High Qts and Bass reflex...newbie Question....
mbon...@hotmail.com
suited for rec.audio.tech, I think that this may be a good
starting point for a discussion of technical figures and
their relationship to quality. -- dsr ]
I have been an analog circuit designer for decades and I have recently
rediscovered my interest in speaker design. I am playing around with
a pair of cheap "full range" coax drivers for the fun of it and as a
way to gain experience in measuring Theile/Small parameters and
playing with enclosure concepts.
The question: I have read that high Qts drivers "will not work" for
bass reflex designs. I measured the small 5.25 inch drivers I am
experimenting with (typically used in car audio) as having a Qts of
1.88. Based on my reading, this is very high (is it even too high a
value to be real i.e. I screwed up the measurement?). I would
appreciate any insight or discussion on why high Qts drivers and bass
reflex don't go together. TIA, Steve
glenbadd
wrote:
>
> I have been an analog circuit designer for decades and I have recently
> rediscovered my interest in speaker design. I am playing around with
> a pair of cheap "full range" coax drivers for the fun of it and as a
> way to gain experience in measuring Theile/Small parameters and
> playing with enclosure concepts.
>
> The question: I have read that high Qts drivers "will not work" for
> bass reflex designs. I measured the small 5.25 inch drivers I am
> experimenting with (typically used in car audio) as having a Qts of
> 1.88. Based on my reading, this is very high (is it even too high a
> value to be real i.e. I screwed up the measurement?). I would
> appreciate any insight or discussion on why high Qts drivers and bass
> reflex don't go together. TIA, Steve
Because you can't achieve a reasonable bass response in a
reasonable box size and tuning, compared to what can be
acheived for similar sized drivers than have a better Qts
range for bass reflex, around 0.3 - 0.5.
Dick Pierce
> [ Moderator's note: while this initially appears better
> suited for rec.audio.tech, I think that this may be a good
> starting point for a discussion of technical figures and
> their relationship to quality. -- dsr ]
>
> I have been an analog circuit designer for decades and I have recently
> rediscovered my interest in speaker design. I am playing around with
> a pair of cheap "full range" coax drivers for the fun of it and as a
> way to gain experience in measuring Theile/Small parameters and
> playing with enclosure concepts.
>
> The question: I have read that high Qts drivers "will not work" for
> bass reflex designs. I measured the small 5.25 inch drivers I am
> experimenting with (typically used in car audio) as having a Qts of
> 1.88. Based on my reading, this is very high (is it even too high a
> value to be real i.e. I screwed up the measurement?).
Well, you may or may not have screwed up the measurements, but
a Qts of 1.88 is not at all out of the realm of possibility.
Generally, car aduio drivers have fairly low compliance suspensions,
because you can't predict ahead of time what the enclosure volume
is likely to be, so you make the suspensiojn compliance the
dominant and controlling mechanical stiffness in the system. All
other things being equal, a high compliance leads to a high
resonant frequency and a high resonant frequency leads to a
high Qts.
> I would
> appreciate any insight or discussion on why high Qts drivers and bass
> reflex don't go together. TIA, Steve
Let's look at one way of viewing what a reflex system does.
Imagine a driver whose combination of motor system and mechanical
properties results a strong motor system (a high Bl product for
the voice coil's DC resistance) and moderate resonant frequency
(relatively low mass and high compliance).
I use the terms "high" and "low" here rather than actual
numbers for the moment because it is really the ratios
of some of these numbers that is important.
Among other things, there are two important results of this
particualr arrangement:
1. The ratio of the motor strength to moving mass contributes
to a higher electro-mechanical efficiency,
2. The ratio of the mechanical reactances to losses in the
driver lead to a fairly high degree of damping at resonance,
indeed, reflex-designed drivers are significantly over-
damped.
Now, put such a driver in a sealed box (or an infinite baffle)
and you end up with a system which, again, is significantly
overdamped at cutoff. Another way of saying this is that the
frequency response at resonance is significantly down from the
passband response.
Now a reflex arrangement adds a second resonant system, consisting
of the enclosure's acoustic compliance combined with the effective
acoustic mass of the port. That second resonant system has two
very tightly interconnected effects on the system:
1. It has the highest coupling to the outside air at its resonance,
2. It provide a very low impedance load for the driver at resonance.
The net result of this is that around the driver's resonance,
assuming the system is designed properly, the acoustical output
of the driver is dramatically reduced, and its acoustical output
is replaced by the acoustic radiation of the port.
In order for this to work (to do its job), the range of electro-
mechanical parameters of the system have to be constrained to a
relatively narrow range of values.
A figure like Qts, the total driver Q at resonance is actually
a combination of other values, notably Qes (the driver Q at
resonance due to electrical damping) and Qts (the driver Q at
resonance due to mechanical damping). But these value are derived
from the mechanical properties, such as the electrical losses,
mechanical losses, moving mass, magnet strength, and so on.
These "Q" figures are simply a measure of the ratio of the between
the reactance, or the amount of energy stored, at resonance to
the amount of energy that is dissipated by losses, mechanical or
electrical.
Be all that as it may, the major benefit of reflex tuning occurs
when the total Q is moderately low (0.3-0.5, give or take), and
the enclosure size and enclosure tuning frequency are in the region
of the driver's effective volume of compliance and resonant frequency,
respectively. Those numbers maximize the reflex's primary benefits:
1. maximized efficiency, bandwidth and enclosure volume ratios,
2. maximum reduction in driver excursion around the cutoff
frequency leading to lower distortion, increased power
handling and reduced vulnerability to mechanical failure,
3. Best compromise of bandwidth vs transient response
Stray outside of that regionj, and these benefits diminish rapidly
to the point where you get beyond a Qts of maybe 0.7-1, there is no
advantage in almost any respect and get much beyond that, most
gross operational paramaters of the system start to get MUCH worse.
Frequency repsonse becomes severely more complicated, transient
response, already pretty poor, gets markedly worse, excursion can
ftn get much worse and, with a Qts already as high as it is, the
driver itself has no efficiency advantages to begin with.
To more thoroughly evaluate what the possibilities are, we need
to know a more complete set of parameters, but let's consider
some hypothetical 5" driver with a Fs=75Hz, Vas=10L, Qts=1.88,
and Xmax = 0.2cm.
IN an infinte baffle, this driver would already have a
significant response peak of 5.8 dB at 75 Hz. It's output
would be excursion limited at 75 Hz to an SPL of around 77
dB.
In a sealed box of 10L, that peak now becomes nearly 8.6
dB high at 92 Hz, with a similar limiting at low frequencies.
Trying to evaluate what it does in a reflex enclosure becomes
difficult, because there are so many choices and if the criteria
are some combination of flat response, best transient response,
enclosure size, power handling or the like, pretty much all the
choices are bad.
Let's just take our 10 liter enclosure, pop a port in it and
tune it to 75 Hz. The result is a 14 dB peak at 120 Hz, a dip
at about 63 and a small peak at about 48. While power handling
is improved over a very narrow band around 75 Hz, it's
demonstrably worse around it.
Well, let's try a different configuration: enclosure is 30
liters and the enclosure is tuned to 30 Hz (a "design" I have
seen advocated for some of the "rice-paper" drivers). Such
a system is remarkable in that it's small-signal response
"extends" down to about 31 Hz. Sure there's a 9 dB peak at
90 Hz, then a 3 dB dip at about 43 Hz and finally a 5 dB
peak at 34 Hz before it takes a precipitous plunge towards
silence below that.
And while its power handling is improved in the 40-60 Hz
region, it is seriously excursion limited below that
compared to any other alignment.
--
+--------------------------------+
+ Dick Pierce |
+ Professional Audio Development |
+--------------------------------+