Hi All,
I'm picking up an old project. The previous owner stripped all the drivers from these horns, so I've been slowly regathering all the parts necessary to complete it.
The horn originally had an TAD 2001 installed on it, but a decade or so later, claimed these were NLA and told techs to replace with Celestion CDX1730. These two drivers have a different exit angles. 2001 has a reported angle of 8 degrees, where the CDX1730 has a 20.16 degree angle. How different will the end result be? Common sense is telling me just to go with the 1730, but I do prefer the BMS 4550 in general, or something newer from the 18Sound catalog. All of which have a lower angle than 20 degrees. The horn crosses at around 1.2K.
Just stick to the 1730? The TAD isn't an option for me as the price of them, even used and blown have gone stratispheric on the used market.
I'm picking up an old project. The previous owner stripped all the drivers from these horns, so I've been slowly regathering all the parts necessary to complete it.
The horn originally had an TAD 2001 installed on it, but a decade or so later, claimed these were NLA and told techs to replace with Celestion CDX1730. These two drivers have a different exit angles. 2001 has a reported angle of 8 degrees, where the CDX1730 has a 20.16 degree angle. How different will the end result be? Common sense is telling me just to go with the 1730, but I do prefer the BMS 4550 in general, or something newer from the 18Sound catalog. All of which have a lower angle than 20 degrees. The horn crosses at around 1.2K.
Just stick to the 1730? The TAD isn't an option for me as the price of them, even used and blown have gone stratispheric on the used market.
It will unload at a higher frequency, which affects all the BW from that point up, but without doing the complex math I assume you can sim both in Hornresp or similar to compare; but yeah, don't believe it's a big deal, so 'penny pincher' that I am would go with new and maybe have to raise the XO point a bit to compensate if needing high power.
It's not as though it will sound 'bad' once tuned in, but you may be missing out on some of how it could be. This kind of mismatch tends to produce higher order modes. On the other hand, some horns do that well enough without any help.
Hey Allen,
What do those sound like (HOMs), what is the threshold of audibility of HOMs vs. frequency, and how do amateurs like us measure HOMs? How do we know when the HOM fears are unfounded if you can't measure them, and you can't tell if you hear them?
I find it amazing that so much angst is spent on this recurring fear and so little effort is spent on quantifying and measuring it acoustically (not simulating it).
If you can't or won't measure the degree of the problem (i.e. by amateur diyAudio participants), isn't it a lot better to cease directing others from focusing on it, rather than offering diy platitudes about "what's good" and "what's not"?
Other than avoiding obvious horn design problems with using horns with throat slots above ~4 kHz (2" diameter throat), I think that HOM-fearing diy'ers (without HOM measurement means) are simply walking themselves away from good, usable horn/driver designs with this kind of "sage advice".
Chris
Last edited:
It's funny Chris, I was thinking the same thing when I decided to say it wouldn't sound bad.
I've spent a little time researching the problem/designing and building various examples and I know there is a difference.
I've spent a little time researching the problem/designing and building various examples and I know there is a difference.
Not good enough...
You really do need to answer some of the "rhetorical questions" I posed first.
I really do tire of the same old HOM "fearmongering" without direct (non-simulated) evidence.
Chris
You really do need to answer some of the "rhetorical questions" I posed first.
I really do tire of the same old HOM "fearmongering" without direct (non-simulated) evidence.
Chris
Sound would diffract there, at the sudden curvature change as throat angle changing suddenly. This is only for short wavelengths, similar in size as the physical features. In case of throat, this is typically about 1", so any issues would be around 1" wavelength, around 10kHz or so, which is problematic anyway with most compression drivers. For this reason, any issues likely go unnoticed.
What happens when sound diffracts? a secondary sound source forms, which emits sound back to where it came from, in opposite polarity. Exact same thing happens at baffle edge, or with any objects. The secondary sound emitted would then reflect back at the phase plug or at the diaphragm or something, and emit again, interfering with direct sound when observed further away, perhaps 5cm or 2" late or so depending where the sound reflects back, likely distance between diaphragm and throat interface times two as it's a round trip. Since response on the top octave is likely already full of anomalies this easily go unnoticed in measurements. 5cm delayed sound in opposite polarity would make destructive interference on 5cm wavelength, about at 7kHz, and combilter from there up.
Same thing with horn mouth, or any part of the device, where curvature changes sound emits back inside and reflects back. Now there is longer delay from mouth back to ~throat where it reflects, so delay of this secondary sound is greater so interference starts much lower in frequency, hence diffraction at mouth likely affects most of the device pass band. Secondary sounds bouncing back inside the device and back out again would make interference that changes to various directions.
If your throat were ~20kHz wavelength in diameter, has continuous curve from phase plug all the way to the mouth and over even all the way around the device, there is hardly any signs of interference = no any secondary sound sources = no HOM = no diffraction. Well, in a simulator zooming close enough there is some, like ~+/-0.1db, which means they are about -40db down as there likely always is some, sound would diffract all the way around the device although it ever emitted the secondary sound! Sound interacts with physical objects and we always have some until non-physical transducer is invented.
A typical commercial waveguide for a PA speaker for example typically have much more severe interference, also the drivers don't have good wavefronts exiting and so on, so the response is usually bit ragged showing issues like this. Bad reflection at mouth with a circular symmetrical device can make quite severe interference pattens on measurements. What is audible and what is not, well, I don't know that but it's worth it to pursue less interference as manufacturing such device costs the same as worse measured performance because difference is in the shape.
What happens when sound diffracts? a secondary sound source forms, which emits sound back to where it came from, in opposite polarity. Exact same thing happens at baffle edge, or with any objects. The secondary sound emitted would then reflect back at the phase plug or at the diaphragm or something, and emit again, interfering with direct sound when observed further away, perhaps 5cm or 2" late or so depending where the sound reflects back, likely distance between diaphragm and throat interface times two as it's a round trip. Since response on the top octave is likely already full of anomalies this easily go unnoticed in measurements. 5cm delayed sound in opposite polarity would make destructive interference on 5cm wavelength, about at 7kHz, and combilter from there up.
Same thing with horn mouth, or any part of the device, where curvature changes sound emits back inside and reflects back. Now there is longer delay from mouth back to ~throat where it reflects, so delay of this secondary sound is greater so interference starts much lower in frequency, hence diffraction at mouth likely affects most of the device pass band. Secondary sounds bouncing back inside the device and back out again would make interference that changes to various directions.
If your throat were ~20kHz wavelength in diameter, has continuous curve from phase plug all the way to the mouth and over even all the way around the device, there is hardly any signs of interference = no any secondary sound sources = no HOM = no diffraction. Well, in a simulator zooming close enough there is some, like ~+/-0.1db, which means they are about -40db down as there likely always is some, sound would diffract all the way around the device although it ever emitted the secondary sound! Sound interacts with physical objects and we always have some until non-physical transducer is invented.
A typical commercial waveguide for a PA speaker for example typically have much more severe interference, also the drivers don't have good wavefronts exiting and so on, so the response is usually bit ragged showing issues like this. Bad reflection at mouth with a circular symmetrical device can make quite severe interference pattens on measurements. What is audible and what is not, well, I don't know that but it's worth it to pursue less interference as manufacturing such device costs the same as worse measured performance because difference is in the shape.
Last edited:
I'm happy to answer a question that's been asked on the forum.
Chris, you're right that some might feel angst at knowing enough to know that something could be, but not actually knowing.. Many responses makes a good thread. If you'd be willing to coach me on how to avoid fear-mongering, I'm available by PM.
Measure it. Then measure the audibility of it. We don't talk about the "aether wind" anymore. There's a reason.
Chris
Chris, you might find this interesting as member TNT moves to larger waveguides which take more control of the wavefront.
Definitely encouraging.
- https://www.diyaudio.com/community/threads/son-of-hf-big-flower.421251/post-7877648
Definitely encouraging.
- https://www.diyaudio.com/community/threads/son-of-hf-big-flower.421251/post-7877648