
Home  Forums  Rules  Articles  The diyAudio Store  Gallery  Blogs  Register  Donations  FAQ  Calendar  Search  Today's Posts  Mark Forums Read  Search 

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 
5th May 2010, 09:19 AM  #1101  
diyAudio Member
Join Date: Jun 2007

Hi Everyone,
Quote:
Attachment 2 shows the AkAbak power response prediction for the same system, radiating into 2 Pi space. Attachment 3 compares the power response data exported from Hornresp with that exported from AkAbak. The results are not precisely the same because of slight differences in the values of the frequencies sampled by the two programs. I can’t see how the model results could be interpreted as being "quite different", though :). Kind regards, David
__________________
www.hornresp.net 

5th May 2010, 10:58 AM  #1102 
diyAudio Member
Join Date: Sep 2008

Hornresp have multiple times been shown to resemble the real world very accurately.
As long as you know what you model and actually build it. Both in directivity response, output and excursion. The biggest gray area is when you start folding a pipe or TH where you simply don't know what you simulate, or how it behave in real world. Q of resonances differs a lot here, but in a good way. Time to take the new version out for a test drive now. 
6th May 2010, 08:40 AM  #1103  
diyAudio Member
Join Date: Jun 2007

Hi Everyone,
Quote:
The following example illustrates the difference between the throat acoustical impedance results produced using a plane wavefront model, and those produced using an isophase wavefront model, for a largemouth 100 hertz exponential horn. Attachment 1 shows the plane wavefront model results, identical to those produced by AkAbak, compared to the results for an equivalent infinite horn. Attachment 2 shows the isophase wavefront model results, as produced by Hornresp, compared to the results for an equivalent infinite horn. Attachment 3 shows the results for the infinite 100 hertz exponential horn, separately. For a constant flare, as the horn becomes longer and the mouth becomes larger, the throat acoustical impedance results should tend asymptotically towards the infinite case. This happens with the Hornresp isophase model but not with the AkAbak plane wavefront model. Kind regards, David
__________________
www.hornresp.net 

6th May 2010, 08:51 AM  #1104 
diyAudio Member
Join Date: Jun 2007

Hi Everyone,
Speaking of source code, some users might find the following of interest. The attachment shows one of the 80column Hollerith input punch cards taken from the original Hornresp source code listing. I still have the complete set of cards for the program  they are now more than 40 years old. The deck of cards making up the program is 8 cm thick. The actual Fortran IV code statement can be seen printed at the top of the card. It represents one line from the subroutine responsible for calculating the throat acoustical impedance of a finite conical horn. The original Hornresp program ran as a batch file on a roomsize mainframe computer :). Kind regards, David
__________________
www.hornresp.net 
6th May 2010, 10:45 AM  #1105 
diyAudio Member
Join Date: Aug 2007

I don't have the data for the horns David posted, but here's a comparison of the throat impedance of a large exponential horn in an infinite baffle (CIR=2), simulated both in Hornresp and in BEM. Again, Hornresp is quite close, and David is right about the fact that the throat impedance gets smoother as the mouth size is increased beyond CIR=1.
Bjørn 
6th May 2010, 08:12 PM  #1106 
diyAudio Member
Join Date: Sep 2008

Using loudspeaker wizard on this data gives me "Runtime error 6: Overflow"
ID=26.40 Ang=2.0 x Pi Eg=2.83 Rg=0.00 Fta=0.00 S1=100.00 S2=100.00 Con=20.00 F12=0.00 S2=0.00 S3=0.00 L23=0.00 F23=0.00 S3=0.00 S4=0.00 L34=0.00 F34=0.00 S4=0.00 S5=0.00 L45=0.00 F45=0.00 Sd=490.00 Bl=8.50 Cms=1.10E03 Rms=1.00 Mmd=63.10 Le=1.60 Re=3.60 Nd=1 Vrc=20.00 Lrc=20.00 Ap1=0.00 Lpt=0.00 Vtc=8000.00 Atc=1000.00 Com=MGR 12" BP 
6th May 2010, 09:38 PM  #1107  
Mark Kravchenko
diyAudio Member

Quote:
I just missed punch cards and Fortran. I'm not that old! But I remember having to take all the fundamentals in Computer Science. Thanks for the trip down memory road David. So if I have this figured out correctly you have been doing horn simulations for what 35 years? Maybe 40? Explains the thoroughness of the program. And I have yet to see anyone do as tight code as you. All the functionality in still less than a MB. Nice. THe computers I took programing on were memory enhanced to 32 Kb and ran on Z80 processors. My high level language was the precursor to C. Turbo Pascal. Those were the days when I had hair! LOL I have but one comment to make on Hornresp's accuracy. The little wierd bandpass variant I worked out checked in as almost perfect to the hz and magnitude of the drivers impedance in box. I have to say I was floored. I think I posted a comment back then and I have unfortunately lost the screen shots but considering that the box was neither a horn or a normal vented or sealed box your method of calulating the resulting output and the impedance magnitude is most impressive. Mark Mark 

6th May 2010, 11:25 PM  #1108  
diyAudio Member
Join Date: Mar 2005
Location: Taiwan

Quote:
__________________
Hear the real thing! 

7th May 2010, 08:34 AM  #1109  
diyAudio Member
Join Date: Jun 2007

Quote:
Many thanks for this. I used a somewhat more extreme example than you did, just so that the trend towards the infinite case could be clearly seen. The relevant Hornresp parameter values are Ang = 2,0 x Pi, S1 = 10,00, S2 = 75335,04 and L12 (Exp) = 244,38, giving F12 = 100,00 hertz, Cir = 4,00 and AT = 1,87 degrees. While a Cir value of 4 is unlikely to be used in practice, the example serves to illustrate the considerable difference between plane and isophase wavefront models when applied to horns with large mouth Fta values. It is certainly necessary to use an isophase model for Le Cléac'h, spherical wave and tractrix horns, to obtain reasonably accurate results. It is interesting that with the BEM simulation results, the resistive component of the normalised throat impedance appears to tend towards a value just above 0,9 rather than to 1, at the higher frequencies. Kind regards, David
__________________
www.hornresp.net 

7th May 2010, 08:48 AM  #1110  
diyAudio Member
Join Date: Jun 2007

Quote:
Many thanks for this! I get the same error  I will let you know when the bug is fixed. Kind regards, David
__________________
www.hornresp.net 

Thread Tools  Search this Thread 


Similar Threads  
Thread  Thread Starter  Forum  Replies  Last Post 
Hornresp Class  FlipC  Subwoofers  8  3rd November 2008 07:23 PM 
Some questions about hornresp  brsanko  Full Range  4  18th October 2008 10:36 PM 
Hornresp help / JX150  316a  MultiWay  0  11th February 2004 04:56 PM 
New To Site?  Need Help? 