Help needed with woofer "buzz"
Paul Winser
> The IMF's are most likely the old transmission line models, which can be
> viewed (despite the strident claims of some) at the cutoff region as
> seriously mistuned vented systems, in the sense that they have whopping
> high vent masses and relatively low enclosure compliances. The result is
> a serious loss of control below system cutoff, a known problem with these
> sorts of systems.
Dick, you've criticised "transmission line" speakers a number of times
on this forum. Are your objections purely on analytical grounds, or
can you relate the "serious loss of control below system cutoff" to
how these speakers sound? Although there have been some bad examples
over the years, there have also been ones that have been praised.
John Wright seems to able to pull off whatever it takes to make such a
speaker sound good. I certainly like them.
Of course these speakers are not behaving as true transmission lines
(at least, not the ones that fit in my apartment), but I and others
I know favour the sound over common reflex speakers of similar size.
Paul
Richard D Pierce
Paul Winser <pa...@prl.philips.co.uk> wrote:
>DPi...@world.std.com (Richard D Pierce) writes:
>
>> The IMF's are most likely the old transmission line models, which can be
>> viewed (despite the strident claims of some) at the cutoff region as
>> seriously mistuned vented systems, in the sense that they have whopping
>> high vent masses and relatively low enclosure compliances. The result is
>> a serious loss of control below system cutoff, a known problem with these
>> sorts of systems.
>
>Dick, you've criticised "transmission line" speakers a number of times
>on this forum. Are your objections purely on analytical grounds, or
>can you relate the "serious loss of control below system cutoff" to
>how these speakers sound? Although there have been some bad examples
>over the years, there have also been ones that have been praised.
>John Wright seems to able to pull off whatever it takes to make such a
>speaker sound good. I certainly like them.
My criticism of transmission lines stems, of late, primarily from the
ridiculous claims made by some of the more vocal and less knowledgable
proponents of the design, most notably some manufacturers who have made
claims that are at complete odds with physical reality: most notable that
they are 1st order systems, which is utter hogwash, secondly that they
are acting as transmission lines where the wavelength is far in excess of
the dimensions of the line, and it is the transmisison line effect that
is responsible for their great low end.
Regardless of whether they have a great low end or not, the extremely
well understood field of physics has some very different things to say
about the action of lines at low frequencies, and does so in a way that
makes testable predictions. For one, if you assume a model based on a
lumped-parameter system topologically identical to a reflex, it predicts
a particular respose. The transmission line paradigm (I won't call it a
model because it has, at best, vague mathematical underpinnings) predicts
something else. For example, there's a lot of hand waving about 1/8 wave
resonances and stuff like that, which is utter tripe, since there is no
such thing.
In actual measurement, indeed, the lumped parameter reflex model more
accurately (and by a lot) predicts the action at low frequencies, and the
empirical line pardigm does not. This goes not only for simple frequency
response, but the impedance and so on.
One of the predictions of both models is that they suggest (albeit to
different degrees) that the woofer becomes unloaded at low frequencies.
And this is something that is born out in actual practice. One of the
consequences of this (and is also a problem with reflexes, although not
necessarily to as great an extent, depending upon tuning parameters) is
the extreme cone motion that results below cutoff. Given a source that
may have a large amount of very low frequency information (record warps,
for example), the result is enough signal (and not very much is enough)
to drive the cone into excessive cone motion. The result is that both the
suspension compliance and Bl product, which are position dependent in ALL
woofers, is now varying in time with the result that any higher frequency
information is modulated by the low frequency junk. This generates LOTS
(and by LOTS I mean 20-50% LOTS) of intermodulation distortion. It also
is unsafe for the driver because it severely reduces the mechnical
excursion margin.
>Of course these speakers are not behaving as true transmission lines
>(at least, not the ones that fit in my apartment), but I and others
>I know favour the sound over common reflex speakers of similar size.
And there are good reasons for that, let me advance a few:
1. Because of the high losses in a transmission line, it may be less
sensitive to driver variations. Driver variability is a problem with
optimized low-lss reflex systems in that the produce fairly wide
variations in response (couple of dB maybe).
2. The system IS actine as a transmission line at higher frequencies. The
rear wave IS being lost and absorbed at those higher frequencies ~150
Hz and higher. This can reduce significantly the coloration caused by
internal reflections and re-emission.
3. As an "accidental" secondary benefit, the construction methodology
employed in many TL designs results in a cabinet that's extraordinarily
well-braced, massive and rigid, all of whichg are very desirable
features in a cabinet. But you can also build (and I have) a reflex
or a sealed box that is also well-braced, massive and rigid and,
interestingly enough, it sounds very much like a TL in the upper bass
and lower midrange. And it has none of the disadvantages of a line.
4. Even with the existance of a lot of sound theory and a tremendous
realm of supporting empirical evidence at their disposal, any idiot
can design a loudspeaker and, unfortunately, many do. It's just harder
to do a reflex right.
John Wrights designs were the result of a little luck and a lot of
persistance in the near total absence of any sound theory. One also has
to appreciate that in the early 1970's, the IMF Studio and Monitor were
exceptionally good speaker as much be design as by default. They faced
such awesome competition as Rectilinear, JBL L100's, AR-3a's and, even,
the Bose 901. The IMF's clearly run simply because most of the
competition wasn't.
But then we saw the "improved" IMF Studio and the ALS-40, two of the more
badly conceived and poorly executed system around. The ALS-40, which used
Bud Fried's "active line" technology, were a joke, sounded lousy and
performed the same. Yeah, John Wright did good work, even if nobody knew
why.
--
| Dick Pierce |
| Loudspeaker and Software Consulting |
| 17 Sartelle Street Pepperell, MA 01463 |
| (508) 433-9183 (Voice and FAX) |