[D OB G]

Hi Martin,

I've just got Praxis (to replace Imp).
I'm still on a learning curve, but these are the traces for each of the two woofers, upper and lower, connected in parallel, of the Audiosyncrasy ("An OB Design") taken with the Nearfield Bass Script, with the mic within a cm or so of the dust cap as directed.

It strictly measures to 72 Hz, but I've octave averaged out to 100 Hz to give a reference level. (I don't know if room effects play any part in this technique). As you can see the vertical scale is 3 dB per octave, and the horizontal goes from 10 to 100 Hz (I've just realised it's a bit hard to read).

So this is the series inductance equalisation method using 32 mH, and with each driver having an Fs of 24 Hz (verified using the impedance function), on a baffle 850mm X 500mm.

It obviously flies in the face of the problem of the rising impedance as Fs is approached, the problem you pointed out, yet I reckon I would put the -3 dB point somewhere below 30 Hz.

I still can't explain why it works, against fundamental theory, except to say that it seems to!

David

[JohninCR]

Nearfield measurement below the Fequal frequency is all but meaningless, because it excludes the effect of the rear wave.  It's only useful as a starting point by telling you what the driver does; but the baffle, placement and room are so dominant in the bass frequencies that I feel they are more important than the driver itself as long as it has reasonable low frequency capacity.

[D OB G]

The issue, JohninCR, that Martin raised, is that trying to equalize below the Fequal with a series inductor can't work because of the impedance hump as Fs is approached.
It seems to me from this graph that we have come mighty close to being 3 dB down at Fs with inductor equalization (which is obviously a trivial matter with active equalization).
If the same driver was placed on the same baffle (remember that this is a low Qt driver) without any equalization, do you think that we would get this extended response either in the near field, or at the listening position?
With the nearfield measuring technique (which in this case was a one second chirp pulse), won't many cycles of the rear wave, and rarefaction, reach the mic in the time it takes to make the (averaged) measurement? (In my latest posting on "an OB design", I take measurements with pulses of 4, 7 and 10 seconds).
How do you think that the response of equalized drivers (as Linkwitz uses) should be measured?
How should a high Qt unequalized driver be measured to ensure a flat response (i.e. that the Qt is indeed right)?
Sorry about all the questions, but issues of driver/baffle matching and baffle/room matching seem to be addressed differently amongst us open-bafflers.

David

[Rudolf]

With the nearfield measuring technique (which in this case was a one second chirp pulse), won't many cycles of the rear wave, and rarefaction, reach the mic in the time it takes to make the (averaged) measurement?

With a 50 Hz wave almost 50 cycles of the rear wave would reach the mic during measurement. But please remember: Your mic is 1 cm from the front cone center, but probably 50 cm from the rear cone center. So what you measure at your nearfield mic position is almost infinite baffle, but not dipole behaviour.

How do you think that the response of equalized drivers (as Linkwitz uses) should be measured?

The path length difference at the mic between rear wave and front wave should be significantly less than 1.5:1 (not 50:1), so 1-2 m distance from the driver seems to be a minimum.

How should a high Qt unequalized driver be measured to ensure a flat response (i.e. that the Qt is indeed right)?

Same as above

Sorry about all the questions, but issues of driver/baffle matching and baffle/room matching seem to be addressed differently amongst us open-bafflers.

There are many different roads to reach the same target: a flat response at the listening position. And that wouldnt be flat (at least in the bass region) without accounting for the room influence.

[D OB G]

Hi Rudolf,

Here is a graph, taken at 1700 mm, overlaying 5 raw traces and 5 octave smoothed traces, on a baffle 850 X 500 mm.



There is obviously a lot of noise below 30 Hz.  The speaker is right next to the TV on one side, and a couch on the other, and about 1200 mm from the wall behind.

There is obviously an in-room peak at 45 Hz, and a shelf from 55 to 85 Hz.

Referenced to the 100 to 200 Hz band, we still seem to be about 3 dB down at about 30 Hz.

My point is not to disagree with your point about room influences, but to demonstrate that a series inductor can equalize close to the Fs, in spite of the impedance hump properties that would imply that it cannot.

(We haven't really altered what the dipole baffle frequency response is when taken by the nearfield technique, so is it any less valid for determining the low frequency response of the baffle itself, separate to where, and in what room, and in whose house, it is?)

David

[D OB G]

Hi Rudof again,

I have just re-read Linkwitz (http://www.linkwitzlab.com/sys_test.htm) under Dipole woofer, and to quote (my emphasis):

"The reponse follows what we would calculate from the impedance curve for a closed box woofer (which as you said is almost infinite baffle) with a -3 dB point at 13 Hz for Qts = 0.7.  It is also the dipole response in the far field, after the +6 dB equalization has been applied."

David

[Rudolf]

Hi Rudolf,
My point is not to disagree with your point about room influences, but to demonstrate that a series inductor can equalize close to the Fs, in spite of the impedance hump properties that would imply that it cannot.

David,
my personal experience was that a series inductor would work indeed, but far less efficient than theory without regarding the impedance hump would have predicted.

We haven't really altered what the dipole baffle frequency response is when taken by the nearfield technique, so is it any less valid for determining the low frequency response of the baffle itself, separate to where, and in what room, and in whose house, it is?

With nearfield measurement you haven't altered the dipole baffle frequency response, but you only measure its nearfield response! I don´t see any practical value in doing that except using it according to John Kreskovsky´s suggestion: http://www.musicanddesign.com/measure_dipole.html.

Rudolf

[D OB G]

Thanks very much Rudolf,

That is certainly a comprehensive site.
One interesting thing I noticed is that what I thought was noise below 30 Hz actually shows the compound result of a room with a resonant frequency of 45 Hz, and the real rising response below 15-20 Hz (not much benefit I admit, except maybe for the overall retrieval of ambience on the recordiing, if it is there) of a room that is not sealed, but opens into other spaces, something I had not come across before.
The site certainly demands some close reading.

David