Modeling Can Surprise You.

I haven't heard anything about a weak lowend on the 15A. But yea, graphically, it looks pretty darn week.

 

Of course, WinISD isn't the end all...

 

Though I haven't modeled up all ED's woofers, I have modeled the 15A, for obvious reasons , a million times. My 1.73ish sealed box has a F3 of 67ish hz BUT, in the tiny cabin of my truck, I end up with a beautifully flat-ish repsonse with more spl than I need. I have a ported box 98% complete (need to do the cutouts [dual baffled] and seal up everything) that will be ~2.75/32.5 which gives better output, obviously, at the expense of a hair of bloat with regards to SQ. Heh, just under a cube for the .707 enclosure, and a F3 of about 65-66, if you decided to go that way. It has little problem dropping low. 2.5 cuft sealed actually gives you a ~2 dB boost down low over most of the "recommended" enclosures for this driver (and a F3 of 70...double )

BTW, Steven, I have all the new ED driver's isd files if you want me to send them to you. Save you the tedious param. entering that annoys all of us One more thing. drop isd like a bad habit. I just got BassBox Pro. That program is a million times better than isd hopes to be. God I love it and it's steeper learning curve...

:ss:

 

Makes perfect sense...
The BL is huuuuge on the 15a. 30, I believe.

I do like the dual spaced spider on the A series, but I don't see the need.. unless it's because that piston cone is so massive, with such a distant center of gravity relative to the voice coil, that there was a high potential for former rocking.

With most cone subs, there isn't the need, single spider solutions work well.

 

i have a question...remember...i am the guy who buys subs...makes a few different enclosures for them...gets them metered and lets you know how they sound to my ear...i am just learning about parameters and how to get a performance idea from modeling

i know that Bl is motor strength....but why is it so high on the 15A?....i haven't seen ot that high unless it was a spl designed sub....what does this mean and what does it do or not do for performance?

 

Your observation is correct... in fact the DD 9515 only has a BL that reaches just below 25!
The BL of the 15A is huge, absolutely.
But also consider that when you measure BL on dual voice coil subs, if you measure it with the voice coils in parallel, BL will be half of what it is if you measure it with the voice coils in series. Interesting? B)
I don't think it's unreasonable to assume that the BL was measured with VC's in parallel here (or else this sucker's got a BL of 60 in series! :lol: ).

Usually, with traditional motors (non XBL2 I mean) you trade off breadth of the BL curve (how broad it is, becoming less and less parabolic, more and more flat) for high peak BL numbers (higher and higher BL, but resulting narrowing and peakiness of the BL curve shape)...
You pick, do you want more Xmag and the accompanying lower BL, or higher BL, and the accompanying less Xmag.
That's oversimplification, but that's a typical design trade off..

Think of it in these terms:
You control the length of the voice coil relative to the thickness of the top plate (the magnetic gap that the voice coil is suspended in).

If there was a 6:1 ratio between them (VC was 6x longer than the top plate was tall), then the voice coil could move quite a distance before the coil count in the gap changed at all (and therefore the BL was the same, for longer). But BL is lower, because at any given time, only 1/6th of the windings are actually suspended in the magnetic gap, working to push the sub in and out.

On the other hand, if you rearranged those windings so that there was only a 2:1 ratio (voice coil only twice as long as the top plate is thick), then you'd have many more coils suspended in the gap at any given time, all acting to push the sub in and out...
But the voice coil wouldn't be able to travel as far before the coil starts to actually exit the gap, reducing coil count in the gap, and therefore BL.

The width of the gap is also a factor in BL...
If you've got a cavernous distance between the top plate and the pole piece, lots of space for the voice coil to fit in that gap, there will be less magnetic flux in the gap - just like if you held two magnets an inch apart from each other... there would be less force pulling them together (or pushing them apart, depending on how you were holding them), compared to if you held them only 1/8" apart.

TC sounds has definitely established itself as a buildhouse that can produce really high BL subs, minimizing the inherent tradeoffs relative to other's designs. They've got a nifty top plate design that looks deceptively small when viewed from the outside, and apparently use quite tight tolerances and minimal gaps.
But the A motor is known to be a typical TC sounds motor, traditional, parabolic in BL shape. (That being said.. eD hasn't released DUMAX plots of any production A series subs, so how parabolic we don't know.)

As far as how BL affects performance...
It's the strength, that tells the cone to follow the signal being fed into it.. the cone position mimicks the signal.
For a pure sine wave... cone position follows the sine wave.. Starts out at the "at rest" position, moves out until you reach the peak of the sine wave - which is the point you reach peak excursion on the sub.. and as the previously steep curve slopes to that peak, cone velocity slows, until it literally stops at that max excursion point, at the peak of the sine wave, slowly accellerating back towards the "at rest" position, passing it, heading towards max excursion point in the rearward direction... when it reaches the trough of the sine wave... finally heading back to the "at rest" position again.
A complex signal is really the same thing... the cone position will mimick the position on the complex waveform.

If your BL is constant throughout your entire excursion range, the subwoofer will be brute-force controlled by the motor, behaving exactly as it should.
That's THE benefit, in terms of SQ, that the XBL2 subs really have.
More parabolic BL curve subs have ideal control over their cones when they are nearest the "at rest" position, and don't have quite as solid control at higher excursion levels.
Extremely peaky BL curve subs can create distortions due to this, as the sub passes through a point where the motor has control over the cone, then literally almost completely gives up control of the sub, letting the cone's momentum carry it out until it stops and is pulled back in simply due to the suspension yanking the cone back into a range where the motor can take control again and fling it in the other direction.

Ideally, you want a BL curve that is flat, across it's excursion range.

Ponder the servo-motor type of subwoofers... such as the PG Cyclone, or the ServoDrive BassTech7 or ContraBass.. a servo motor, no matter how far it turns, will always have a constant BL. It could turn 360 degrees and it's BL doesn't change a bit. That - and about 10x higher efficiency than a voice coil motor - are inherent benefits of servo motors. That's cool. B)

BL also affects both the type and size of enclosure that a subwoofer is regarded to be "ideal" in...
A higher BL subwoofer will work better in smaller enclosures...
But as a higher BL sub will also have a lower Qes (they roughly show the same thing, those two specs.. that is a lower Qes sub also indicates a stronger sub), and since Qes is a component of Qts, a lower Qes drives the Qts to be lower...
And since low Qts subs are typically ideal in ported enclosures, while high Qts subs are ideal in sealed enclosures, it also directly drives what application the sub is best in.

Finally, read up on "Hoffman's Iron Law"... it dictates the most fundamental design tradeoffs, describing them in terms of three mutually exclusive items (in both raw driver design, and enclosure design with respect to a given driver):
1) high efficiency
2) small enclosure size
3) low frequency extension
Some factors must be traded off to maximize others, based on your design goals.

Important.. primarily with respect to BL... because Fs tends to rise as BL rises... see what's happening? High efficiency and small enclosure size naturally increase with BL (suspension design factors remaining constant), so low frequency extension will naturally be traded off.
It's a good rule to learn and understand.

Phew!
Hope that helps.. all those words and it really doesn't do justice to how BL, and primarily curve shape affect performance...
But that's an intro anyway.

 

I missed that. Can you repeat that Geo?
:lol:




Typically, a wordy, thorough, well written repsonse. Nice job once again, Chris.

 

Take the published T/S parameters for the e15A with a grain of salt; the listed Re is 6.77 Ohms, which would be a dual 4 Ohm design. I suspect the rest of the parameters are probably not indicative of the actual products.

Anyone in the Seattle area with an e15A? Bring it by the shop and we can run it through the test system to get the real T/S parameters. We can also do a BL/Cms measurement as well, if desired.

Dan Wiggins
Adire Audio

 

THAT would be most interesting.

All this "you don't need to see the DUMax report" crap when ED loved to show off DUMax previously really leaves one to question the facts about the driver. Particularly when you make mention of the driver's RE. Seems a bit fishy.

I wish I had one here Dan, I'd fly out to Seattle tomorrow for this chance.

I'm interested in the drivers #s, can someone dump the TS for the driver here?

 

i don't know how much, if any a difference this will make...i have

qts - .33

vas - 96.47

re - 6.77

 

I believe ED used a dual 4 ohm proto to do dumax testing and used those results as "final results" for the "public"...Don't quote me, but I'm pretty positive Stephen or Ben said that after they were released.

I'm sortakinda close to WA...I was once asked by one of CAF's more respected members to send my 15A to DLC/Detroit to have it DUMAX'd. $160 for the test and two way shipping is a little steep, though.


Here's something to think about...The 12A was reviewed by T. Nousaine in the Oct/Nov issue of ME. The following are ED published specs on the 12A, and the published specs from TN's testing. Note: Both test results were arrived via DUMAX.

ME Specs on the 12A:

Fs=25
Qms= 2.91
Qes= .31
Qts= .28
Re (vc in series)= 3.6 ohms
Vas= 2.1 cu. ft.
Xmax= 23.8
Sensitivity (2.83v input measured at 1 meter)= 89

ED specs (in my isd file. I didn't enter the data, though. isd file was part of a ED package that included all '03 isd files for ED drivers)

Fs= 27.3
Qms= 4.71
Qes= .450
Qts= .41
Re= 3.61 ohms
Vas= 1.669 cu. ft.
Xmax= 26
Sensitivity= 85.33

Some numbers stick out, such as Qms, Qes, Qts, Vas, Fs, and of course xmax is over 2 mm's less on ME's dumax results. The difference in Fs and xmax isn't all that big a deal. The other specs...

 

i don't know how interested everyone will be in a couple of months...but i would be more than happy to send out my 15A in the spring...as long as somebody pays the return shipping

 

This is very interesting indeed. I wonder why people over on CAF wont pool together money and get a 15a dumaxed like they did for that DD sub that got railroaded over there when the results came in..

 

one thing that i can honestly say that pisses me off...i did'nt check the specs before i purchased the driver...i rarely do...i would of purchased the 15A regardless

but when i go to model it up in winisd and design a enclosure...i expect the parameters to be from the driver sitting on my kitchen table...not some proto-type that i will never use :ranting:

 

I'm not 100% positive on the proto being used (close, though). I did read it on the internet, so take it for what it's worth I, too, would have purchased the 15A anyways. I figure, once you get to this level of woofer, the differences between supersub A and supersub B is probably more dependent on install than the driver's actual "ability" to reproduce high output, low distortion bass. I actually planned on placing my order for the Brahma the next day through Seth (search CAF, I asked him, RThompson, and teal a million questions regarding this driver). Money was the deciding factor for me.

 

Ahhhh, I was just posting up the ME Dumax results (in which Nousaine said that Stephen Ponte had nothing to do with) when I saw that luvd already had...

I'm not really gonna speculate on this one because it just leads to incessant bashing by ignorant people as well as people taking the speculations as truth. I'll leave my decision up to a DUMAX report (since that's the system that the original model was measured with).