length 
calculated to tune to a chosen bass frequency half-waavelength.  At that 
frequency, the wave at the port entrance (ultimately coming from the back of 
the driver) is in reinforcement phase with the wave from the front of the 
driver, extending the speaker's freq. range.

Good points of TP are deeper extension of bass response for a cabinet size 
and relatively flat(tish) bass response below 100Hz down to the tuned freq.

Bad points are the low-frequency response can actually have a peak at the 
tuned frequency, then it will plummet off a cliff just below it, the phase 
of the bass tends to change dramatically with frequency, the speaker has a 
lumpy impedance graph - presenting a very low impedance* to the amp at the 
tuned frequency and high-impedance at nearby frequencies (more difficult 
load for an amplifier to drive).  If not carefully designed, TP can 
over-emphasise sound near the tuned freq., causing the effect known as 
"one-note bass" or "boom box".  The port's resonance has inertia and stores 
energy which the driver takes time to overcome, it can continue transmitting 
for several cycles after the amplifier has stopped the signal, which means a 
TP speaker's extended bass response comes at the cost of "mushing up" the 
sound (time-smearing).

Just an example, y'know.  :-)

* [impedance could be the opposite way to my memory, but point is, it's 
still lumpy]

>>>> Hard to see how you could deliberately make the system less efficient 
>>>> without just plonking a huge 50W resistor in series (which would 
>>>> totally screw up the quality of the sound).
>>>
>>> And why would that be?
>>
>> The answers lie in the way passive networks of electronic components work 
>> and in speaker mechanics.
>> [...]
>> So if you plonk a great big resistor in series with all that, the amp 
>> can't damp unwanted speaker motion, the crossover detunes, the frequency 
>> response goes lumpy, the sound goes muddy, and you think "bleccch!"  :-)
>
> I'm sure that would be the case.
>
> Now, what if that resistance was part of the design right from the 
> beginning?

Possibly...  But I think it would prevent the amplifier from sinking the 
back-EMF current from the speaker.  The amplifier's output stage should have 
an apparent impedance (Z) as close to zero as possible, but instead its Z 
would be the resistor's value, allowing a back-voltage and thus impacting 
sound quality.  I don't think there'd be a way to design around that problem 
in the speaker alone.

> By the way... The documentation for my mum's amplifier claims that it 
> contains special circuitry to cancel out the electric signals generated by 
> the speakers themselves. (I have no idea whether this is special or 
> whether it's something all modern amps do and they just wrote about it to 
> make themselves sound cleaver...)

"We have carefully designed into our mower, a special feature to cut grass!"

Every power amp does it, with negative feedback taken from the output stage. 
Any voltage that is not equal to the amp's instantaneous output drive is fed 
back, differenced, inverted, amplified and output again, cancelling the 
error.  That's also how amplifiers get to look like zero impedance or short 
circuit to error signals appearing at the output transistor stage.

>>>>> Hmm, I think my amplifier (nothing special) is rated at 60 W per 
>>>>> channel. (IIRC, into 8 ohms at 1 kHz.)
>>
>> For listening at home, the real reason for high-powered amps is not to 
>> make louder noise, but for fidelity.
>
> Hmm, isn't that what *I* said several posts ago? ;-)

Sorry, I probably either missed or forgot that you did.  :-)  Anyhow it was 
more that I thought the others were pulling the discussion of amp power 
specs off-course.

>>> (It's like all toasters have a special setting that transforms bread 
>>> into charcole. We don't know why, but they all have it.)
>>
>> <Ahem>  Just because you didn't figure it out, doesn't automatically make 
>> the manufacturers into idiots.  The high setting on toasters is needed to 
>> get even a _little_ colour into *crumpets*.  But, as you apparently 
>> haven't toasted a crumpet before, they obviously don't exist and every 
>> manufacturer out there is stupid for gratuitously making toasters with a 
>> nuclear setting.
>>  :-P
>
> Wait... Surely to toast a crumpet you want a *lower* heat? (Since it's 
> thicker and hence nearer to the heat source - and also you want to cook it 
> slower so it cooks through completely.)

I see what you mean, but a lower heat would dry it out before beginning to 
brown it, and then you'd have a circular paver for your driveway.  :-) 
There's a fair bit of moisture in crumpets, and despite the honeycombing, 
they have a dense and rubbery texture.  It needs to be high heat to brown 
the outside while leaving the inside soft and mmmmm.  :-)  You don't want 
them "cooked through".

>>> (I once tried connecting a line-level output to the phono input. Big 
>>> mistake...)
>>
>> Ow.  :-(
>
> Sounded a bit like a metal guitar. Also made the sound come out on *all* 
> input selections. o_O

That's because it's the same effect as metal guitar:  massive overdrive / 
extreme clipping.

Post a reply to this message