Large midrange for OB??? Scott G ?

Ah.. A challenge!

How large? How eff.? What nominal impeadance?

What sort of off-axis performance for a given freq.? (i.e. do you want the off-axis performance to coincide with a dipole's radiation on a small baffle?) Whats the off-axis performance of the waveguide?

Do you want a response deviation and/or a high "q" driver to compensate for the loss in spl from the dipole? (or are you going to eq. it?)

(..also of course how "flexible" are you with the above?)

Using the Edge.. we have a problem here with directivity and a free-air open baffle..

we are looking for a -3db point at 45 degrees off-axis (when the dipole is acting maximally as a dipole).

a 10" or even an 8" driver with as narrow a baffle as possible won't really display this behaviour until just below 300 Hz.

Now you *might* be able to achieve something more directive with a side aperiodic cardoid venting (..don't know, haven't tried this myself).

effectivly then what will likely have is a horizontal polar response that becomes less and less directive down, (i.e. wider dispersion), to about 450-500 Hz, and then the dipole will start "kicking-in" and dispersion will narrow and be constant somewhere below 300 Hz.

this may or may not be a concern to you.

by the way - the driver you mentioned looks very nice except for inductance (which increases IMD dramatically).. Their "Active Impeadance" series drivers (10 inch only) are conversly AWESOME (in this respect). Though less eff., perhaps the 10NDA520 would be a better solution (when considering inductance and dispersion).

Note:

I have know idea what the off-axis response of the driver is.. but, the Ciare OEM 12.75 NdMR looks pretty sweet in every category but excursion (but its SOOOO freakin eff.). (..and because its a 12" driver it should be more directive down to a lower freq.) However spl does drop quite a bit below 400 Hz..

http://www.ciare.com/oem/index.html

Price is not too bad either:

http://www.assistanceaudio.com/08_Ciare.html

Paul W said:

Good...I got the guy who knows every driver made by man!

Click to expand...

I'm not sure about that..

All else equal, the more eff. a driver the greater the resolution.. of course the same can be said for mms for a given sd.. the lower the mass the greater the resolution.

DDS Eng 1-90 Pro?

Hmmm, other drivers high resolution, high output, low distortion, low cost (relativly anyway). Tough one.

Well, there is the 18sound 8NMB420.. good excursion, LOW MASS for a pro driver (good resolution), decent inductance at 1 kHz for an 8 inch pro driver, and substantial excursion capability for an 8 inch pro driver. The only thing wrong with it (other than possible dispersion issues) is its lower fs. (..sounds odd huh?) In this case the lower fs means the driver will naturally want to extend its freq. response, thereby increasing excursion, thereby increasing distortion. Of course a good steep filter should fix that problem.

B&C PE21..frankly not as good as the 18sound 8 inch, but not bad (limited in output however because of x-max and particularly x-damage vs its sd). No off-axis data.

Then there is the new BMS Pro 8N215.. Mass is a little heavier than the above.. good excursion, good inductance - a nice driver.. No off-axis data though.

I've looked at a LOT of other Co.'s drivers.. all turned me "off" from one aspect or another.. Things like high inductance, poor excursion for a given sd and fs, high mass for a given sd, low eff for a given sd, poor high freq. extension, poor "break-up"/increased "ringing", availability, etc. ALL "played" a part in my driver sifting thought process. (..and of course cost was an over-riding factor.)

Of all the driver's I've mentioned.. I'd prob. go for the 12 inch ciare all things considered. It will likely be more detailed than the rest, while having the capability for greater spl's and a more directive freq. response.

IF I wanted a less directional driver I'd choose either the 8" 18sound or BMSPro.

From a purely "build-quality" perspective I'd almost certainly go with the BMSPro. ..Though I suspect all of the drivers listed are built VERY well.

So there you have it.. a very limited selection of drivers that *might* be well suited to your needs.

Just because it seems to fit in here - dispersion pattern of my ca. 48H 24W 7.5D (inches) shallow U-frame. Driver is a SS8543 (6.5"); data are without filters, the rolloff is due to low Q plus dipole effect. With hindsight the active impedance 18Sound 10" would have fit in here nicely.

Purpose of my post though is just to show the off axis response of such a baffle, I expected more dipole like behavior until up high but it seems it went towards cardioid in the mids w/o even trying. Back of the baffle completely open, no stuffing.

MBK said:

Just because it seems to fit in here - dispersion pattern of my ca. 48H 24W 7.5D (inches) shallow U-frame. Driver is a SS8543 (6.5"); data are without filters, the rolloff is due to low Q plus dipole effect. With hindsight the active impedance 18Sound 10" would have fit in here nicely.

Purpose of my post though is just to show the off axis response of such a baffle, I expected more dipole like behavior until up high but it seems it went towards cardioid in the mids w/o even trying. Back of the baffle completely open, no stuffing.

Click to expand...

Yours doesn't seem to hit actual dipole behaviour until about 230 Hz. The freq. response from 500 to 1 kHz is effectivly omni (in the forward plane) and only starts into a cardoid response above 1 kHz - strictly due to the dispersion character of the driver.

There is also the 12 inch active impeadance driver from 18sound that extends the limited dispersion down to lower freq.s.. unfortunetly the inductance of the driver (assuming it isn't a mis-print), is almost 20 times as much as the 10 inch driver.

I personally have the active impedance 10" driver although I havent got it playing yet in a finished speaker.
I regret buying it after reading about studies that conclude non linear distortion plays a small role in our perception of sound.

I spent a small fortune on it( relative to my pocket money) although it is a relative bargain compared to other offerings in the same class. In terms of non linear distortion it should be the best in its class, the active coil thingy I read is even better than faraday rings. There is literally nothing else out there with such low inductance and very high efficiency.
Although if I had to decide again i would simply buy the considerably cheaper 18sound mb600, eminence, or paudio 10"s without the bells and whistles.

To ultilize their potential(as well as the Ciare MR series mentioned) you want to hornload them because their frequency response drops considerably in the midbass.

For dipole use the Ciare midbass 12" 12.64NdW1 is more suitable than the 12.64NdMR. Still very efficient, much flatter response, and greater xmax.

The active impedance 12" is great too, not sure about the high inductance but im guessing it may not matter as its impedance is dead flat like its 10" relatives.

I should add some caveats to my above data: in-room and closer to the baffle than I would like (80 cm or 32") for such a large baffle - the full baffle effects likely don't show up this close for the higher wavelengths since the rear-to mic distance is considerably longer than the front-to mic distance.

The freq. response from 500 to 1 kHz is effectivly omni (in the forward plane) and only starts into a cardoid response above 1 kHz - strictly due to the dispersion character of the driver.

Click to expand...

Agree on the dispersion above 1k. But even a cardioid would be failry omni in the +-60 degree range. I should complement this to a complete 180 degree polar view.

Paul W said:

Scott,
The Ciare's choppy impedance curve is scary...I'm thinking floppy cone which isn't for me. On the other hand, the 10NDA520 looks better than the 10M600. I was wondering why I saw no mention of shorting rings in the 10M600...now I know where Eighteensound puts their technology! I'll check out the others you mention. What would you do if you were to break my bank?

Click to expand...

I noticed that to with the ciare - but it doesn't seem to manifest itself in the freq. response curve. Now it *might* with linear decay.. but here I'd be more concerned with getting the inital decay clean (say within 9+db down less than a milisecond). It would NOT be a floppy cone - in fact this cone should be incredibly stiff considering both the mass and material/shape, as well as the motor's field strength. If there is a mechanical problem, that would likely be at the transition between cone edge and surround (..i.e. stored energy). Again, this would prob. mean that the linear decay would be "messy" around 1.5 kHz, but it might not be a problem with a 4th order filter at 1kHz. Conversly I'd expect the driver to be otherwordly "clean" between 250 Hz and 800 Hz. Its "give and take" - no perfect driver and all that. Note however that stored energy IS something that can be significantly improved by a DIY'er with fairly modest modifications.

The more I think about the 18sound active impeadance drivers - the more I begin to suspect that the 12 inch driver's inductance value is a misprint by 10 times the value. (i.e. it should be .09 instead of .9) I'd check this out, and if it is in fact true, then I'd prob. select this driver even with a price premium. Effectivly there is nothing to suggest any sort of problem with this driver from what I'm seeing until at least 1.8 kHz. (..and the build quality is extraordinary for this series.) It could however be 50 to 100 US more than you were willing to pay however.

Breaking the bank..

I'd personally go in another direction here..

I purchase the Supravox 165 GMF and front horn load it (wave-guide) all the way down to 200 Hz to obtain the exact dispersion patteren I wanted. (..prob. 40 degrees vertical and 90 degrees horizontal.) Note that I would still operate the driver as an open baffle (..soley to keep the driver from non-linear compression and reflections).. and again, I would still crossover around 250 Hz LR 4th order. The driver's impulse response is absurdly clean for initial decay - and this translates into EXTREME subjective detail.

http://www.supravox.fr/anglais/haut_parleurs/165_GMF.htm

The problem here is that I don't know of a good wave-guide for a 6.5" driver. So you would have not only the extra expense of the drivers as well as the shipping, but also the time and material of building and experimenting with waveguides. NOT and easy course to "plow" (..on the other hand I would say that with respect to making your own line-source true ribbon driver )

EDIT: Apparently DDS also does custom horns.. then it may only be a "small" matter of cash.

http://www.ddshorns.com/catalog.php?page=products

Paul W said:

Tech,
Do you have a link (other than Eighteensound) for comparisons of AI and shorting rings? Why do you regret buying the AI driver? Only price or is there something else I didn't catch?

Click to expand...

Theres a thread here: http://www.diyaudio.com/forums/showthread.php?s=&threadid=62247&highlight=

Although there really very few material on this as no other drivers have this technology.


I regret buying it after reading about studies concluding that non-linear distortion plays a small role in our perception of sound (obvious exception is when abusing the driver).

If nonlinear distortion is such a small factor, I should have bought a cheaper alternative.

tech.knockout said:

Theres a thread here: http://www.diyaudio.com/forums/showthread.php?s=&threadid=62247&highlight=

Although there really very few material on this as no other drivers have this technology.


I regret buying it after reading about studies concluding that non-linear distortion plays a small role in our perception of sound (obvious exception is when abusing the driver).

If nonlinear distortion is such a small factor, I should have bought a cheaper alternative.

Click to expand...

I'd prob. be the first on this forum to say that THD as non-linear distortion doesn't matter much unless very high order products are elevated in level.

However..

You are only getting part of the information..

IMD is a form of non-linear distortion. It is also VERY audible at even fairly low levels. In particular - it frequently occurs at the upper freq. limit of fairly large diameter loudspeakers. Moreover inductance is prob. THE contributor to IM distortion at higher freq.s..

So cheer-up.. You purchased a *REALLY* good driver and its technical prowess will almost certainly improve the subjective quality sound if you operate it anywhere near its upper freq. limit.. So its likely you shouldn't have purchased a cheaper driver IF you wanted such a large eff. driver as a midrange. (..note though it also depends on how significant the operating passband is.)

ScottG said:

I'd prob. be the first on this forum to say that THD as non-linear distortion doesn't matter much unless very high order products are elevated in level.

However..

You are only getting part of the information..

IMD is a form of non-linear distortion. It is also VERY audible at even fairly low levels. In particular - it frequently occurs at the upper freq. limit of fairly large diameter loudspeakers. Moreover inductance is prob. THE contributor to IM distortion at higher freq.s..

So cheer-up.. You purchased a *REALLY* good driver and its technical prowess will almost certainly improve the subjective quality sound if you operate it anywhere near its upper freq. limit.. So its likely you shouldn't have purchased a cheaper driver IF you wanted such a large eff. driver as a midrange.

Click to expand...

I also got this impression as people talk about supravoxes, eminence beta 8s, paudio coaxes, very old 15" altecs, etc. These drivers from what I can tell use conventional motor structures. The common denominator seems to be low mass, high efficiency, and lack of cone breakup.
It seems these three matters; instead of buying a midrange driver that must have the bells and whistles in the motor which was my main priority at the time.

tech.knockout said:

I also got this impression as people talk about supravoxes, eminence beta 8s, paudio coaxes, very old 15" altecs, etc. These drivers from what I can tell use conventional motor structures. The common denominator seems to be low mass, high efficiency, and lack of cone breakup.
It seems these three matters; instead of buying a midrange driver that must have the bells and whistles in the motor which was my main priority at the time.

Click to expand...

To a large degree - however, it also limits very high output though. i.e. most of these motors can't take the heat from lots of current and won't work as well substantially beyond their reference sensetivity.

So while I advocated a supravox 165GMF as a driver - practically speaking it shouldn't be "seeing" much more than about 32 watts. On the otherhand at "lower" spl.s (i.e. +3-6 db above their reference) its a non-issue. (..also factor-in about 5 db of gain for the waveguide I've spec'ed - that would place the driver around 100 db eff. 1watt/1 meter) Doing the "math" indicates that 100 db + 4-5 times the gain puts the output max around 112-115 with a waveguide. Thats still extreme spl's in my "book".

(note.. I should have mentioned that while IM distortion is a significant factor at higher freq.s - its often the lower freq. inductance value, (..not the inductance at higher freq.s), coupled with even a low spl lower freq. "tone" and a higher freq. tone that generates the problem.)

The more I think about the 18sound active impeadance drivers - the more I begin to suspect that the 12 inch driver's inductance value is a misprint by 10 times the value. (i.e. it should be .09 instead of .9)

Click to expand...

I entered the 12NDA520 data into SL's handy spreadsheets. If Le was 0.91 mH, we'd expect 7.3 OHms at 1kHz and 59 Ohms at 10kHz. At 0.091 mH we'd expect 5.2 and 7.8 Ohms for 1 and 10 kHz, which is much closer to the data sheet. In fact the data sheet's graph shows even lower reactance than that.

If you are doing more measurements, a full 360 degrees would be quite interesting.

Click to expand...

I'd love to comply... but I have to find a cleaner setup to get truly trustworthy data. Indoors and without suspending the speaker, any measurement beyond 32" / 80 cm or so, becomes increasingly haggard due to room effects and especially, floor bounce. But 80 cm is too small to fully account for the baffle effects - at 80 cm front-to-mic distance the rear wave travels a full 80 cm farther than the front wave, which means it is 6 dB down.

Just for info, what are the prices like? In Singapore there is one distributor but on the web they only list one model (the 10" AIC), and w/o prices or further info.