Hornresp

[David McBean]

It would be very useful if the power (Y axis) would scale similarly to the other charts.

Hi NiilaH,

Thanks for thanks, and for the feedback. I will try to include the change in the next update.

Kind regards,

David

 

[LORDSANSUI]

Hello David,

Regarding the bug, even creating a new input from scratch the issue remains, so probably could be in the Linux environment, at wine, don't know, the hornresp should be simple enough using visual basic language to be properly translated. I can live with the issue and fix it manually all the time.

If it help the last check, the data file is saved correctly as you can see below, looks like the issue happen while reading the data and showing in the screen.

|CHAMBER PARAMETER VALUES:

S6 = 800.00
Con = 14.00
S8 = 1200.00
Con = 180.00
S9 = 1200.00
Con = 1.50

 

[LORDSANSUI]

Hello David,

Me neither. Perhaps it could be worth checking with the experts on the Paraflex Facebook group to see how they go about it?

I tried but i don't see them publishing the input parameter, it's always just the output, maybe they like to keep their secrets.

If I don't disturb you too much, I think understanding hornresp is enough, from math or programming standing point, you defined 3 volumes: H7, H3 and H4, all with respectively throat and mouth, how did you made the connection? there is a high chance that if the math works, probably the volumes in the real world might work too.

One more questions.

Considering the knowledge you have, would it be possible to define a side connection in hornresp as you can see in the attachment #2?

I will not even try to ask you new feature, i'm glad you hekp me a lot and probably the right way now is to find how people are using the current model available, asking Brian Steele help would be an option once he has spreadsheet for paraflex type. In a while I'm restless because I can't figure out how to make the current connection in the real world.

link H7 mouth to H4 troat
link H7 mouth to H3 mouth
link H3 mouth to H3 troat

I think it's only possible if all are POINTS, being a volume it's not maybe it's not feasible. Two objects cannot occupy the same space simultaneously.

 

[b_force]

Hi NiilaH,

Thanks for thanks, and for the feedback. I will try to include the change in the next update.

Kind regards,

David

There is a workaround for this, which is exporting the graph, and importing it in another program.
But yeah that's a little cumbersome sometimes.

If it's not to difficult, it would also be nice to change the freq axis as well.
The only two options are 1 - 2k and 10 - 20k.
Although the loudspeaker wizard is only 10 - 2k (an can't be changed as far as I know?)

 

[LORDSANSUI]

Hello David,

Just a complement, for my sanity (LoL) I checked the system model for Taped Horn, see attachment, because that model is very well know and I build a Loudspeaker using that model witch result is very close to the simulation. That model also consider a connection (node) of tree volumes, in this case the driver and it's placed in between at S2 between H1 and H2, a kind of side feeding like I did for the paraflex, but in paraflex I tough that the area of the H7 would work in a similar way so I placed H7 at S4 between H3 and H4.

 

[David McBean]

If it help the last check, the data file is saved correctly as you can see below

Thanks, but unfortunately it doesn't really help me much . The difficulty is that everything works just fine for me so I don't know where to go looking for a problem. There have also been other instances in the past where Linux / Wine and Hornresp haven't liked each other, which I have not been able to do anything about.

 

[David McBean]

I tried but i don't see them publishing the input parameter, it's always just the output

That's a pity - there is little to be gained by anyone from doing that. How can things be improved without sharing information? Hornresp is a classic example of the benefit to come from sharing, with the improvements made over the years being mainly suggested by users of the program, not thought up by me .

from math or programming standing point, you defined 3 volumes: H7, H3 and H4, all with respectively throat and mouth, how did you made the connection?

The simulation model assumes that the three connecting lines are acoustic paths and calculates the volume velocities through them and the pressure at the common connection point. In effect, exactly the same as calculating the currents and voltage in an electrical circuit.

would it be possible to define a side connection in hornresp as you can see in the attachment #2?

The 1P (one-parameter) horn theory used in Hornresp means that horn segments are in effect two-port devices, and that it is not possible to have a third port in the side of a segment. The only way that an offset can be applied is to have it at the interface between two adjoining segments, as is currently done.

I think it's only possible if all are POINTS, being a volume it's not maybe it's not feasible. Two objects cannot occupy the same space simultaneously.

H3, H4 and H7 are lumped element segments and have volumes - but yes, in effect they are connected together at a single point.

 

[David McBean]

If it's not to difficult, it would also be nice to change the freq axis as well.
The only two options are 1 - 2k and 10 - 20k.
Although the loudspeaker wizard is only 10 - 2k (an can't be changed as far as I know?)

The default frequency range is 10 - 20000 Hz.
From the Input Parameters window the Tools > Frequency Range menu commands can be used to change the range to 1 - 2000 Hz.

If the frequency range is set to 10 - 20000 Hz the default range in the loudspeaker and filter wizards is 10 - 2000 Hz. If no filling material is specified the wizard range can be changed to 100 - 20000 Hz by double-clicking on the Frequency (hertz) label at the bottom of any chart.

If the frequency range is set to 1 - 2000 Hz the default range in the loudspeaker and filter wizards is 1 - 200 Hz. The wizard range can be changed to 10 - 2000 Hz by double-clicking on the Frequency (hertz) label at the bottom of any chart.

For many reasons, this is not going to change .

 

[GM]

Hornresp is a classic example of the benefit to come from sharing, with the improvements made over the years being mainly suggested by users of the program, not thought up by me .

For real! I lost track of all the wishes that you basically said was never going to happen and did with way more than I even knew to ask for!

 

[David McBean]

For many reasons, this is not going to change .

It is of course possible to export chart data as a .csv file and open in Excel or similar, where the frequency axis range can be changed if required.

 

[David McBean]

I lost track of all the wishes that you basically said was never going to happen and did with way more than I even knew to ask for!

The attached list might help you to remember. That's just the main version updates - bug fixes not included .

 

[GM]

If I may since there's so many computer savvy folks here, a way off topic subject:

Last week I could see videos everywhere I've gone since 'forever' and this week I can't see them anywhere! Here, youtube, FB, you name it!

They just show a black screen with a slow circling semi-circle, so can anyone help me or where best to ask as I'm basically clueless WRT computers as Googling tech help has IME become a very dangerous 'game' for the uneducated in recent years. I have Win 7Pro HP 450 G3 Probook, Google Chrome, Avast freeware is all I know to list.

TIA

 

[LORDSANSUI]

Hello David,

Regarding Linux/wine, no problem, I will find a way that I may can change something in the settings in order to fix it or maybe install a windows in a virtual machine environment so hornresp fell at home.

The simulation model assumes that the three connecting lines are acoustic paths and calculates the volume velocities through them and the pressure at the common connection point. In effect, exactly the same as calculating the currents and voltage in an electrical circuit.

I see, I love electrical / electronics, one advantage is that the node can have as many connection as we need, probably the simulations run in the same way, but in real life you have much more flexibility with electrical circuits, two capacitors can be connected in a node but in the circuit board place "far away" each other, in the acoustic environment the volumes are vary constrained.

are lumped element

Sharing above sentences cleared a lot for me, thank you again. and also based on that and as I pointed in the post #13,045, also the driver feeds in a 3-node connection, and if you change driver cone size, the input parameter for the volume H1, H2, etc, will not be affected, see attachment. So, I think the approach I had for paraflex is the same and don't know why you said in the post #13,034.

The connection of H7 to H3 and H4 as shown in your physical design is not the same as that assumed by the system model. You have the mouth of H7 feeding into the sides of H3 and H4, not into their mouth and throat ends respectively.

H7 mouth (S9) placed between H3 and H4 (S4) in a PH1 model correlates to driver D (Sd) placed between H1 and H2 (S2), don't?

Looks like the driver is feeding the sides of H1 and H2.

 

[b_force]

The default frequency range is 10 - 20000 Hz.
From the Input Parameters window the Tools > Frequency Range menu commands can be used to change the range to 1 - 2000 Hz.

If the frequency range is set to 10 - 20000 Hz the default range in the loudspeaker and filter wizards is 10 - 2000 Hz. If no filling material is specified the wizard range can be changed to 100 - 20000 Hz by double-clicking on the Frequency (hertz) label at the bottom of any chart.

If the frequency range is set to 1 - 2000 Hz the default range in the loudspeaker and filter wizards is 1 - 200 Hz. The wizard range can be changed to 10 - 2000 Hz by double-clicking on the Frequency (hertz) label at the bottom of any chart.

For many reasons, this is not going to change .

Yeah I was aware of that

Just a general question, but out of interest, why does this seem to be a such problematic thing in software?

 

[David McBean]

So, I think the approach I had for paraflex is the same and don't know why you said in the post #13,034.

Sorry for the confusion - I was in a bit of a hurry at the time and would perhaps have worded things differently if I had a little longer to think about it .

From a theoretical point of view your design is indeed consistent with the PH1 system model, but from a practical perspective L23, L34 and L45 are relatively short compared to the diameters of Sd and S9 and I suspect that this may reduce the accuracy of the predictions compared to measured results. The only way to know for sure would be to build the system and see.

 

[David McBean]

Just a general question, but out of interest, why does this seem to be a such problematic thing in software?

There is no inherent reason why it should be a problem - it is not particularly difficult to specify a logarithmic scale given lower and upper limiting values. I can't speak for other software but as far as Hornresp is concerned it is simply because I prefer to control all the outputs. It is the essentially same reason why the speed of sound and air density values are fixed, and cannot be changed by the user.

One practical consideration with a variable frequency range is that it takes a little more coding effort to compare chart results produced using different frequency ranges.

In Hornresp with the 1 - 2000 Hz range selected it is possible to sample the main charts to a tolerance of 0.1 Hz at any frequency between 0.1 and 2000 Hz. With the 10 - 20000 Hz range selected it is possible to sample to a tolerance of 1 Hz at any frequency between 1 and 20000 Hz. Surely this is enough .

 

[LORDSANSUI]

Sorry for the confusion - I was in a bit of a hurry at the time and would perhaps have worded things differently if I had a little longer to think about it .

From a theoretical point of view your design is indeed consistent with the PH1 system model, but from a practical perspective L23, L34 and L45 are relatively short compared to the diameters of Sd and S9 and I suspect that this may reduce the accuracy of the predictions compared to measured results. The only way to know for sure would be to build the system and see.

Hi David,

Thank again for clearing, i'm mind calm now, LOL.

I will do a lot of simulations and optimizations runs and regarding the results, build may not happen if any improvement is found compared to the TH.

I was checking if the model and my understand was right before investing time, you know, if shxt in, shxt out.

Have a nice day.

 

[David McBean]

Hornresp Update 5430-220925

Hi Everyone,

CHANGE 1

The Driver Electrical Input Power chart in the Filter Wizard is no longer limited to a maximum value of 800 watts. Post #13,038 refers.

CHANGE 2

A number of minor loose ends associated with the recently added TL Design tool have now been tidied up.

Kind regards,

David

 

[SamAnytime]

Power in GIGAWATTS!!!

 

[David McBean]

At the other end of the scale, 0.1 picowatts. Hopefully the Filter Wizard driver power range will now cover everyone's needs .