Crossover will probably change because I moved the port hole positions slightly. But if you leave the holes where they were in the previous version, then the same crossover would work. The new hole positions may or may not be an improvement (but I do have them further into the corners of the horn, which should be better in that respect).
I don't know what the spl capabilities are.... basically whatever 4 NS6s give you in a large ported box (I crossover to a sub at about 60Hz), I'm pretty sure those will be the limiting factor, since others are horn loaded. More than loud enough for me, but I'm not into macho volume stuff! Foam does require software changes due to the HF rolloff. Haven't tried it on these, haven't really felt the need. I had it on the SEOS15s and it maybe helped, hard for me to tell. For looks and WAF, though, the foam is a definite negative.
Hey Bill. How is the midbass punch in theese btw? the four NS6 do have about the same surface area as a 12" driver, is the punch comparable to the Seos 12" kits?
Punch is in good supply. Of course they are in a big ported box unlike most SEOS builds so they go much deeper. Drivers don't have high end 12" type excursion though so probably wont go as loud as a pro 12. Not a. issue unless you plan to start a club or are into huge spl.
Yeah I agree. I keep toying with the idea of buying a Dunlavy SC-IV, but most of the 'hifi' phase-coherent speakers have poorer directivity and lower headroom. As far as where I live, I work at home and own a bunch of homes in the Pacific Northwest, so I tend to bounce around between Oregon and Washington. For the most part I'm in the Seattle area.
Speaking of phase coherent speakers- Does anyone know how these guys are doing it? http://www.psiaudio.com/technology/cpr probably the same as how this works? http://www.thuneau.com/arbitrator.htm Apparently the PSI's are all analog (so people say, maybe not under the hood). Was curious if this is something that could also be accomplished in NI's reaktor, some sort of all pass filter? Sorry for the sidetrack...
Sorry I don't know much about that company, but the first set of graphs there already makes me wary. They show a 'regular' speaker with all sorts of phase wraps, and then their wizbang speaker with nice flat phase. Trouble is they removed all the fixed delay in the plot for their speaker but leave it in for the regular one (which wouldn't look as bad if it were shown fairly). The phase plot they show for theirs is quite good, though, so perhaps its only their marketing department that is shady. I bet that's for only one vertical angle, though.
Hi bwaslo, This is looks like a real winner, one great thread. A few questions have been rattling around my mind: Is there any design guide for the angle and length of the secondary expansion (mouth flare), or, how did you arrive at this one? In Danley Synergy and Unity horns going back to the ServoDrive TD1 there seems to be a much shorter secondary flare, and much less of an angle change. And two questions regarding the spreadsheet: 1. When using the same angle (e.g.: B2 and B3 both @ 60) for horizontal and vertical coverage in the spreadsheet, shouldn't E2(w) be approximately E2(H)+1 (@ a thickness of B9-0.5)? 2. The input value for L12 does not seem to take the path length inside the compression driver into account (e.g.: about 7.2cm for the BMS 4550). Shouldn't a nominal value be added as the midrange drivers interact with the horn all the way to the inside end of the compression drivers throat horn? This would also reduce the value of S1 (again a nominal value for a 1" compression driver could be assumed, e.g.: 2.5cm^2). Thanks again for posting this nice example of a DIY Synergy horn. Regards,
These filters only account for phase issues at a single point or a small range of points. You can't fix a physical issue like pathlength differences with a filter. The Synergy design achieves this in the physical realm (of course the crossover itself can create phase issues so that must be handled appropriately too). Those DSP filters aren't necessarily bad, but they don't address the issue as well as the Synergy design. They could be used in conjunction with the Synergy design though. I played with Thuneau products a while back and it was just not worth the hassle for me, but it does work as advertised...you just have to understand the inherent limitations. There are pretty advanced VST crossovers out there but I haven't played with them much. The digital console world is something I have little experience with.
TB46, The formulas used are from Keele, see his "What's so Sacred about Exponential Horns" paper. You can adjust ratw to make it smaller if you want. The very wide flare angle also gets into it. Also notice that those aren't the actual dimensions, they are just for drawing guide lines before cutting. Actual widths there are C3(h) and C3(w), the corners are cutoff from the E dimensions. These will all be changed, by the way, in the next version of the spreadsheet. That would probably be more accurate for HornResponse modeling (but don't do it for cutting wood!). I imagine it varies for different drivers, though. Here's a shot of the latest version of the waveguide, CoSyne II, mounted in a cabinet (same one as used before). Made from oak veneered plywood (so cuts had to be as accurate as I could make them).
Hi bwaslo, Thanks for the additional information. I like CoSyne II, makes me wonder how you are connecting the drivers to that horn. Regards,
The HF CD is mounted to a plate made from thin plywood glued to the horn throat. The midranges and woofers are mounted with short woodscrews over the port holes (also cleared out is room for the rubber surrounds of those drivers so they don't run into wood when moving).
They look good Bill! How much better do they measure with the ports right in the corners compared to along the sides.
Hello everybody, just found this thread and I'm very interested. I did build the original Unity horns as sold by Nick at Lambda, I think I was one of the first to build a system and I looooooove it. But the itch to build something that might be even better is always there. I'm in Portland, OR. Happy building horns New Year to everyone !!! Alex
mtg90-- I haven't measured the on-axis/off-axis differences yet. But in an unfortunate way, they're a little worse- the little 2" mid driver ports turn out to be farther from the tweeter, which causes both a drop in the high end (from the quarter wave notch ) as well as a second peak above the notch - I didn't figure the port entry point right. it appears. It can still work, I even have a mostly linear phase crossover designed for it, but the crossover is more complicated than for the black horns and doesn't model quite as well for phase (though probably a little better in magnitude flatness). For now if doesn't much matter for me since I'm using a minidsp crossover, but I don't have an active crossover figured for the new horn. I also don't have the parts to make the new passive one, so I'm back to listening to the black horns. Not sure how to proceed just now - buy all the parts for the new crossover, work up an active one (not so easy) or try to machine or move the existing ports (even less easy ).
Alex - are you building a horn this season? My wife and I plan to retire to Portland in a few years, nice place.
I don't know how soon I could start to build the horns; my immediate project is to build something (and I have parts for a line array) for the home theater (5 + 1) setup that I badly need. Plus I have a ton of other house project that my wife really wants to be done before I start any other project... : Portland is great, been living here since '85... Alex
Bill, Quick question, are Q1 and Q2 the miter angles for pieces 1 and 3 in your current spreadsheet revision? It appears that way, just want to verify.... Scott (P.S. I did get Praxis re-calibrated and all is working well. )
Hi Scott, I'm not sure what that angle is called. It's the angle on the edge of the board (the edge there isn't perpendicular to the large plane, nor are other edges on those two boards.). I actually have a (partially complete) revised spreadsheet I'd be happy to send you (send me your email address). The new plan is much easier to assemble and avoids any of the "perpendicular sled" cuts. I hope to do a full "how to" with photographs of my last build when time permits. Bill
Thanks for sharing fantastic project, Bill Looks like in the original SH, the woofers are arranged as simple reflex with port tuning around 80ish hz. From the internal drawing on DSL website I suspect Tom uses Faital 12HP1020 on some of SH models, which has high Qms/Qes ratio. Does it have any relation to synergy concept? can the woofers arrangement above be simulated in Hornresp?
Bill, I'll PM you my email address. I've gotten permission from the "Sergeant at Arms" or "She Who Must Be Obeyed" to build a set of horns for our home theater fronts. <squeee...SO CAN'T BELIEVE SHE OKAYED THAT!...squeee> I'm still not sure how that happened. It'll all be slowed down by the fact that in July we started fostering two little girls, ages 4 and 7. Our world has exploded in pink and perfume. Sigh. I'm more than willing to help with photo's and jig ideas for the build process. I've done quite a bit of furniture and cabinetry, and I'm more than willing to post my ideas in this thread. It'll just be slower than I want due to the fact that I'll have to be the U.N. to two little girls fighting over Barbies and Elmo dolls. I've got another question regarding the existing spreadsheet, I'm assuming you're using S1 as the entry point for a 1" diameter compression driver. A 1" circle has an area of 1.27*1.27*3.14 or 5.06 cm^2. Did I miss something, because the 2.53 reported in V3 is the area of a circle with a diameter of .707 or the h x w of the square formed by the end of the compound miter pyramid. Is that right? It makes very little difference in the rolloff of the horn, though it does possibly ever so slightly change the optimum chamber volume/port area in a hornresponse simulation. In practice I don't think it'll make one bit of difference because it's nearly impossible to get *exactly* the right chamber volume. Scott
Any new thoughts on the viability of a two-way based on the NS-6? What port/driver placement issues would be encountered by just using larger drivers? I've got the mini-dsp and amps, looking for a fun learning experiment. Thanks!
I put a D360 in with 4 NS6es, same port locations. It can work (though not linear phase - not enough overlap. I doubt that really matters though).
Is that the same port location as the original NS-6 placement -Or- using the NS-6 with placement port location closer to the gento's location? I'm too much of an audio newb when it comes to understanding the impact of linear phase, I really want to experiment without totally screwing up early version of the experiment. After all, I just made my first horn without losing any fingers! Would this issue be resolved by utilizing something smaller in a two-way? Say the Vifa TC9FD-18-08 http://www.parts-express.com/pe/showdetl.cfm?partnumber=264-1062
Re: Effect of large woofer port holes on waveguide Bill, You may want to look at the old Marshall Buck patents, US4,283,606 and US4,619,432 where he uses a perforated metal sheet to allow the low frequencies from a woofer behind a horn/waveguide to leak through while providing an undisturbed waveguide surface for the high frequencies. (See Figures 8, 9, and 12) He optimizes the perforated metal as a low pass filter for the woofer in some embodiments. He does not provide much specificity relative to the optimal porosity to minimize impact on the tweeter. He suggests approximately 35% open area, but he is crossing over at ~2.5 kHz and he is attempting to balance the lowpass characteristics desired for the woofer with the high frequency integrity of the waveguide. I have sent him an email asking about the parameters for the types of applications we are considering. Anyhow, I think you may find it to be an interesting possibility for covering your open slots in the Synergy horn constructions. I can't attach the patent, because it is 2MB and the stated limit for attachments here is 194kB total attachment. So, I'll attach the pertinent drawings, and hopefully, you can get the patent if you want to see the rest, or I can email it to you if you like. All the best, - James