Scanspeak 18w/4531g

[ErinH]

For test setup, procedures, and information on how the data was obtained see sticky here:
http://www.mobilesoundscience.com/f11/driver-testing-methods-setup-frequency-response-harmonic-distortion-693/
Do not ask setup questions in this thread. This thread is only for driver discussion.

To understand how to read the data, try here:
How to read Frequency Response and Harmonic Distortion Plots - DIYMA.com



Driver Tested:
Scanspeak 18w/4531g:
[no pictures taken]


Impedance and T/S Parameters:



Frequency Response at 0 degrees, 3dB scale:


Frequency Response at 0, 15, 30, and 60 degrees, 5dB scale:



Harmonic Distortion at 95dB:

[cvjoint]

One of the few true wide bandwidth 7"s out there imo. Smooth Fr and very good off axis performance. In the HD department 2nd order is merely average from what I've seen but the rest is class leading. One thing I've noticed with the Scans is that they tend to have a little hiccup seen in the HD and impedance plot, this one is around 800hz. Nothing sizable for sure.

[Adam_MSS]

It's interesting that the impedance anomaly you mentioned also lines up with a sharp drop in 3rd order distortion. Certainly not problematic, but I'm interested to know more about what causes this. Mechanical phenomena? Edge damping of cone resonance? The little cone ridges?

[86mr2]

Isn't that the surround diffraction area for a 7"?

[cvjoint]

My best guess is cavity resonance. By cavity I mean the dustcap to pole piece gap.

[86mr2]

I found this after a quick search:

Presence Notch
Context: The speaker has a broad presence notch .
Description: The loudspeaker has a major fundamental resonance which is mainly determined by the spider and the mass of the cone. A second resonance occurs at the surround. It's not major, but can have an dramatic effect on the frequency response of the speaker. The surround plays an important part in the overall sound, because it acts as a terminating impedance for the waves travelling out from the voice coil to the edge of the cone. If the surround goes into an oscillation of its own, it will either disconnect from the cone and refuse to terminate the waves, sending them back towards the voice coil, or at certain frequencies will completely absorb the waves. This surround resonance is around 800 to 1000hz. Usually, in a frequency response plot, it appears as a dip around 800 to 1000hz. This dip can be as much as 6db deep and can be as broad as 200hz. Since a 10db difference in sound level is perceived as a doubling or halving of sound, the 6db drop at the presence notch would certainly be noticeable. There would be a pronounced drop in the mids. The player would have a tendency to turn up the volume to restore the mids, with a resultant increase in the lows and highs. Hence the term presence notch.

I think I've seen another explanation that predicts a notch/peak combination, but I can't find it right now.

[cvjoint]

^^Great find. It fits in with the observed data perfectly. My theory would have predicted a peak anyway.

[ErinH]

Indeed, great find. Anymore links like that?

[86mr2]

Like I said, I found this site after a very quick search, and it is actually an instrument speaker site. I know damn well that I have seen better discussions of the various resonance artifacts in FR and HD plots, but I'm kind of thinking that they are scattered across 10 or 15 years of DIYAudio posts in the loudspeaker sub-forums.

Speaker Terms