Tetsuo111

Hi @ClaudeJ1, I strongly agree that tapped horns sound better the lower the crossover. I can hear the nastiness coming from the upper-end. I don't think it makes TH's a "bad speaker", simply that they need to be implemented intelligently and within their performance envelope. Sadly, 80 Hz is a difficult target to hit (for me), while also satisfying my physical size limits and upper X-over plans (~ 400 - 500 Hz).

Gents - First, regrets for a long-delayed reply. It's simply been super-busy around here lately, and there's never enough quiet time to focus on horns after satisfying other time commitments. Second, a HUGE THANK YOU to JWC and Seti for your support. JWC's dual driver horns were the inspiration for this project, mostly for his results, but even more for a logical, well-documented thought-process. Third, I moved the thread to a different forum, assuming (rightly or wrongly) my DIY project might be a distraction in this Klipsch-centric forum. For those interested in following, here you go: https://www.diyaudio.com/forums/multi-way/336290-upper-bass-tractrix-flh.html#post5779878

Thanks, Moray. I've communicated with John in the past. I need to pay him a visit sometime!

Hi, This is my first post, although I've been a long-time follower and longer-time Klipsch user. A bit about myself: I've been listening to horns for awhile. My first "high-end" system was Klipsch Heresy powered by Counterpoint tubes and sourced with vinyl (back in the day). I took a detour with non-horn loudspeakers for a few years, but returned to horns with Dr. Edgar's Edgarhorns, and now listen to a multi-way horn system of my design. I'm moving forward on a dual-driver, tractrix FLH upper-bass horn, with these design goals (but I'm still uncertain if all are achievable): 1) FLH architecture for a clear upper end 2) BW: 100 - 600 Hz (2.5 octaves) 3) Directivity: 90 degree BW at 500 Hz minimum, and as high as possible 4) SQ focus on transition from upper-bass to midrange, rather than ultimate low-frequency extension due to: - Human hearing is more sensitive in the midrange than low-frequency - Deliver dynamic, visceral "thwack" over upper-bass extension 5) Tractrix curve - to integrate well with the midrange 6) Crossover ~ 500 Hz to 1) Edgar 350 Hz round tractrix / B&C DCM50, and later to 2) K-402 7) Crossover ~ 90 Hz to stereo Danley SPUDs 😎 Mouth of sufficient size to minimize ripple 9) Length balanced for extension and minimal group delay Assumptions: - Half-space - Quarter-sized - Dual drivers to minimize horn length and driver excursion - Fits through doorways (~ 31" max) One source of inspiration for me has been @jwc bass horn projects. I find his projects well-considered and clearly articulated. Also, @setifor documenting so much of interest to horn-lovers like me. Thanks, @jwc and @seti! First, I selected B&C 10" 10MD26 woofers because I have some familiarity with them from past builds. They are very sensitive, with lightweight paper cones, and strong motors. I chose 10" drivers (rather than 12" or 15") to integrate with the midrange. https://www.bcspeakers.com/en/products/lf-driver/10-0/8/10md26 Using Edgar's and PWK / Delgado AES papers as a guide, the following parameters are calculated: MASS ROLL-OFF Fhm = 2 * Fs / Qes = 2 * 76 / 0.22 = 724 Hz THROAT AREA St = ( 2 * Pi * Fs * Qes * Vas ) / c = ( 2 * Pi * 76 * 0.22 * 0.20 ) / 343 = 58.47 cm^2 BACK CHAMBER VOLUME Vb = Vas / [ ( Fc / Fs * QEs ) ] -1 = 20 / [ 90 / ( 76 * 0.22 ) - 1 ] = 4.3 L REACTANCE ANNULLING CHECK Fc = Qes * Fs * ( 1 + Vas / Vb ) = 0.22 * 76 * ( 1 + 20 / 4.3) = 90.19 Hz Finally, to the reason for my post....! My experience is mostly running Hornresp. Since I need data in AutoCAD compatible format for the build (as well as curiosity about cross-checking David's Hornresp with Edgar's / Keele's maths) I chose to model on Volvotreter's excel tool download. I'm running into an obstacle though, because I haven't figured out how to model a dual driver design on Volvotreter's tractrix calculator. ***Can anyone share guidance on this? Also, Hornresp uses a forcing function to calculate horn length, given inputs of throat area, mouth area, and Fc. But the result seems to differ from Volvotreter's worksheet. Usually, Hornresp results in a longer path length. Naturally, I want to minimize bulk while still attaining my design goals. I think JWC's full-sized horn was 61 cm deep. So perhaps I'm guilty of operator error because I get a differing path length when I model his horn on Volvotreter's tool. In any case, I suspect I'm making a simple mistake and am here seeking a little oversight and adult hand-holding. Thanks for any clarification. Cheers!