The cardioid (upper bass to midrange about >150Hz AFAIK) in "Krypton" and the old "Xenon" models also is a "resistance box" design: The triangular shaped "flow resistors" at the sides are quite characteristic.
https://www.hsound.co.uk/wp-content/uploads/2014/11/Amphion-Krypton-3-Loudspeaker.jpg
Although i do not know which materials, dimensions etc. used in detail.
But i listened longer to an early "Krypton" model at a show in germany near Frankfurt, where all the manufacturers present had to get along with empty hotel rooms as listening rooms.
I told the amphion guys after - thoroughly - listening to their "Krypton" and even a smaller model, that i would send the other manufacturers right home, because the amphions were the only ones - to my ears - who got their speakers working on that particular show under that particular conditions.
That happened more than 10 years ago, so even the models changed by now.
I remember well, most other conventional designs playing in those rooms could produce just "mud" e.g. in upper bass to lower midrange ... dragging the mid-to-high range down by masking.
Also irregularities in highs because of non matching radiation patterns of the drivers (Woofer/Midrange vs. Tweeter) were commonly audible at the other manufacturer's products ...
Sometimes bad conditions at a show may help in learning about loudspeakers a lot more, than good conditions ever could.
Cardioids are IMO definitively a way to go, at least from mid/upper bass to the midrange IMO.
https://www.hsound.co.uk/wp-content/uploads/2014/11/Amphion-Krypton-3-Loudspeaker.jpg
Although i do not know which materials, dimensions etc. used in detail.
But i listened longer to an early "Krypton" model at a show in germany near Frankfurt, where all the manufacturers present had to get along with empty hotel rooms as listening rooms.
I told the amphion guys after - thoroughly - listening to their "Krypton" and even a smaller model, that i would send the other manufacturers right home, because the amphions were the only ones - to my ears - who got their speakers working on that particular show under that particular conditions.
That happened more than 10 years ago, so even the models changed by now.
I remember well, most other conventional designs playing in those rooms could produce just "mud" e.g. in upper bass to lower midrange ... dragging the mid-to-high range down by masking.
Also irregularities in highs because of non matching radiation patterns of the drivers (Woofer/Midrange vs. Tweeter) were commonly audible at the other manufacturer's products ...
Sometimes bad conditions at a show may help in learning about loudspeakers a lot more, than good conditions ever could.
Cardioids are IMO definitively a way to go, at least from mid/upper bass to the midrange IMO.
Last edited:
What do you mean exactly?
EDIT: My question was directed at mayhem13. I had not seen LineArray's post when I responded. I'm assuming Oliver answered mayhem's question?
EDIT: My question was directed at mayhem13. I had not seen LineArray's post when I responded. I'm assuming Oliver answered mayhem's question?
Last edited:
Should read: "A very beneficial and interesting way to go", i was not implying it being the only way possible.
But that's understood, i guess 😉
(Maybe "a way to go" vs. "the way to go" already said that, but english is not my mother-tongue)
Last edited:
My new design plan calls for an MTM section but only 4" Scan 10F mids.....im thinking now to go with 150mm mids and cross lower......maybe 200 hz.....but with the resistive enclosure mayby start at 400hz but go 1st order on the high pass?
@mayhem13:
Im not quite sure whether i understood your concept correctly.
If you want to bring a cardiod into play above say 400Hz (1st order) (and use a conventional monopolar woofer system below ?) this seems fairly high to get a significant advantage from it IMO. But it may also work well ... depending on your concept as a whole and the actual application of that loudspeaker.
If you are afraid of overstressing the midrange drivers mechanically, you should calculate the possible max. SPL according to driver's displaced volume ("undistorted") and the effective dipole pathlength of your design.
E.g. Linkwitz' sheet gives you an estimation of the (excursion x Sd = Vd) needed for dipoles, which is rather pessimistic for a true cardioid:
http://www.linkwitzlab.com/spl_max1.xls
As Earl stated before, a true cardioid get's away with sqrt(2) times per octave the Vd of a monopole (below Fequal), while the dipole needs 2x Vd per octave in relation to the monopole.
(approximately the dipole needs 8x excursion per octave lower, monopole just 4x, cardiod could end up with 5,7x as an estimation)
So you can estimate/calculate, whether your midrange drivers have sufficient Vd for your desired level e.g. at 200Hz.
A first order highpass btw. is not sufficient to even keep the driver's excursion constant below the cutoff frequency ...
A driver not equalized will have 4x excursion per octave lower due to "natural mass hampering" while a
first order filter will just reduce that rise in excursion to a factor 2x ... the excursion is still rising even in the stopband of the first order filter (to be observed fairly above resonance of the driver):
Don't know if everyone using a first order highpass is being aware of that: Stressing the driver (at least increasing distortion) in the stopband (distortion will also affect the passband) is not a fiction, as also spectral statistics of most music genres will "play against the driver".
Last edited:
Unibox gives me the luxury of simming a leaky sealed box which in the past has proven fairly accurate in matching with the resistive enclosures i've tried. My first pass at this with a passive HP filter was difficult as the impedance peak of the drivers is dramatically smoothed with a resistive alignment.
@mayhem13
I tried the "unibox" modeling right now (because i am always curious) and do not think this is very close to what's happening in a (even simple version of a ... ) "resistance box type" cardioid:
Typical flow resistances needed seem far lower in reality IMO than even in the "minor leaks" setting in unibox (which is the most "leaky" setting). That will often cause the resonance peak to flatten even more in reality.
Furthemore that software gives no clue about the dipole component radiated (polar pattern), which depends on the effective dipole pathlength also, which is a fundamental enclosure parameter not accounted for in the model (because it is made for a "sealed box", not to blame the author of the software):
But when not knowing how to equalize exactly - due to dipole component in radiation - then it's hard IMO to estimate the drivers excursion (Vd) needed also: How should this be done using such an incomplete model ?
In many "resistance box" designs, you also have a considerable component of mass reactance in the "port" so a pure flow resistor is not always realistic in modelling (depending on chosen resistances as well), especially in the rolloff region of the rear source.
So finally "i am not convinced" ...
Before i could believe some curves coming out of a piece of software, i'd like to be convinced, that the modelling is adequate at least for the main components of a system: I am not seeking for "perfection" here.
Cheers
Last edited:
The measured impedance of the system was very close to the simulation so I just assumed the remaining parts of the model were accurate. I can confirm my next sim by measuring the model driver's actual excursion at power and freq using a laser we use at work for measuring mechanical stroke within .5mm
Either way, with what we've accomplished with DSP cardoid subs in professional applications and what I've experimented with as well as others, I do believe a passive cardoid system could extend uniform directivity below 400hz in a practical manner, without large drivers, baffles or horns.
Either way, with what we've accomplished with DSP cardoid subs in professional applications and what I've experimented with as well as others, I do believe a passive cardoid system could extend uniform directivity below 400hz in a practical manner, without large drivers, baffles or horns.
mayhem13 said:
@maiyhem13:
My intention was not just "being fussy" here ...
The "unibox" model for sealed cabs in insufficient to model a resistance box type cardioid even in a very simple type of design.
So even in case the "cone movement" related curves (excursion over frequency, impedance curve) of the driver happen to match (by accident) in some cases:
You don't have a radiation model accounting even for the dipole component.
So where is the input field for "effective dipole pathlength" in the "unibox" sheet, which is made just for modelling "moderate leakage" in a conventional sealed box ?
Without a - even simple - radiation model, you don't have any hints how to compensate for SPL dropping at LF.
As long as you have a model that incomplete, it is pretty useless IMO trying e.g. a laser interferometer just to prove that "isolated cone movement data" is close to reality sometimes ...
It is much like comparing "apples to oranges" IMO, because the incomplete model won't tell you anything e.g. about the cone movement you really need for flat SPL especially at LF.
Last edited:
agreed.......only FR measurements on and off axis to 180 degrees are going to give us the more relevant data. I'm just using the model for a rough approximation of the efficiency and excursion requirements needed to narrow down the field of midwoofers that would be effective.......not nearly as taxing as a dipole alignment
Hi Badman,
We use recycled polyester wool, such as this:
I've also had good results with rockwool and fiberglass, but those materials are possibly hazardous to your health.
We use recycled polyester wool, such as this:
I've also had good results with rockwool and fiberglass, but those materials are possibly hazardous to your health.
- Status
- Not open for further replies.