But didn't derive linear decay.. d@mit.(..or it least doesn't have it there.)
Can you get that from the impulse data?
Yes. (..in fact from what I remembered seeing before (not Augerpro's measurements) - Earl had provided the data-points and someone else had derived the CSD plot with them.)
Of course what would still be missing is linear decay in the context of the overall design (i.e. filter, midbass driver, baffle included). It isn't misleading to have these absent, but it's nice to get a clearer picture of what you can expect in a resulting loudspeaker.
Last edited:
So no one has posted an impulse of Geddes Summas ?
Seems like some owner out there would have done that by now, they are kits and I have never seen kits not measured?
???
http://www.diyaudio.com/forums/multi-way/100392-beyond-ariel-108.html#post1821292
An externally hosted image should be here but it was not working when we last tested it.
An externally hosted image should be here but it was not working when we last tested it.
An externally hosted image should be here but it was not working when we last tested it.
An externally hosted image should be here but it was not working when we last tested it.
Michael
Thanks Michael!
That's it...and yes, it displays the sort of linear decay problems I expected. I should also note that the B&C driver has it's own resonance pretty well suppressed, so not much of what's displayed here is the result of driver resonance.
Again though, at the intended reference axis nearing 30 degrees off of the 0 degree axis, it's probably swamped by the low-pass filter.
I'd like to see Jean-Michels design's CSD off-axis as well (..though I believe he recommends listening on the 0 degree axis).
Well, I say OS because over 1/3 of the curve is OS (not 90deg included agle though) for the initial expansion, although the first 4mm of the throat is slightly modified due to the concaved dome. Junction between the driver and WG makes a difference because normally it cannot match perfectly, and effects the high frequencies. So that needs to be accounted for as well. Of course the lip is done as I explained before. In the end, it does not "look" like an OS waveguide.
It's like doing the "Thomas Edison" approach on the computer. We've done that quite often to get some sexy aircraft curves.
A typical compression driver has a power response (on a plane wave tube) that resembles a lump, similar to a low-order bandpass, usually centered (for 1" throats) around 1-2kHz. CD devices emulate this, hence CD EQ requirements. Non-CD geometries tend to boost highs more, at the cost of additional directivity, and nonlinear power response. The beaming of the driver contributes some top-end differences but the above generalizations are largely accurate. JBL and some others post plane-wave responses of their drivers, take a look at the JBL 2426h datasheet (or an un-eq'd OS FR)
The power response follows the velocity response if the power is directionally independent, as it is in a plane wave tube. The velocity of the diaphragm in a compression driver is a highly damped single degree of freedom system, hence a single peak that is basically flat at resonance (high damping) and falls at 6 dB/oct on either side of resonance. Hence this is the power available. At low freq the response is always wide directivity so all compression drivers on horns must be HP because of this. Then there is a relatively flat region arround resonance and then the response tedns to become completely dependent on the horn because the power response is falling, but the directivity can be changing so any number of things can happen. IF the axial response does not fall then the device cannot be CD, its that simple.
Where?? We are talking specifically about compression drivers mounted on horns/waveguides not flat plannar drivers. You also need measurements well beyond 30 degrees to call a device CD. Can you post links to where your measurements are that use a compression driver and extend past 30 degrees??
Rob
Michael, can you link to where? I've looked, and I know it hasn't been too long ago that you were disputing the claim, but I don't ever recall you actually posting a graph that was constant directivity. I remember you posted some graphs that were not too far from it though. Could you also post the graphs with just the driver and no horn or WG?
Thanks,
Dan
Thanks,
Dan
Shouldn't an AMT driver have a different acoustic loading profile at the vibrating membrane compared to a planar driver or a voice coil driver? This makes a hell lot of difference in the gain.
The acoustic loading of a horn or waveguide is independent of the type of source at its throat.
If the pressure response is independent of direction, i.e. CD and the axial response is independent of frequency, then the velocity of the source must be independent frequency, just as the power response is. How is this possible? No source can have a velocity that is independent of frequency (unless its EQ'd of course). Did physics somehow change and nobody told me? Oh its the new Mige0 physics! Now I understand.
If the pressure response is independent of direction, i.e. CD and the axial response is independent of frequency, then the velocity of the source must be independent frequency, just as the power response is. How is this possible? No source can have a velocity that is independent of frequency (unless its EQ'd of course). Did physics somehow change and nobody told me? Oh its the new Mige0 physics! Now I understand.
Last edited:
May I suggest - we'd better might look at the topic the other way around !
So
– who's to explain where *exactly* those CD behaviour can be found in the (new) SUMMA polar plot below ?
Summa
An externally hosted image should be here but it was not working when we last tested it.
Just overlay a high lighting frame on the CD area in that plot - and then re-post again - *please*
Michael
Last edited: