Hi everyone, I'm currently working on a speaker project for my home theater and am REALLY interested in the Denovo DNA-360 driver and SEOS-12 waveguide. I was originally planning (and have actually completed the crossover design for it) on using the B&C DE10 CD and ME10 horn combination, however after reading about the SEOS and Denovo am SERIOUSLY, SERIOUSLY reconsidering. My only problem is, I need the frequency response and impedance graphs (or data files, really in any format, I can convert from pretty much anything to what I need) to plug into Passive Crossover Designer. I see the FR graph, but it doesn't specify what waveguide it's on, and I've had no luck finding an impedance chart. I'm really looking for the DNA-360 mounted to the SEOS-12 graphs SPECIFICALLY, otherwise PCD really won't be all that accurate. Can anyone help me out? FYI for anyone interested, it's a bit of an odd project... Instead of going with a single, large woofer, I'm using 4-6" woofers (Aura NS6-255-8As). I'd say my sensitivity is about 94dB/1W on average, a little higher on the lower end (about 96-97dB at <300Hz) because of my enclosure design; I'm using a ported box tuned to 65Hz, high-passed at 60Hz, with the big baffle they're mounted on, it should take care of the BSC without having to low pass a pair of them and lose some midbass. With the DE10+ME10 combo, I was using a 3rd order electrical HPF at 2000Hz (the acoustical rolloff is close to a 4th order, though), any recommendations or proven crossover designs for the Denovo? Thanks! Jason
Jason, the Frd and Zma files for the seos12 and 350/360/DE250 should be contained in http://libinst.com/SEOS/SEOS12+Designer10/Xover%208.zip
The DNA-360's crossover is quite a bit more complex than what I designed for my DE10+ME10 combination, so I may have some questions after I sit down and study it... I have a strong electrical background, but I discovered early on in my crossover design that my textbook filter knowledge was a lot more useful getting A's on tests than actual, real life implementation, and the only acoustical know-how I have is self-taught. It's gonna take me a minute to try and understand what the purpose of the additional circuitry is (specifically the dual parallel RLC networks, the series resistor in the inductance leg of the filter, and the 3rd capacitor for the filter... Dunno if I'll have time today tho). With my baffle design (well, lack of electrical BSC), the lower I can cross the woofers to the tweeter, the better off I'd be, I believe (like I said, I'm really pretty green when it comes to technical acoustic aspects). A 1000Hz crossover point would put the horizontally-positioned woofers and all the woofer to CD C-C spacing at less than 1/2 a wavelength a part, and even the vertical woofers criss-crossed would be within 1 wavelength. In case anyone was interested, here's my work with the 4 Aura NS6's and the DE10+ME10: Here's the FR... I'm very happy with it; the crossover point is 2000Hz: This is as good as I was able to get the impedance. Most important was staying above 6 ohms, I have a lower model (Harman Kardon AVR-247, it's 55W/channel at 8 ohms is the main driving force in a high-sensitivity speaker) receiver that'll run a 6 ohm speaker but I didn't want to go any lower: The crossover schematic. I obviously wanted to have a SINGLE crossover for all 4 woofers, but I couldn't figure out a good way in PCD to do that, other than calling 2 woofers and 2 midranges and designing identical circuits. I was going to convert the values to make a single crossover, but I'm not so sure anymore because I don't have really any way to judge if it'll negatively effect performance. The 2nd order Butterworth I used on the woofers resulted in a near-exact 3rd order Butterworth curve, and the 3rd order Butterworth on the tweeter yielded a near-perfect 4th order, although it rolls off even steeper below about 1.75kHz. Like I said, I don't have rhyme or reason for some of the things I did, I just messed around and figured out what changes resulted in what outcomes, and this was a combo that worked well: The baffle, drawn to scale. The woofer horizontal C-C spacing and the woofer to CD C-C spacing is less than a 2000Hz wavelength, and really everything else is as good as I could get it. I modeled it after several theater speakers, most notably the QSC SR-46 cinema surround. I'm going to make the baffle out of fiberglass so I can rear-mount the woofers and have the surround baffle edge contour down to the surround. The Auras have an ugly paper gasket stuck to the front, but other than that it's not bad looking, just a plain old paper cone.
It looks like your simulation uses manufacturer traces??? That might be why it seems more complex. Likely those don't measure better.
Sorry, get the file http://www.libinst.com/SEOS/SEOS12+Eminence%20pro%20woofers/SEOS12%20DeltaLite2512.zip. Unzip that and there's a file named "SEOS12 White Driver.zma" (or something like that). The "white" refers to a color that Erich had some of the first samples painted. This was a 360, but the 350 is basically identical (as is the DE250) except for mounting bolt types (and apparently unit-unit consistency).
Hi Bill, I just noticed you posted the new zip file, I've been with working with the original Designer 10 file. I took a look at the newer posted file, and the data appears slightly different... Were these two individually tested units (I saw you mention the Deltalite file uses a 360), if so, were they the same model and just different units, or were they different models? The sensitivity in the newer-posted files, IIRC, was notably lower than what was in the Designer file. Sorry for the poor descriptions... I'm trying to peck this out on my phone. I can get screen captures if you want to see what I'm talking about in the morning. Thanks, Jason
Here's what I'm talking about, some visual aides: These are the tweeter files from the DeltaLite .zip: And these are from the Designer 10 .zip: Just curious as to which would be the more-accurate file to use to work with the crossover for a DNA-350...
The measurements were made on different days, probably with different drivers. Maybe at different angles, too (some I measured on-axis, later ones I did at about 22deg off horizontal). BTW, you can't just grab any tweeter response files and use them with any woofer files unless drive levels, mic positions, and distances were controlled when measured, you have to take them as pairs -- only use the woofer file with the tweeter file measured in the same setup, so the drive levels and mic setup were the same. When I do crossovers, I don't go to effort to use any particular drive level (it often depends on who is around that I might annoy when doing it!), so don't use my tweeter response files except with the woofer response file from the same set! If I were more organized, I'd set and control levels overall, but these were done relatively casually, I didn't set out to make a driver library! Impedance files, though, shouldn't matter, since they don't vary with drive level or mic distance or all that other stuff. That said, I can't explain the diff in the impedance files. The responses (both FRD and ZMA) do look a bit off compared to the ones from the Deltalite folder.
Understood about the levels. I never really considered that, I digitized the FR and Z plots for the Aura woofer that were on the cut sheet, then ran them through some the FRD tools to get phase and baffle step effects prior to dumping them into PCD. The impedance curves were so close that I figured that it could just be differences in individual units, and for that matter the FR is fairly close curve-wise, the Deltalite version is just a solid 5dB
What the hell?! My computer decided to quit working half way through my message.... Anyway... Where was I? The DeltaLite version is just a solid 5dB down from the Designer version, so I would imagine you used different drive levels. I don't have the equipment to test drivers (no software, and the only mic I have is the crumby little omni that came with my HK AVR for the auto EQ, so certainly not the best), seems like I may be SOL... The Aura cut sheet states it was taken from 1m, but guess that doesn't matter since the SEOS files are required to mate to their respective woofer files to be truely accurate... I'm not sure where to go from here, maybe shoulda figured this out prior to actually dropping the $$$ on 3 of em. :-[ The only thing I can think of, without testing abilities, is to use one of your files for curve shaping and use an adjustable L-pad for level matching, but my attenuating resistors also provide some response tweeking as well. Damn.
You might be able to get measurements close enough to avoid nulls around crossover using your mic and freeware REW. Use the crossover sections from existing designs for the SEOS12 and what you've come up with on the Auras. Set them playing and do some trial and error, inverting tweeter if necessary -- chances of success the trial-and-error go up if you put the SEOS in a separate cabinet (or no cabinet) so you can adjust its distance relative to the woofers in real time. Don't expect the mic to measure flat at frequency extremes, but it should be good enough to tell if the drivers are happy near crossover, and you can assume the SEOS behaves above there.
Probably a noob question but, is there any way to rescale the frd file for the SEOS 12 to be compatible with 1W/1M level frd files from other woofers? . I have DE250's and want to try and model an XO for some other woofers. Thanks, Foggy