Hornresp

David McBean · 2019-01-10 8:17 am

Kolbrek said:
As T increases, distortion decreases, but keeping the horn end dimensions and length the same, so does the cutoff frequency

To make matters worse, for a fixed Eg, the sound intensity at the horn throat also varies with T, further affecting the calculated distortion figures

.

Kravchenko_Audio · 2019-01-10 3:55 pm

David McBean said:
To make matters worse, for a fixed Eg, the sound intensity at the horn throat also varies with T, further affecting the calculated distortion figures .

So you mean it's complicated to calculate!

johnnyjo · 2019-01-11 3:55 am

Regarding Multi-Entry Horn Wizard...

When doing an ME design the ND record created by the wizard is for the high-frequency driver. In this record am I supposed to enter parameters for the HF driver? If so, which parameters should I enter? HF drivers don't have many parameters and I couldn't find any examples for using the wizard. Just need some direction.

David McBean · 2019-01-11 6:44 am

mwmkravchenko said:
So you mean it's complicated to calculate!

It's not that difficult to calculate, but it is another variable that affects the distortion result at a given frequency, along with the values of T and fc.

David McBean · 2019-01-11 6:59 am

johnnyjo said:
In this record am I supposed to enter parameters for the HF driver? If so, which parameters should I enter?

Hi johnnyjo,

If you don't know the actual parameter values for your HF driver, then use the ones given in the multiple entry horn Nd record created by the Input Wizard.

Sd = 20.00
Bl = 8.00
Cms = 1.60E-04
Rms = 3.00
Mmd = 2.00
Le = 0.05
Re = 6.00
Nd = 1

Vrc = 1.00
Lrc = 8.00

Fr = 40000.00
Tal = 4.00

Vtc = 10.00
Atc = 20.00

Kind regards,

David

johnnyjo · 2019-01-11 3:39 pm

The manufacturer doesn't give any specs on the driver. That's why I'm asking. I'm trying to sim a BMS 4594 coax.

Specification

David McBean · 2019-01-12 6:21 am

Kolbrek said:
The limit as T->1 is 1. It is also valid for T>1, where both the numerator and denominator are complex, but the result is real.

...as shown in the attached chart, which plots Bjørn's Hypex correction factor Dh from T = 0 to T = 15.

David McBean · 2019-01-12 6:26 am

johnnyjo said:
The manufacturer doesn't give any specs on the driver.

I am yet to find a compression driver manufacturer that does. They are all very secretive about such things, for some reason

.

papasteack · 2019-01-12 8:18 am

David McBean said:
I am yet to find a compression driver manufacturer that does. They are all very secretive about such things, for some reason .

^^
Modelling compression drivers

Kolbrek · 2019-01-12 8:46 am

David McBean said:
They are all very secretive about such things, for some reason .

I think there are several reasons. While a lumped parameter model may be used early in the design process, most of the design is done using FEA. In many cases a driver is a modification of a previous model. It is possible to fit a lumped model to an FEA model, but it is extra work that may not pay off in more sales.

When the diaphragm starts to break up, a lumped model can be quite misleading, and show too little HF output compared to the real driver. There are also extra cavities and channels that require a more complex lumped model than the standard woofer model. A simple lumped model will only be a rough approximation, more so than is the case with a woofer.

But the main reasons are probably two things: too few (large) customers ask for it, and there is no readily available method to measure the lumped parameters of a compression driver, in the way you can do it for a cone driver.

Kravchenko_Audio · 2019-01-12 6:22 pm

Kolbrek said:
I think there are several reasons. While a lumped parameter model may be used early in the design process, most of the design is done using FEA. In many cases a driver is a modification of a previous model. It is possible to fit a lumped model to an FEA model, but it is extra work that may not pay off in more sales.

When the diaphragm starts to break up, a lumped model can be quite misleading, and show too little HF output compared to the real driver. There are also extra cavities and channels that require a more complex lumped model than the standard woofer model. A simple lumped model will only be a rough approximation, more so than is the case with a woofer.

But the main reasons are probably two things: too few (large) customers ask for it, and there is no readily available method to measure the lumped parameters of a compression driver, in the way you can do it for a cone driver.

Sounds like a paper for a talented guy to research and write!

cogito321 · 2019-01-13 12:14 am

Multisegment Estimator

Hi,

I am modeling an single segment exponential horn. Can find the menu for conversion of single segment horn into multiple segment horn.

Anyone?

kazap · 2019-01-13 1:47 am

Please can I ask if there is a new way to accurately model a "Ripole" open baffle woofer in Hornresp? If so please attach the Hornresp file.

David McBean · 2019-01-13 7:26 am

cogito321 said:
I am modeling an single segment exponential horn. Can find the menu for conversion of single segment horn into multiple segment horn.

Specify your single-segment exponential (Exp) horn as an equivalent hyperbolic-exponential (Hyp) horn with T = 1, then use the Hypex Approximator tool.

David McBean · 2019-01-13 7:30 am

kazap said:
Please can I ask if there is a new way to accurately model a "Ripole" open baffle woofer in Hornresp?

Another one for AkAbak...

.

kazap · 2019-01-13 8:31 am

David McBean said:
Another one for AkAbak... .

Thanks

jimmosher · 2019-01-15 12:23 am

cogito321 · 2019-01-15 1:54 am

David McBean said:
Specify your single-segment exponential (Exp) horn as an equivalent hyperbolic-exponential (Hyp) horn with T = 1, then use the Hypex Approximator tool.

Got it. Thanks.

David McBean · 2019-01-15 5:48 am

jimmosher said:
This is the first installment of an index file. A short description is followed by a link to the exact post.

Hi Jim,

You have been busy!

Many thanks for doing this - it is a great resource.

Kind regards,

David

David McBean · 2019-01-15 7:27 am

David McBean said:
Hi freddi,

Many thanks for this excellent feedback! You are not "being an idiot" - far from it . Hornresp simply should not do such things under any circumstances - there is something wrong somewhere.

It seems that you have identified two issues; firstly, that the diaphragm displacement is not what we would expect to see from a DBR system, and secondly, that it is possible to crash the wizard as you have shown. I have tried to replicate the problem you experienced, but so far without success. Would it be possible to provide details of the sequence of events that leads to the blank input drop-down list box being displayed? If I can generate the problem myself, then with a bit of luck I should be able to track down the cause and fix things.

Thanks and kind regards,

David

Further to my message above...

Try as I might, I still have not been able to replicate the problem experienced by freddi. On examining the relevant source code carefully, I cannot see how it could possibly crash the Wizard in the way shown. In the absence of any further feedback I suspect that freddi may have experience a momentary glitch with his computer that did not allow the Hornresp screen to refresh as it should.

With regard to the diaphragm displacement issue, I have now checked that the result produced by the Wizard (see Attachment 1) is indeed valid, by comparing it to the one produced when the same double bass reflex system is specified independently of the Wizard (see Attachments 2 and 3). A completely different model is used in this case, but the displacement results remain the same, as anticipated.

At this stage, I intend taking no further action on either matter, unless of course someone can provide me with a sequence of events that generates the problems experienced by freddi.

Hornresp

Member

Member

Member

Member

Member

Member

Member

Attachments

Member

Member

Member

Member

Member

Member

Member

Attachments

Member

Member

Member

Member

Member

Member

Attachments