jaya000 said:
Hi jaya000,
As a general rule, just like Bjørn, most Hornresp users tend to set Atc = Sd. It is usually close enough for practical purposes.
Note that if the 'Resonances Masked' option is selected, then the value of Atc has no effect on the SPL response prediction anyway
.You haven't given the value for Sd, but I see from the plan that the driver is nominally 18 inches in diameter. If we assume that the equivalent piston diameter is 16.5 inches then Sd = Pi * (16.5 / 2 * 2.54) ^ 2 = 1379.51 sq cm, which is pretty close to the 1400 figure being used.
You would really need to ask the designer how the figure was arrived at, if you want to know for sure
.Kind regards,
David
Hi yaja000,
I am designer of that cabinet.
Atc is averaged from throat chamber dimensions. It is not set on exact number but "rounded" to first "round" value. It is not such critical because change of 40-50 cm2 will not make sizeable difference.
(as long as you do not do to extremes or set Atc area lower than Sd area)
B.T.W. .. that one is not real "scoop" by design. Because of it's relatively short horn(port) and relatively big throat chamber it behaves more like scoop/Bass reflex hybrid.
Best wishes,
Stipe
I am designer of that cabinet.
Atc is averaged from throat chamber dimensions. It is not set on exact number but "rounded" to first "round" value. It is not such critical because change of 40-50 cm2 will not make sizeable difference.
(as long as you do not do to extremes or set Atc area lower than Sd area)
B.T.W. .. that one is not real "scoop" by design. Because of it's relatively short horn(port) and relatively big throat chamber it behaves more like scoop/Bass reflex hybrid.
Best wishes,
Stipe
Hornresp Version 20.30
Hi Everyone,
Just letting you know that the 'S2 - fixed' option as suggested by Oliver ('tb46') has now been added to the Hornresp Tapped Horn Wizard tool. Version 20.30 refers.
My thanks to Oliver for the suggestion!
The S2 and S3 values (for a 3 segment tapped horn), or the S2 and S4 values (for a 4 segment tapped horn), are now also shown on the wizard schematic diagram.
Could you please let me know if you find any bugs when using the new functionality. While I try to pre-test things thoroughly, there is always the chance that I might miss something - it has happened in the past.
Also, please note that after giving it much thought I have decided not to include any signal processing functionality in Hornresp. Because input-signal manipulation is not really part of horn loudspeaker theory (which is my main interest and focus) I simply could not find the motivation or enthusiasm necessary to take on such an ambitious task. Sorry.
Kind regards,
David
tb46 said:
Hi Everyone,
Just letting you know that the 'S2 - fixed' option as suggested by Oliver ('tb46') has now been added to the Hornresp Tapped Horn Wizard tool. Version 20.30 refers.
My thanks to Oliver for the suggestion!
The S2 and S3 values (for a 3 segment tapped horn), or the S2 and S4 values (for a 4 segment tapped horn), are now also shown on the wizard schematic diagram.
Could you please let me know if you find any bugs when using the new functionality. While I try to pre-test things thoroughly, there is always the chance that I might miss something - it has happened in the past
.Also, please note that after giving it much thought I have decided not to include any signal processing functionality in Hornresp. Because input-signal manipulation is not really part of horn loudspeaker theory (which is my main interest and focus) I simply could not find the motivation or enthusiasm necessary to take on such an ambitious task. Sorry
.Kind regards,
David
S2 Fixed
Hi David,
Thank you very much for this new feature.
I have given it only a quick try (I'm half asleep), and it looks as if everything is working fine. One additional suggestion: it might be more useful to have the default as "Horn - S2 Fixed" instead of "Horn - S2 Variable", otherwise S2 will be altered if one leaves the Wizard and then re-enters.
Anyway, Great work once again,
Regards,
Oliver.
Hi David,
Thank you very much for this new feature.
I have given it only a quick try (I'm half asleep
), and it looks as if everything is working fine. One additional suggestion: it might be more useful to have the default as "Horn - S2 Fixed" instead of "Horn - S2 Variable", otherwise S2 will be altered if one leaves the Wizard and then re-enters.Anyway, Great work once again,
Regards,
Oliver.
G'day David
Another addition that might be worthwhile is the ability to add some loss in the line. I've just put up a comparison between a HornResponse model and the measured results for a 35Hz Tapped Horn on my webpage. The measured Tapped Horn shows lower Q in the peaks than the model, and the actual horn was undamped. Tom Danley uses damping in most or all of his Tapped Horns and it has a great effect on the amplitude of the peaks. A simple variable lumped loss would allow builders to fine tune their designs to even more closely match the model. Once a damping figure was established for the way a builder constructs a typical line, that figure could be used in future designs. I believe Tom's AkAbak script uses a loss component.
That said, I still find it remarkable just how close your model comes to measured results. All the main features are at exactly the right frequency.
Thanks again for a great job!
Cheers
William Cowan
Another addition that might be worthwhile is the ability to add some loss in the line. I've just put up a comparison between a HornResponse model and the measured results for a 35Hz Tapped Horn on my webpage. The measured Tapped Horn shows lower Q in the peaks than the model, and the actual horn was undamped. Tom Danley uses damping in most or all of his Tapped Horns and it has a great effect on the amplitude of the peaks. A simple variable lumped loss would allow builders to fine tune their designs to even more closely match the model. Once a damping figure was established for the way a builder constructs a typical line, that figure could be used in future designs. I believe Tom's AkAbak script uses a loss component.
That said, I still find it remarkable just how close your model comes to measured results. All the main features are at exactly the right frequency.
Thanks again for a great job!
Cheers
William Cowan
Greets!
At least as remarkable is that this is with published specs! Are all Peerless published specs this close to reality?
Yes, if you go back to where I used MJK's excellent MathCad software to show that end loaded THs are basically driver-in-mouth TLs, BLHs (if a filter chamber is used), then you can see how much damping affects its overall response in general and its HF response in particular.
GM
At least as remarkable is that this is with published specs! Are all Peerless published specs this close to reality?
Yes, if you go back to where I used MJK's excellent MathCad software to show that end loaded THs are basically driver-in-mouth TLs, BLHs (if a filter chamber is used), then you can see how much damping affects its overall response in general and its HF response in particular.
GM
G'day GM
In never fails to amaze me just how close those Peerless drivers are to spec. I've measured many of them (12" CC, XLS and XXLS only) and they are always right on the money as long as the test is done at ~20degC. I can't say the specs of many drivers come out that close. A Morel MW166 I measured the other day had a Qt that was almost double the spec!
One test I did when all this started was to place damping at the velocity maxima for the first out of band peak. This did reduce the Q of the peak, but I also lost some sensitivity. I think Tom's approach with just lining the walls is a better one. Better response without the loss in sensitivity.
Cheers
William Cowan
In never fails to amaze me just how close those Peerless drivers are to spec. I've measured many of them (12" CC, XLS and XXLS only) and they are always right on the money as long as the test is done at ~20degC. I can't say the specs of many drivers come out that close. A Morel MW166 I measured the other day had a Qt that was almost double the spec!
One test I did when all this started was to place damping at the velocity maxima for the first out of band peak. This did reduce the Q of the peak, but I also lost some sensitivity. I think Tom's approach with just lining the walls is a better one. Better response without the loss in sensitivity.
Cheers
William Cowan
Greets!
Good to know, thanks! My limited experience with the brand dates back to when RadioShack 'rubber stamped' them for sale. They too were great value and while the only DIY specs were Fs, usable BW and recommended cab, they performed as advertised with high consistancy.
I agree, lining is much preferred to stuffing. The pioneers of audio did extensive testing on the subject and the only alignments I'm aware of that they stuffed were to create ~aperiodic ones in the huge folded theater bass horns. All others were typically only lined on one of each parallel walls with none on the one closest to the vent if used. Much later, stuffing came into its own with the advent of acoustic suspension alignments and was begun to be used to EQ intentionally under-damped and/or 'noisey' vented alignments.
GM
Good to know, thanks! My limited experience with the brand dates back to when RadioShack 'rubber stamped' them for sale. They too were great value and while the only DIY specs were Fs, usable BW and recommended cab, they performed as advertised with high consistancy.
I agree, lining is much preferred to stuffing. The pioneers of audio did extensive testing on the subject and the only alignments I'm aware of that they stuffed were to create ~aperiodic ones in the huge folded theater bass horns. All others were typically only lined on one of each parallel walls with none on the one closest to the vent if used. Much later, stuffing came into its own with the advent of acoustic suspension alignments and was begun to be used to EQ intentionally under-damped and/or 'noisey' vented alignments.
GM
GM said:
It also depends what frequency you want the lining/stuffing to be effective at, in other words whether you're trying to damp low-frequency longitudinal resonances or higher-frequency lateral ones.
For a tapped horn the resonances we're most worried about are those just above the passband which are longitudinal, to damp these out you want to damp airflow along the horn not across it, preferably without adding too much loss at the quarter-wave resonance which defines the LF cutoff.
In this case lining the walls may not be the best method since the airflow slows down there, it might be better to have a lengthwise damping layer down the centreline of the horn (like a lengthwise partition) where the air velocity is a maximum.
If possible it might also be preferable to position this along the horn only close to the velocity maxima for the higher resonances but not the fundamental one.
This isn't so easy since Hornresp can't plot the pressure and velocity profiles along the horn at a given frequency, though the program must know this -- David, any chance of adding this?
(I guess the answer is no since it's a fundamental change)
Ian
IME the only time you want anything but a lining in any type of TL loading (TL, TQWT, horn or mass loaded (vented) variants) is when there's excessive peaking at Fb, so unless there's a need to deliberately design an under-damped alignment and acoustically response shape it, stuffing should be avoided. Since all these are fundamentally 1/4 WL resonators, then obviously the max broadband benefit of any damping is at the closed end.
GM
GM
Re: Version 20.30 Product Number 2030-090201
Hi Oliver,
It is a feature, not a bug.
When the "L23 + L34 Fixed" option is selected on a four segment tapped horn, the path length between the two sides of the driver remains constant. Although the values of L23 and L34 can change, the sum of L23 and L34 is fixed. This can be readily seen by shifting the L12 or L45 sliders while viewing the schematic diagram. The L23 and L34 sliders are disabled because their values are dependent upon the values given to L12 and L45.
Kind regards,
David
tb46 said:
Hi Oliver,
It is a feature, not a bug
.When the "L23 + L34 Fixed" option is selected on a four segment tapped horn, the path length between the two sides of the driver remains constant. Although the values of L23 and L34 can change, the sum of L23 and L34 is fixed. This can be readily seen by shifting the L12 or L45 sliders while viewing the schematic diagram. The L23 and L34 sliders are disabled because their values are dependent upon the values given to L12 and L45.
Kind regards,
David
iand said:
Hi Ian,
You are correct in assuming that Hornresp knows the pressure and velocity values, but because it would require a major change to the program to implement, as you have guessed, at this stage the answer has to be 'no'. Sorry
.AkAbak could always be used, of course
.Kind regards,
David
iand said:
Hornresp knows the pressure and velocity values at S1, S2, S3, S4 and S5 (if in use). Plotting the distribution along the horn is something different...
I can't resist attaching a screenshot of how such a plot would look like in polar coordinates
The discontinueties at 15 and 365 cm are the points where the driver enters the horn.Bjørn
