
Home  Forums  Rules  Articles  diyAudio Store  Blogs  Gallery  Wiki  Register  Donations  FAQ  Calendar  Search  Today's Posts  Mark Forums Read  Search 
MultiWay Conventional loudspeakers with crossovers 

Please consider donating to help us continue to serve you.
Ads on/off / Custom Title / More PMs / More album space / Advanced printing & mass image saving 

Thread Tools  Search this Thread 
10th June 2019, 12:58 PM  #451 
diyAudio Member
Join Date: Sep 2008

I proposed in another thread using the Dayton PT Mini in a vertical line along with the TC9 specifically to address the off axis response.
https://www.partsexpress.com/pedocs...ifications.pdf 
10th June 2019, 01:26 PM  #452  
diyAudio Member
Join Date: Sep 2008

Quote:
My listening experiences with Magnepan speakers is they do this very very well. A very recent listening of the 1.7i was a revealing experience and what’s possible with their new LRS is astonishing for the $$$. Older posts, I know......but worth addressing attempts to combat horizontal directivity in nearfield arrays. Would an array of open baffle or dipole TC9s accomplish this?.....the TC already has a very high QTS making it somewhat suitable and the new SBAcoustics 3” has an even higher QTS. 

10th June 2019, 01:35 PM  #453 
diyAudio Member

Tend to agree with nc535 that out of box its hard to get those nice smooth BMR factory curves, in red trace below is TEBM46C20N4B factory curve verse my diy 192kHz rate session on a relative big dual cardboard baffle of size around 100120 centimeter times 120140 centimeter, green trace is 1/24 smoothing and black trace is FDW (6 width in cycles), zooming into DATS measurement we can see the two blib where response is worst. That said driver here was simply front mounted so who knows if one works with that weird frame figure to get it implemented smooth and flush into enclosure help with a smoother curve.

11th June 2019, 11:30 AM  #454 
diyAudio Member
Join Date: Mar 2013

Thanks for the kind words
I'm unsure whether the phase response offaxis is identical to ditto onaxis. Is it that simple? So that the only thing needed to know phase between two identical drivers is the difference in distance? I made a new version of the excelsheet with this assumption to experiment with vertical directivity of each driver and to see how this affects the combined response. Directivity modelling Directivity is modeled by a frequency dependent response as function of offaxis angle. So on axis the frequency response is typically 0 dB from 10020000 Hz. The at increasing angles the response decreases. You can either make your own curves in the Directivity tab (090 deg) or use the included assumption that the curves can be described using the formula: on_axis_responseseparation*(frequency/max_frequency)^slope*off_axis_angle with this formula you can get  constant directivity by setting slope = 0  something resembling normal tweeter behavior by setting slope = ~1 and appropriate separation (0.5?) Input it  separation, which sets the separation between lines at 20 kHz multiplied by angle (so separation = 1 gives 45 dB at 45 deg).  slope, which sets shape of curve Onaxis response I also added a simple way to choose an onaxis response to counteract the downward slope with the formula (frequency/max_frequency)^slope_0deg*elev_20 where slope_0deg sets the shape of the curve and elev_20 sets hos much output is increased at 20 kHz. Note that increasing onaxis output also increases offaxis output. Examples With FTCarray of 42 drivers at 4 m and parameters slope = 2 and separation = 0.5: directivity 0.5db_deg, 4m, 42 drivers, dirplot.png directivity 0.5db_deg, 4m, 42 drivers.png No huge difference to monopole behavior. What if we increase directivity (waveguide?) and set parameters slope = 1 and separation = 1: directivity 1db_deg, 4m, 42 drivers, dirplot.png directivity 1db_deg, 4m, 42 drivers.png Looks a little better, more even response above 7 kHz (less to fix with DSP). What if we add that output increases with frequency by setting slope_0deg =0.6 and elev_20 = 12 dB: directivity 1db_deg, 4m, 42 drivers, sloped, dirplot.png Note that all curves are modified. directivity 1db_deg, 4m, 42 drivers, sloped.png Not bad! (Note also that the "Without directivity"curve is unaffected). Let's take it to the extreme, what if we magically can make an array where the directivity only allows positive summing. So that when the offaxis angle causes negative summation (phase angle of more than 90 deg) the output drops to 100 dB: only pos sum, 4m, 42 drivers, dirplot.png only pos sum, 4m, 42 drivers.png Increased output all the way and a lot more even frequency response Not possible to physically make though File The spreadsheet is attached, just rename to *.zip and unpack. /Anton 
13th June 2019, 07:05 AM  #455 
diyAudio Member
Join Date: Mar 2013

Error for "bottom driver" in last excelsheet
I found an error in the excelsheet in the previous post. It only affected the "bottom driver" so long array behavior was almost unaffected. But when I tried simulating a single driver the error became apparent. Attached is an updated version.
I was experimenting with constant directivity vs traditional speakers. A highly directional constant directivity speaker seems to behave well: 1 driver, 4 m, constant directivity, dirplot.png 1 driver, 4 m, constant directivity.png Compared to a traditional (nondirectional up to a few kHz and then narrowing): 1 driver, 4 m, traditional.png Might be quite difficult to achieve that level of directivity though. /Anton 
13th June 2019, 05:10 PM  #456 
diyAudio Member
Join Date: Apr 2011

Hi Onni:
I love this spreadsheet. Its very timely as I'm building/experimenting with line arrays now. You might find it interesting to compare your results with those produced in Vituix, which can also model line arrays taking driver directivity into account, although with fewer maximum drivers. A suggestion re' directivity: it would be "nice" to just enter a piston diameter and then have you compute the directivity roll off from that. Thanks for doing this. Jack
__________________
My Synergy Corner Horns and Bass Bins http://www.diyaudio.com/forums/multi...bassbins.html 
18th June 2019, 01:29 PM  #457  
diyAudio Member
Join Date: Mar 2013

Quote:
Good suggestion, not sure how to do it though... Is there an easy formula? /Anton 

18th June 2019, 04:49 PM  #458 
diyAudio Member
Join Date: Apr 2011

__________________
My Synergy Corner Horns and Bass Bins http://www.diyaudio.com/forums/multi...bassbins.html Last edited by nc535; 18th June 2019 at 05:12 PM. 
18th June 2019, 06:37 PM  #459  
diyAudio Member

Quote:


18th June 2019, 07:42 PM  #460 
diyAudio Member
Join Date: Jun 2005
Location: Fort Worth, Texas

The directivity equation for a circular piston uses Bessel functions, which fortunately Excel has as a built in intrinsic function.
In the attached screen grab, you can see the required equation in the Excel formula bar. Note that the polarity inverts at each offaxis lobe as you go up in frequency The spreadsheet is also attached in *.zip file for those who happen to have Excel and want to play around. The “ka” parameter is the ratio of the piston circumference to the wavelength. ka = (2*pi*f/c) * (d/2) = (pi*d) / (c/f) where: f = frequency (hz) c = speed of sound = 343 (m/s) d = diameter of piston (m) 
Thread Tools  Search this Thread 


Similar Threads  
Thread  Thread Starter  Forum  Replies  Last Post 
Short Line Array (line source) build  perceval  Full Range  226  4th January 2015 02:08 PM 
Achieving Optimum Line Array Performance Through Predictive Analysis  Patrick Bateman  MultiWay  2  27th June 2014 11:45 AM 
Shunyata DTCD power line analysis  jackinnj  Power Supplies  1  16th December 2011 12:30 AM 
Dynamic load line analysis using music and speaker load  Michael Koster  Tubes / Valves  0  7th March 2008 08:47 PM 
line source diffraction  david yost  MultiWay  8  5th June 2005 02:01 PM 
New To Site?  Need Help? 