| Living with Low-Q Loudspeakers | Return to Index |
Before Thiele and Small analysed the bass-reflex speaker system, and published the results in 1961
(sic), the design of such speaker cabinets was empirical. In conjunction with the low power amplifiers of the day, one emphasis put on speaker drive-unit design
was to obtain high efficiency. This lead to speaker designs with large, powerful magnets, particularly if
the mass of the cone was not to be lowered to the point where it compromised other aspects of
performance (such as an irregular mid-range response).
These historical aspects have left a pool of older-style drive units that have powerful magnet systems.
Of course, not all old drive units have powerful magnets, and likewise, some modern drive units also
feature powerful magnet designs (e.g. Altec/JBL/Tannoy).
The Thiele & Small parameter that reflects the performance of the "motor" system, including the
magnet design, is the "Qe" parameter (e for Electrical). In general this parameter varies from around
0.2 through to 2. The speakers with more powerful motor systems have a low Qe. The total Q of a
speaker is the combination of the mechanical Q (Qm), and the electrical Q (Qe). In a speaker with a
low Q such as 0.25, the contribution from the mechanical Q (say around 3) is quite small, so the total Q
may be about 0.2.
Conventional wisdom when designing a loudspeaker, is to tune the system so that the frequency
response is as smooth and as extended as possible - no peaks, and no early roll-off. This in turn implies
a specific Q (of around 0.5 to 0.7) for the combined speaker-in-box. When a drive unit is placed in a
box, the Q is raised. A small box raises the Q more, and a large box raises the Q less.
|
With a driver that has a very low Q to start with, a small box is required to achieve an overall Q of 0.5. For example, for a 15" Tannoy with Qt=0.2, to raise the Q to achieve a maximally flat response the required box volume is only 1.3 cubic-feet (and the reflex port is tuned to 47Hz)! As can be seen from the graph, this gives a very high cut-off frequency of about 75 Hz - atrocious performance from a 15" drive unit! But then again, no one puts 15" drive units into 1.3 cubic-foot boxes - do they?
B: Larger Box
The intuitive reaction is to make the box bigger. As Gilbert Briggs (of Wharfedale) used to say: "smooth runs the brook where the flow is deep". In other words, get a big box, and you get smooth response. If the box is made larger (7 cubic feet), but the port frequency is left unaltered, then this kind of isn't true - see curve B where a peak in response becomes evident from the large volume - but the peak is not really severe and could be acceptable to some tastes.
C: Extended Response
Finally, take the larger 7 cubic-foot box of B, and shift the tuning downwards so the port frequency is 30 Hz instead of 47 Hz. What appears is a gradually drooping response from 200 Hz downwards, propped up by the port resonance at 30Hz, and then rolling off steeply below that. This response is sometimes referred to as a low frequency shelf.
The disadvantages of this alignment are:
-
it does not follow conventional wisdom, which demands a maximally-flat approach to alignment.
it could, under certain circumstance sound bass-light, or thin.
-
it lends itself to operating the loudspeaker in a room, where reflections conspire to give a boost to low frequency output from a speaker, particularly when near a corner
it could be easily equalised to remove the low frequency shelf - in fact the normal Baxandall tone control exhibits, when set to give about 4dB boost, an equalisation very close to that required
it gives, when either equalised, or used in a room, the best low frequency extension for a low-Q drive unit
Summary
Living with low-Q speakers requires an un-orthodox approach to the enclosure alignment - one where a larger-than-predicted box should be used, and the port should be tuned lower than predicted. When this is done, practical circumstance such as room placement, or use of tone controls, leads to satisfactory low frequency response.
It is interesting to note that this is more or less exactly how "decent" speakers from the 1950's were operated - large boxes, placed in the corner (or, smooth runs the brook where the flow is deep). This contrasts with the current design approach where the Q of the driver is chosen (higher) to get maximum extension and smoothness from smaller boxes, but placed away from room corners.
© Steve Spicer, February 2002 Steve Spicer
s.spicer@trl.oz.au
No part of this work may be reproduced without permission of the author.