OSWG Acoustic Impedance

[ivica]

May be for some DIY members it would be interesting to see OSWG horn acoustic impedance.
I have done it for 1, 1.5 and 2 inch horn throat, with 30 deg and 45 deg off-axis angle.

Regards
Ivica

[ivica]

May be for some DIY members it would be interesting to see OSWG horn acoustic impedance.
I have done it for 1, 1.5 and 2 inch horn throat, with 30 deg and 45 deg off-axis angle.

Regards
Ivica

it seems that lower frequency loading (under 2kHz) is not so good as I have expected

[Ruediger]

Did you measure or compute the curves? What was the setup? Where does the fine ripple come from?

Ruediger

[ivica]

Did you measure or compute the curves? What was the setup? Where does the fine ripple come from?

Ruediger

Hi Ruediger,

No, these are iteratively calculated acoustic impedance curve.
The ripples are expected as this is not infinite length horn, so there has to be reflections from the horn mouth.
Putting larger length in the calculations such ripples would become smaller,

My point is to emphasize that relatively low real part of the acoustic impedance would be expected from 500Hz to 2khz using OSWG horn wall curvature, so for higher power application using metal-type suspension of the diaphragm (such as JBL 2441, 45, 46,47, 50,51,...2452) has to be done with caution, I think that 2441 (Aluminum), 2447, 2450SL, 2451 are especially vulnerable

regards
ivica

[fpitas]

Is this real surprising? I was under the impression that waveguides in general don't offer the same acoustic loading as a horn, being mainly optimized for dispersion.

[Ruediger]

What irritates me is the fine ripple even at 10 kHz and above. That means that the (effective) horn mouth is too small even at 10 kHz and a substantial part of the wave gets reflected back into the horn resp. wave guide.

Are you sure your calculations are right?

Ruediger

[ivica]

What irritates me is the fine ripple even at 10 kHz and above. That means that the (effective) horn mouth is too small even at 10 kHz and a substantial part of the wave gets reflected back into the horn resp. wave guide.

Are you sure your calculations are right?

Ruediger

Hi Ruediger,

These are mathematical ideal solutions, an approximations, but general trend is important. If You ever have done any more detailed resolution measurements You have seen a lot of 'bouncing' due to the reflection from edges, nearby objects etc. Here again I want to emphasize that OSWG horn curve seems to be not so good driver loading for lower frequency (say "under" about 2kHz) , but they are very good in directivity control and low HOM (high order modes)
Much more can be find in the:

http://gedlee.azurewebsites.net/downloads/AT/Chapter_6.pdf

Be careful while reading Chapter.6 as the author, Mr. Earl Geddes , has used C = 2*Pi*f*Ro/(Co* sin(theta)) on the - figures 6-9 and 6-11
instead of the frequency (f), where Ro -is throat RADIUS, and Co is the speed of sound * 345m/s), and theta is the off-axis angle.

http://gedlee.azurewebsites.net/Books/AudioTransducers.aspx

When I have done rescale of my calculation I have get Mr.Geddes figures, so I believe that the calculations are good, but some other members of the form can, may be confirm that or not.

Regards
Ivica

[ivica]

What irritates me is the fine ripple even at 10 kHz and above. That means that the (effective) horn mouth is too small even at 10 kHz and a substantial part of the wave gets reflected back into the horn resp. wave guide. Are you sure your calculations are right? Ruediger

Interestingly that using 2-inch drivers with internal short horn such as 2445, 2446, 2450, 2441 with OSWG
would "help" from the point of acoustic impedance that would load the driver. Here is the acoustic impedance for
OSWG horn with 2-inch throat, where a short conical horn of 4.6cm that would convert 1.5 inch driver
output to 2-inch horn throat (simulated short internal horn for 2446 driver)

regards
ivica