XMAX

[Arjank]

Interesting read

Invalid Link Removed

 

[MuzikMan]

Man I am going to need about a 500 page primer for this one but I'll try to pick out what I can

 

[craig.p]

Fig. 19's nasty (disturbing?) d3 curve is VERY interesting.

"What price Xmax?"

 

[HolmeBass]

Very cool. Talk about the deep end of the pool. Unfortunately for us, in the early part of the paper he reveals a disturbing assertion: different speaker manufacturers determine Xmax, possibly our most revered parameter, with different methods rendering direct comparison between brands problematic. Furthermore, none of those methods (geometric or 10% total harmonic distortion or other measures of distortion) are very good!

Then things get complicated as he searches for a better way to compute/estimate/measure Xmax. I guess one of the big changes he suggests is not using just fs, the resonant freq of the driver, but rather using intermodulation components by using two tones to drive the harmonic distortion tests, a "bass tone" close to fs and a "voice tone" of higher freq. There is a cool table relating his various distortion parameters to actual physical causes. I agree fig 19 shows high values of 3rd order modulation distortion compared to relatively low values of THD, but can we really hear that? What does it sound like?

Now I am interested to hear what our resident speaker cabinet design experts have to say- a real problem, or not really that big of a deal when working on an actual speaker cabinet design? How sensitive are speaker box sizing computations to slight variations in the value of Xmax? Subscribed.

 

[greenboy]

Might as well tie this in to some of the other threads that have addressed this, it's not news. OldogNewTrick I think started one, maybe AlexClaber too, or at least both participated. Some of it was not that long back.

Until one starts actually measuring for large signal tests, the ears tell enough of the story. So rather than belaboring the thing, if a driver measured under this method or that method performs nicely for its purpose, I don't think too many of us are sweating the methodology here, because variance is typically not such a wide difference as to scuttle things.

It sure matters for Shoot-outs though ; }

 

[billfitzmaurice]

Not exactly new information. A few years back DuPont stopped providing Kapton to the industry for voice coil formers, forcing those using them to find new suppliers. Eminence was one, and since the new formers resulted in some T/S spec alterations they had to re-measure them. They used that opportunity to change over to Klippel Analysis. All of the xmax figures of their drivers since the introduction of the Deltalite II, and all drivers still in production that pre-date them, are Klippel measured. If Klippel changes the standard Eminence will follow suit.

With respect to electric bass high xmax isn't problematic, as we don't use very high xmax (over 10mm) drivers anyway. Most are between 4 and 6mm, and don't present a particular problem with respect to harmonic distortion. It's HT and PA subwoofers that can exhibit very high THD, for the reasons explained by Paul Klipsch a half century ago. His cure, the folded horn, remains the way to obtain the highest possible output with the lowest distortion of any speaker alignment.
http://www.readresearch.co.uk/loudspeaker_papers/klipsch_modulation_distortion_article_3.pdf

 

[seamonkey]

Not exactly new information. A few years back DuPont stopped providing Kapton to the industry for voice coil formers, forcing those using them to find new suppliers. Eminence was one, and since the new formers resulted in some T/S spec alterations they had to re-measure them. They used that opportunity to change over to Klippel Analysis. All of the xmax figures of their drivers since the introduction of the Deltalite II, and all drivers still in production that pre-date them, are Klippel measured. If Klippel changes the standard Eminence will follow suit.

With respect to electric bass high xmax isn't problematic, as we don't use very high xmax (over 10mm) drivers anyway. Most are between 4 and 6mm, and don't present a particular problem with respect to harmonic distortion. It's HT and PA subwoofers that can exhibit very high THD, for the reasons explained by Paul Klipsch a half century ago. His cure, the folded horn, remains the way to obtain the highest possible output with the lowest distortion of any speaker alignment.
http://www.readresearch.co.uk/loudspeaker_papers/klipsch_modulation_distortion_article_3.pdf

Since drivers are maybe going up in price - maybe we'll see a return to folded horns? Horns sure are loud and accurate. Just have to convince the manufacturers they still need to add mid-range drivers. They could put horns on those too.

Check out your own site for the best horn designs in the industry.

 

[billfitzmaurice]

Since drivers are maybe going up in price - maybe we'll see a return to folded horns?

They never went away, they still own the top end of the pro-touring sound market.
Sharp increases in driver prices do make horns, folded and otherwise, more attractive now in the lower price ranges. They dominated in the 60s before the advent of cheap watts, and now that driver prices are skyrocketing they may dominate again since you only need to use half as many of them for the same output.

 

[greenboy]

seamonkey, a lot of us like direct radiators for MI use. And we like a modicum of harmonic distortion. It's been part of the instrument voice forever, as well as in the playback systems. It actually does some things I prefer greatly, though I prefer not to have too much of it.

 

[knuckle_head]

It'd be interesting to know how/whether this affects sub-50 Hz . . . uuhhh . . . subs. I have seen so very few horn designs toward this end - folded or otherwise.

 

[craig.p]

Sharp increases in driver prices do make horns, folded and otherwise, more attractive now in the lower price ranges. They dominated in the 60s before the advent of cheap watts, and now that driver prices are skyrocketing they may dominate again since you only need to use half as many of them for the same output.

Which means less and/or smaller power amps.

Which means less current draw off the main panel.

Which means less chance of tripping that main breaker up top.

Which can mean the difference between being able to hit a room hard and being able only to be heard. (Or being able to play the room at all.)

It's all still relevant, and it's all still good.

 

[Arjank]

And we like a modicum of harmonic distortion.

I don't mind some 2nd order harmonic distortion, but 3rd order makes it ugly.
Instruments sound are based on even order harmonics, natural harmonics. A bad speaker though can add uneven harmonics, we don't want that. So, when you hit like 10% 3rd harmonic distortion things start to sound ugly.
When you base the xmax on the 10% distortion limit of the Fs of a speaker it can be 10% 2nd order or 10% 3rd order distortion. And that is not what most manufacturer mention. A speaker where the 2nd order overrules the 3rd order distortion will sound better.
Thats why I keep saying that manufacturers need to show harmonic distortion plots at different spl's.

 

[greenboy]

Have you actually analyzed musical instruments? I have. I used to do a lot of synth programming, some of it commercial. It ain't what you are saying there. I think you've hashed it up with some of the noncey balderdash found on some tube amp sites as well.

 

[craig.p]

I don't mind some 2nd order harmonic distortion, but 3rd order makes it ugly.

Yep, that's what sort of made me want to puke on my boots when I saw Fig. 19. Lots of 3rd even at relatively low excursions.

 

[Arjank]

I think you've hashed it up with some of the noncey balderdash found on some tube amp sites as well.

Nope

Ok, then use a sinewave generator and dail in 1khz, 2khz and 4 khz. How does that sound, better.
Now dail in 1khz, 3khz, and 5khz. How does that sound, ugly!

 

[seamonkey]

seamonkey, a lot of us like direct radiators for MI use. And we like a modicum of harmonic distortion. It's been part of the instrument voice forever, as well as in the playback systems. It actually does some things I prefer greatly, though I prefer not to have too much of it.

It' kind of the luck of the draw then if your direct radiator is going to give you the area of harmonic distortion you want. Everyone is going to be different depending on the drivers, and placement.

And if you get a perfect sound, then you have to mic the cab and hope the FOH can accurately reproduce the sound. Not add more distortion.

As one TB celebrity states
"If you can hear a difference, you can measure it, if you can measure a difference, you can model it."

Cabs are usually modeled with convolution (reverb). Dynamic Convolution. Take some impulse at different levels and run it through convolution. An accurate cab will reproduce this within human hearing capabilities. Convolution are showing up in pedals, and iPod apps.

Many well known colored cabinet have impulse response you can pick up to model them, but you have to have an accurate cab if you expect the modeling to work accurately.

Off topic from xmax - except horns are going to give you lower distortion levels.

 

[greenboy]

Here's a decent book on the subject of actual music instruments, with tons of FFT and 'scope examples of each instrument family in various ranges, as well as a crapload of text and references to the sample gathering for each instrument: A Synthesist's Guide to Acoustic Instruments by Howard Massey.

Like me and my partner back in the day, Massey had early mac digidesign product to gather the data samples. I've looked at lots of this stuff.

 

[OldogNewTrick]

Might as well tie this in to some of the other threads that have addressed this, it's not news. OldogNewTrick I think started one, maybe AlexClaber too, or at least both participated. Some of it was not that long back.

Yep... this one....

 

[greenboy]

horns are going to give you lower distortion levels.

Until you get up to a certain frequency anyway ; } ...Gotta laugh if you think saying "Horns!" and clicking your heels three times transports you to the land of audio perfection. That's why the constant push for better CD devices, truer waveguides less dependent on acoustic coupling efficiency and more interested in directivity control and accuracy and sound quality - and in recent years a lot of interest in HOM control - Higher Order Modes distortion. Google Earl Geddes (who Duke Lejeune has worked with, BTW).

 

[Arjank]

Gonna check that out, greenboy
I know you can get odd overtones on a stringed instruments when you hit it on the "wrong" spot.
But, odd order harmonics produced by a speaker just doesn't sound good.

Here's a decent book on the subject of actual music instruments, with tons of FFT and 'scope examples of each instrument family in various ranges, as well as a crapload of text and references to the sample gathering for each instrument: A Synthesist's Guide to Acoustic Instruments by Howard Massey.

Like me and my partner back in the day, Massey had early mac digidesign product to gather the data samples. I've looked at lots of this stuff.