2009-2010
June: Album art presentation by Rich Shelton. Rich is a artist and musician who is fascinated by album art and what it tells us about the relationship between popular music and the graphic art of its packaging. Richard will do a slide show with some of his favorites from his collection of 18,000 albums.
May: Speaker unveiling of new club SB Acoustics setup with all passive crossovers.
April: Audio by Van Alstine partners with Nate's audio (local loudspeaker guy Nate Field) to put on a nice show.
March:
February: Bel Canto demonstrates some new products and high resolution audio recordings - and brings in a fantastic high-resolution, high-sensitivity loudspeaker featuring a fantastic new horn by Peavey. Ceiling tiles were bouncing that night ;)
January: Bring and Brag
December: A special Loudspeaker 101 meeting with Ron Ennenga, and a concurrent Turntable setup meeting with Wally Malewicz of WAM Engineering.
A brief (!) wrapup of the Loudspeaker meeting from a forum post by Ron is as follows:
Alright, just to get things going, here are some more complete,
written answers to some of the questions that I recall were asked at
the meeting. Loudspeakers are a complex topic and a thorough treatment
of just box design could take a whole year of monthly one hour meetings.
If you want to contact me privately with questions, you may also do that.
If there are questions about individual parameters and what they mean, perhaps
the best resource is the thiele small article at wikipedia. It has
reasonably good descriptions of the individual parameters.
1) How does Vas affect box size?
Box size is a function of Qts and Vas for both sealed and vented
boxes.
for sealed boxes .Qtc/Qts=sqrt(Vas/Vb+1) => Vb=Vas/((Qtc/Qts)^2-1)
for sealed boxes box size varies inversely with the square of the
ratio between initial and final Q.
In a vented box the most instructive way to discuss this aspect of box
size is to use a curve fit:
Small-Margolis (1981 JAES article) says box size varies as: Vb=Vas*20*Qts^3.3 and Fb =
0.42*Qts^-0.96
D.B. Keele, Jr says: Vb = Vas*15*Qts^2.87 and Fb=0.42*Qts^-0.9
....so Box size varies approximately with the cube of Qts and in
proportion with Vas. Tuning frequency varies inversely with Qts.
In both sealed and vented, box size is doubled when Vas is doubled,
assuming all other variables remain constant.
2) Will 2 smaller woofers have better transient response than one
larger woofer?
Not necessarily. Transient response (in the sense of step or impulse
responses) is a function of several things. First is high frequency
response and second is low frequency response. A smaller woofer will
generally have better high frequency response so a measured step or
impulse will be sharper (faster) when no crossover is used. Once a
crossover is added, there is no advantage to wider bandwidth, since
both now essentially have the same high frequency bandwidth. In low
frequency response, the Q and cutoff frequency (or Vb and Fb in vented
boxes) determine the response shape and the resonant behavior of the
speaker. Low Qtc in sealed boxes is often identified as the source of
"fast sounding" bass. I have heard speakers with Q's as high as 1.3
attributed with the adjective "fast" or punchy. I think this is
sometimes due to a bump in the "punch" region 100-150Hz and smaller
woofers are more likely to have this.
Two smaller speakers _with the same response shape_ as one larger
speaker should not sound different at low frequencies from the larger
woofer. Smaller woofers often have higher cutoffs (F3 or Fb), though,
which brings us to a third factor: how the speakers match with the
room. Room resonances tend to decay in times on the order of seconds,
where speaker system resonant decay in normal systems tend to decay in
milliseconds. The room can thus dominate the sound in this area.
Below the frequency where your room has its lowest standing wave ( F=c/
(2*L), where c is speed of sound and L is the room's longest
dimension) the room starts to pressurize and there is a theoretical
12dB/octave gain in response. In practice, leakage and doors to
neighboring rooms reduce this and the resultant is more likely to be
less than this in practice. This gain, when coupled with flat response
speakers, can cause a lifting of bass response that results in narrow
peaks for a vented system, and broader peaks for a sealed system.
Designing the system with a drooping response cna match the room
better and this is one advantage to rolling your own, in that you can
desighn the system for your room, Larger rooms should have less of an
issue with room gain causing bloated bass. In a smaller room, a
dipole source that does not pressurize the room at low frequencies
might be considered.
Another factor is distortion. Distorted bass sounds slow and fuzzy.
Use good quality drivers and don't push them too low.
Another factor is voice coil inductance. Smaller woofers have smaller
coils with lower inductance (usually). Excessive voice coil inductance
can cause response shape changes in the low frequency area that are
not modeled by most software. The result is commonly a bump in the
50-70Hz region that happens to coincide with the standing wave from
floor to ceiling in most rooms.
So the short answer is that bass response is more complicated than
just woofer size. ;)
3) What about dimension ratios?
Various dimension ratios have been put forward as having beneficial
effects for speaker boxes. The usual argument is that they distribute
the resonant modes more evenly. While this is true in an undamped
box, they have a much smaller effect if there is box damping and there
is no effect of frequencies are restricted to frequencies lower than
the standing wave frequencies.. For bass-midrange boxes, even if well
damped, avoiding dimensions that are simple multiples is still a good
idea, but the golden ratio, (phi=1.6180339887499... ;) ), is not some
magic number that is any better than many other ratios. Dimension
ratios are more important in mounting drivers on a baffle. On a
rectangular baffle, the smoothest response results from dimensions
that are different (nonrelated) distances from each of the 4 edges.
Now on to some stuff that I didn't get to expound upon cleearly enough
in the meeting. We are in an awesome era where tools to design
loudspeakers are more affordable and more accurate than ever. The
tools below are free or very cheap. We start with measuring T/S
parameters. Mfg T/S should generally be used as a guide for driver
selection only. Their spec sheets can be very useful for hard to
measure Specs like Cone area (Sd), etc...:
Measurement Software:
HolmImpulse - measures frequency response
Speaker Workshop - measures frequency response and T/S parameters.
Jigs and mics could cost up to ~$150-200, but can be made for a few $
if you are handy with a soldering iron and have some parts in a junk
box..
Woofer Tester - http://www.woofertester.com/ costs ~$160 - measures
driver impedance, and they have more expensive products that also
measure frequency reponse, etc...
Once you have T/S parameters, you can model box size:
Box modeling software:
AJDesigner - simplest http://www.ajdesigner.com/
WinISD (WinISD beta is simpler but less informative than WinISD pro
alpha)
Unibox - most complex - latest versions may be found at:
http://audio.claub.net/software/kougaard/ubmodel.html
Now to figure out how your box shape will work - only the BDS below
will do nonrectangular and rounded edge modeling.
Diffraction modeling software:
The Edge http://www.tolvan.com/edge/ - Svante also makes a nice
looking inexpensive box modeler called Basta (I have never used)
BDS and many other tools - http://www.pvconsultants.com/audio/frdgroup.htm
Jeff Bagby's Box and Room modeler (just out so pay attention for
updates) - http://audio.claub.net/software/jbagby.html
Now you measure the response on the baffle and start doing a
crossover:
Crossover modeling software:
Jeff Bagby's PCD (Passive Crossover Designer) - http://audio.claub.net/software/jbagby.html
Speaker Workshop - very powerful, optimizer, etc. (Jeff's tool above
does polar response, though.) SW is also open source:
http://sourceforge.net/projects/speakerworkshop/
Various versions of Spice. LTSpice from linear technology is a free
variant very popular with hobbyists
http://www.linear.com/designtools/software/
November: ARC Tour - we do a factory tour of Audio Research Corporation in Plymouth Minnesota and listen to some of their new "digital" amps and some very high end speakers in their listening room - more details and pics at our Blog!
October:
September: