Hornresp

Yup, that would be fine by me.

Hi Brian,

The good news is that I have managed to come up with a reasonably efficient way of finding the lowest resonance frequency, which means that it should be possible to remove the 100Hz limit and use the default chart maximum of 2000Hz instead, so that the frequency range tested is no longer an issue :).

I ended up identifying the resonance frequency from the electrical impedance rather than from diaphragm displacement. The results are significantly more accurate and consistent when impedance is used.

Hopefully the lowest resonance frequency feature will be included in the next release.

Kind regards,

David
 
When talking about portet enclosures (bassreflex), the impedance resonance peaks don´t correspond to the resonance frequency of the port. Excursion Minimum is close, but not exactly the same. I´d have to look up the formula (Vance Dickasson has it in his measurement book) how to calculate port-resonance from the imp-chart. But I guess David already has something figured out ?
 
When talking about portet enclosures (bassreflex), the impedance resonance peaks don´t correspond to the resonance frequency of the port. Excursion Minimum is close, but not exactly the same. I´d have to look up the formula (Vance Dickasson has it in his measurement book) how to calculate port-resonance from the imp-chart. But I guess David already has something figured out ?

Yes. That is true.

It is the same situation with a bandpass.

Impedance minimum.

It''s those differences that make an easy solution a little bit more difficult to achieve.
 
In a few situations it is a combination of phase angle at the zero crossing point and the impedance at that point.

But I to have to look if there are exceptions to this.

So many things to keep in mind that this work never ceases to impress upon me what bits of knowledge I am missing. Sometimes because of lack of use. Sometimes because of not knowing.

It keeps me always wanting to learn more and ask and seek the answer to different questions.

Never a unhappy moment in this business.
 
Hi David


Is the port dimensions such that I have to divide them? like the ap1 is 12 cm wide and has 113 cm2 total for 4 ports, (two each woofer).

It is for to be shure, because I think I have the answer already.

when divide them horn is more happy keeping material on him.

thanks
 

Attachments

  • ScreenHunter_945 Jan. 28 18.35.jpg
    ScreenHunter_945 Jan. 28 18.35.jpg
    46.5 KB · Views: 155
Last edited:
Hi Everyone,

Before I go any further and just so that I understand exactly what is required, given the attached impedance and displacement charts, can we reach a consensus on what is considered to be the 'lowest resonance frequency' of the sample system?

It would seem from recent posts that we have three options to choose from:

1. First impedance peak - 41.20 hertz.
2. First impedance minimum after the first peak - 61.49 hertz.
3. First displacement dip - 62.38 hertz.

(The frequency of first impedance minimum after the first peak is usually very close to the frequency of the first displacement dip).

What is the 'lowest resonance frequency' of this system?

Kind regards,

David
 

Attachments

  • Attach_1.png
    Attach_1.png
    61.6 KB · Views: 140
  • Attach_2.png
    Attach_2.png
    61.9 KB · Views: 132
  • Attach_3.png
    Attach_3.png
    62.3 KB · Views: 137
Is the port dimensions such that I have to divide them? like the ap1 is 12 cm wide and has 113 cm2 total for 4 ports, (two each woofer).

Hi Kees,

Hornresp has no way of knowing how many ports there are, so you need to specify the total area.

For the example you give, if Ap is shown in Hornresp as 113 cm2 and you intend having four ports, then you are assuming that each of those ports has a cross-sectional area of 28.25 cm2 and a diameter of 6 cm.

Kind regards,

David
 
Definitely not #1. Resonances happen between impedance peaks. Excursion is already way out of control at the lowest impedance peak so there's no point at all in marking the lowest impedance peak.

#2 might be more technically correct than #3 but it hardly makes any difference, it's less than 1 hz apart. if there's a question of which is easier to implement I'd go for the easier option between #2 and #3.
 
Theoretically, it's the mean between the two peaks IIRC, but don't know how HR calculates the peaks in #2 and dip in #3, which would theoretically be at the same frequency, so no clue why there's a slight offset. Regardless, I use #3 as it's plenty close enough.

GM
 
Loudspeaker Driver Design

While dated, the following book, published by the ASA, may help those that have interest in the subject.

"Electroacoustics -The Analysis of Transduction and Its Historical Background"
Author: Frederick V. Hunt

Electroacoustics

Note that inductance is a function of the total number of turns in the voice coil, while motor strength (bl product) is a function of the length of voice coil wire effectively immersed in the magnetic gap. Thus, a smaller diameter voice coil typically will have a much higher inductance than that of a voice coil of larger diameter where the dcrs and bl products of each driver are comparable.

WHG
 
Last edited: