Front horn loaded with narrowed expansion

[blackgotan71]

Hi everybody, i am currently working on a Jubilee style project: 2 horn crossed at 500Hz. After extensive research to find the good driver, I went to an Audax PR300M0 (Fs 17Hz, Qts 0.26) and started simulations on Hornresp. You will find below two simulations of front horn loaded, closed chamber. 

One has a hyperbolic expansion, the other one is more atypical with a narrowing after first expansion. This intrigues me with its unconventional shape while it offers a much wider and smoother response curve in the low frequencies with a small loss of sensitivity (driver, chamber volume,  surface Sd, horn length are identical for the 2 simulations). 

I turn to you for advices: should I trust the Hornresp simulation? Should I focus on this narrowed expansion which seems to provide a 1 Pi response from 40Hz to 500Hz with a single 12' driver?

Thank you in advance for your advice

[grindstone]

The short answer is that it's your experiment to make. 

 

First longer answer is that I have noticed that behavior & geometry too when breaking down the (65+ yr old) EV regency, although not a fronthorn.

 

Second bit is that hornresp has been trustworthy in my 14+? yrs using it Provided You Understand what it's really simulating and how different actual conditions and geometry are.    If you can build a straight cylindrical horn 5 ft long that all radiates into one pi conditions, your power response will have strong similarities to the simulation.  If you think about it, radiation near the xo point will likely be into more nearly 2 pi conditions, so there will be a transition region.  If your design modifies the straight horn to fit into more typical living room conditions, expect other changes that may not be directly modelable in the available segments in HR.  Beyond that, there's the business of any driver's response not being flat in and of itself.  Now, presumably you will stick this low frequency unit in some room which will have its own room gain and standing waves.  Then there's the business of group delay.  Then there's that pesky crossover with polars and directivity.  All this is just DIY audio

 

You will need to do multiple simulations and adjust for things in your head and/or use more sim tools and measurement.  In the end it's your labor and money and sawdust.  Good luck and please keep us posted.  HR is always right for what it does, we just never build what we can model with it.

 

 

 

[Chris A]

On 2/14/2021 at 12:29 PM, blackgotan71 said:

the other one is more atypical with a narrowing after first expansion. 

This will generally kill your effective bandwidth of the horn/driver if you're reducing the area after the compression slot plane.  But you can certainly try it and measure the effects. 

 

The general rule is that the area expansion is positive all the way through, with no reductions in area anywhere along the way.

 

Chris

[PrestonTom]

I am skeptical of that the simulation (or its set up is correct). The horn length and area at the mouth are usually too short & small to get true horn loading at the low frequencies (below 100-200Hz). In the jubilee (51 inches in length) there are peaks and dips at about 100, 200, 300 Hz. The differences (peak to dip) are over 6-8 dB (see the JAES paper). Now corner placement and DSP will help (yours are more likely 2 pi space however)). Your simulation shows a much smaller peak-to-dip ratio (only a few dB at most). So I would take you predictions with a grain of salt. If you are convinced, then by all means make some sawdust and find out. Perhaps a quick & dirty prototype is called for. I did not see the units - is this horn incredibly large?

Good luck (and I have been wrong before),

-Tom