Beyond the Ariel - Page 346 - diyAudio
Go Back   Home > Forums > Loudspeakers > Multi-Way

Multi-Way Conventional loudspeakers with crossovers

Please consider donating to help us continue to serve you.

Ads on/off / Custom Title / More PMs / More album space / Advanced printing & mass image saving
Reply
 
Thread Tools Search this Thread
Old 30th April 2008, 12:29 AM   #3451
diyAudio Member
 
Lynn Olson's Avatar
 
Join Date: Jul 2004
Location: Northern Colorado
Well, that's what a CSD shows - a mechanically "perfect" driver would just have a narrow wall at the back of the display. Unfortunately, a CSD can be scrambled and rendered unreadable by an unsuppressed floor reflection, but I seem to get usable (not perfect, but good enough for my purposes) results with 20~30 dB of attenuation of the floor bounce.

My web page is a once-over-lightly on what you can expect to see when the impulse response is transformed to a CSD and FR plot. My favorite odd duck is the Eton CSD, shown below. I've never seen this behavior before - the obvious resonance (that long mountain range) is actually frequency-modulated! Remarkable! That shouldn't even be possible, if you stick to lumped-element theory. The real world is always a lot more complicated than our simplistic models - and the deeper you look, with more powerful measurement techniques, the more you find.

The original CSD was D.E.L. Shorter's chopped-sinewave method described in Wireless World in the early Sixties. I wish I had copies of these articles - they were as significant as Theile/Small theory, but were not appreciated in the USA when they came out. (That transatlantic accent, y'know, along with a good dose of NIH in the American speaker industry.) Shorter described what he was looking for as "buried resonance", since these were too narrow and too small to show up on a swept-sinewave display - and they were real, measurable, and clearly audible, not just an academic hand-waving hypothesis. These techniques formed the basis of more than a decade's worth of BBC monitors.

Shorter's method used a slowly swept sinewave generator with an external gate that counted the sinewaves and chopped them on and off with a user-selectable 4, 8, and 16 on-and-off duty cycle, with the provision of extending the "off" interval as long as desired. The scope is then triggered by the on-off signal pulses (a secondary output of the switcher). In practice, the apparatus requires a time-of-flight delay for the switching pulse for the scope trigger. The area of the scope display that is of interest, of course, is the "off" interval. As the sinewave generator is slowly swept across the spectrum, certain frequencies show high-Q, quite narrow resonances that can be only a few tens of Hz wide. With a sparse musical spectra, these are rarely stimulated, but with a dense spectra (like a chorus or an orchestra), they can be quite audible.

I used a system like this at Audionics in 1976 minus the time-of-flight delay, so it was only good for looking at the current going through the voice coil. Even so, I was able to clearly see not one, not two, but three narrow resonances in the KEF B110 in the vicinity of 3.2~3.5 kHz - as I recall, they were around 3.2, 3.3, and 3.4 kHz, and were about 30~50 Hz wide each. These were dustcap resonances, of course, but simply removing the dustcap didn't get rid of them - new ones appeared elsewhere. (You always gotta measure, not just assume!)

Similarly, the KEF B139 bass driver had a humongous resonance at 1.5 khz from the expanded foam diaphragm - and I was aware of no loudspeaker that used a notch filter to remove it. I found that with my first loudspeaker, the Audionics TLM-200, that it required a 3rd-order lowpass filter at 200 Hz to remove the midrange coloration from audibility - 2nd-order was not enough, at least with pink-noise test stimulus. Yes, the filter was expensive, but flip the switch, the coloration would obviously come and go.

Note that all of the measurements I'm speaking of here - CSD and Shorter's chopped-sinewave method - look at what happens when the electrical stimulation is cut off.

No relationship to BL product, cone mass, and rise time at all - these techniques look at the behavior of the mechanical system, which is what we care about most, since the correction techniques are so limited in scope.

Yes, you can use notch filters, but I can say from experience they don't completely remove the original coloration (despite what measurements may say), and there is a subtle residue of what I call "grayness" to the resulting sound. (Instrumental tone colors are subtly diminished after aggressive notch-filter equalization.) The best approach to remove these gremlins is to modify the mechanical system (or do the right thing and select another driver).

What truly surprises me is that design methodologies that became standard practice across the BBC (a world-famous organization that publishes in English!) during the late Sixties never found their way into the US high-efficiency manufacturers - Altec, JBL, Electro-Voice, and University. This seems like a country determinedly clinging to black-and-white TV just because their engineers couldn't do color, or more to the point, the surprising absence of plastic-based AC-biased magnetic-tape recording technology in the English-speaking world before WWII. (Prewar magnetic recording in the USA and UK was limited to steel wire and DC bias, with very low-fi results. Low-noise, hifi recording was limited to first-generation acetate masters.)
Attached Images
File Type: gif eton-wtr.gif (12.1 KB, 1268 views)
  Reply With Quote
Old 30th April 2008, 01:20 AM   #3452
diyAudio Member
 
auplater's Avatar
 
Join Date: Apr 2005
Location: KyOhWVa tristate
Default measuring 'colorations"

So you seem confident that CSD's allow you to correlate measured imperfections in the response of drivers to audible artifacts with a high level of confidence? Are there circumstances where the perceived quality of a driver may not correlate with it's measured response, and how would you resolve that conflict? And what sort of limits would you impose on a CSD wrt low level disturbances and their audibility?

Lots of questions, but I'm curious as to how one best characterizes drivers beyond your own hearing so that you can have confidence you're not fooling yourself..

thnx...

John L.
__________________
"...His brain is squirming like a toad..." Jim Morrison
  Reply With Quote
Old 30th April 2008, 01:23 AM   #3453
Salas is offline Salas  Greece
diyAudio Chief Moderator
 
Salas's Avatar
 
Join Date: Oct 2002
Location: Athens-Greece
Quote:
Originally posted by Lynn Olson
amplifier damping is really only effective at the lowest frequencies - it's not going to do very much for resonances at 1 kHz and above.
Will it still control a vc better up to 20k, hence the preference to an autotransformer vs a resistor if some HF unit has to be attenuated?
  Reply With Quote
Old 30th April 2008, 07:07 AM   #3454
diyAudio Member
 
Lynn Olson's Avatar
 
Join Date: Jul 2004
Location: Northern Colorado
Default Re: measuring 'colorations"

Quote:
Originally posted by auplater
So you seem confident that CSD's allow you to correlate measured imperfections in the response of drivers to audible artifacts with a high level of confidence? Are there circumstances where the perceived quality of a driver may not correlate with it's measured response, and how would you resolve that conflict? And what sort of limits would you impose on a CSD wrt low level disturbances and their audibility?

Lots of questions, but I'm curious as to how one best characterizes drivers beyond your own hearing so that you can have confidence you're not fooling yourself..

thnx...

John L.
Well, let's put it this way. I didn't come up with this protocol; the BBC did some forty-plus years ago. I've been to Britain, visited the BBC Research Labs, and heard their most advanced recording and playback systems. To this day that was some of the best sound I've ever heard. So I'm partial to the BBC/Spendor/Quad philosophy, both esthetically and sonically. Other people prefer other schools of design.

I don't think there's anything magical or mystical about speaker design; their "personality", if you will, simply reflects the priorities of that school of the design, the skill and tastes of the designer, and the attention to detail. That's really all there is to it.

Different schools of design sound different; this isn't surprising considering how bad loudspeakers are in objective terms, and designers are forced to trade one set of virtues against others. It's the same as cars: a Porsche 911, a Hummer H1, a Ford F250 with dual rear tires, a Subaru Outback, and a Prius all do different things, and no, you can't combine them into a "supercar."

The claim that any speaker can "do it all" is simply absurd - it usually reflects a young, just-starting-out designer who thinks they are smarter than anyone else in the world. I lost that illusion about thirty years ago - I'm just another worker in the vineyard, tending the grapes, and hoping for a good harvest.

As for sonic correlation, well, you win some, you lose some. I have pretty high confidence in time and CSD data, and I have plenty of company in the industry. But taking just one example, fairly audible diffraction effects don't show up that much in the CSD, yet if you walk around the speaker while playing pink-noise, you can hear the little secondary sources coming from the cabinet edges. And physically radiusing the edges and covering them with felt may measure the same (on the CSD), but they sound quite different in spatial-rendition terms.

Crossover caps are the most annoying of all; the measured differences in DA and DF are really really small, yet the subjective effect can be as large as physically replacing drivers. I have no explanation for that at all. I think JBL's idea of applying 9V DC battery bias to a pair of caps is a good one, though - anything to reduce this annoying coloration.

P.S. Sorry about the lack of info on the new project - I'm doing a lot of off-line research and co-ordination with other folks on compression drivers and horns. As for the comment a few days ago - "why no patents on plastic-film-diaphragm compression drivers?" Well, good question, but from what I can see, Altec and JBL didn't do a lot of research on these - this seems to be a European thing. So, for all of you eager beavers out there who can read more languages than I can, if you can dig up the original B&C, Beyma, and BMS patents, why not post them here?
  Reply With Quote
Old 30th April 2008, 04:24 PM   #3455
diyAudio Member
 
Graham Maynard's Avatar
 
Join Date: Dec 2003
Location: UK
This 15"er is snuck away in the Eminence range.

http://www.eminence.com/pdf/kappalite-3015lf.pdf

Reasonable price, so two in parallel might be useful for OB use ?

Cheers ........ Graham.
  Reply With Quote
Old 30th April 2008, 07:08 PM   #3456
diyAudio Member
 
auplater's Avatar
 
Join Date: Apr 2005
Location: KyOhWVa tristate
Default Re: Re: measuring 'colorations"

Quote:
Originally posted by Lynn Olson


<snip>

Crossover caps are the most annoying of all; the measured differences in DA and DF are really really small, yet the subjective effect can be as large as physically replacing drivers. I have no explanation for that at all. I think JBL's idea of applying 9V DC battery bias to a pair of caps is a good one, though - anything to reduce this annoying coloration.

Personally I believe the "sound" of caps may be due to unrecognized dimensions to the loss tangent and/or dielectric constant. Since the permittivity and the loss tangent assign numerical values to the charge capacity and "mobility" of the dipoles within the dielectric respectively, there may be anisotropic motion not clearly represented by these two values; i.e., "steric" hindrance in rotation or relaxation about one or more axes perhaps, depending on the material. I dunno.. if I were younger I might try figuring this out experimentally, I know the rf engineers i used to work with were all pretty touchy about parasitic capacitance...

thanx for the info.

John L.
__________________
"...His brain is squirming like a toad..." Jim Morrison
  Reply With Quote
Old 1st May 2008, 08:11 AM   #3457
diyAudio Member
 
Brett's Avatar
 
Join Date: Jan 2002
Default Re: Re: measuring 'colorations"

Quote:
Originally posted by Lynn Olson
P.S. Sorry about the lack of info on the new project - I'm doing a lot of off-line research and co-ordination with other folks on compression drivers and horns. As for the comment a few days ago - "why no patents on plastic-film-diaphragm compression drivers?" Well, good question, but from what I can see, Altec and JBL didn't do a lot of research on these - this seems to be a European thing. So, for all of you eager beavers out there who can read more languages than I can, if you can dig up the original B&C, Beyma, and BMS patents, why not post them here?
This is likely a response to my post. I wasn't asking you to do research, merely asking for any opinions and or insights into these diaphram materials you may have picked up over time as I find your comments often insightful and worthy of further investigation on my own. The post was merely an observation that all your comments related to metallic or phenolic diaphrams.

PS: JBL now buy in BMS CD's for some of their speakers.
  Reply With Quote
Old 1st May 2008, 09:25 PM   #3458
diyAudio Member
 
Lynn Olson's Avatar
 
Join Date: Jul 2004
Location: Northern Colorado
My comments on diaphragm materials aren't especially brilliant or profound. If we could take these compression drivers, unscrew the back cap, knock out the phase-plug with a mallet (Alexander's suggestion), and turn them around and listen to the naked diaphragm, what do we hear?

Well, even "small-format" domes are pretty big at 1.75", and traditional large-format diaphragms look downright huge at 2.88" and 4". Compared to direct-radiators, these are big domes. The suspensions vary from tangential, to diamond, to the same kind of thing we see in dome tweeters and midranges, a corrugated roll of compliant plastic or treated fabric.

Compression drivers don't get any exemptions from the laws of physics. The bigger the dome, the lower the breakup frequencies, and this is directly correlated to size. Damping techniques merely damp resonances, but don't do all that much to change the frequencies where breakup starts.

So if an audiophile metal dome that's 1-inch across starts to go into uncontrolled breakup in the 22~26 kHz region, we shouldn't expect wonders from a diaphragm made of the same materials, with similar suspension, that is three to four times bigger, and has nine to sixteen times the area. That's why I am so skeptical of large-format drivers operating up to 20 kHz - what direct-radiator of that size would be free from breakup? I can't think of any.
  Reply With Quote
Old 2nd May 2008, 02:35 AM   #3459
diyAudio Member
 
Join Date: May 2003
Location: Colorado
Quote:
Originally posted by Lynn Olson
So if an audiophile metal dome that's 1-inch across starts to go into uncontrolled breakup in the 22~26 kHz region, we shouldn't expect wonders from a diaphragm made of the same materials, with similar suspension, that is three to four times bigger, and has nine to sixteen times the area. That's why I am so skeptical of large-format drivers operating up to 20 kHz - what direct-radiator of that size would be free from breakup? I can't think of any.
Exactly right.

A 1" aluminum dome breaks up at roughly 25 kHz. A 2" aluminum dome breaks up at roughly half that.

So when you are looking at a 4" large format compression driver, the diaphragm is breaking up in the 5 - 6 kHz range. Past that and all you have are a myriad of resonances.

The smallest of the "small format" compression drivers have 1.75" domes. So if they are of aluminum or titanium, they will go up to around 15 kHz, best case. The only way around this is to use beryllium. That is why the TAD 2001 (designed by Bart Locanthi, one of the true audio geniuses) is so well regarded among certain cognoscenti -- it will go to around 22 kHz before breakup.

And you can use it down to around 1 kHz. So this is a good solution, as long as you don't care about money. The driver is over $1000 each, and a nice wooden horn is around $300 a pair just for starters.
  Reply With Quote
Old 2nd May 2008, 08:06 AM   #3460
jirka is offline jirka  Czech Republic
diyAudio Member
 
Join Date: Apr 2005
Location: Czech Republic
Send a message via AIM to jirka Send a message via MSN to jirka Send a message via Yahoo to jirka
I have been listening to the new Manger MSM C1 active studio monitor now. I cant imagine better speaker for nearfield listening at healthy levels. Simple pleasure machine. Razor-sharp portraits hanging in the air.
__________________
Jirka
  Reply With Quote

Reply


Hide this!Advertise here!
Thread Tools Search this Thread
Search this Thread:

Advanced Search

Posting Rules
You may not post new threads
You may not post replies
You may not post attachments
You may not edit your posts

BB code is On
Smilies are On
[IMG] code is On
HTML code is Off
Trackbacks are Off
Pingbacks are Off
Refbacks are Off



New To Site? Need Help?

All times are GMT. The time now is 10:27 PM.


vBulletin Optimisation provided by vB Optimise (Pro) - vBulletin Mods & Addons Copyright © 2014 DragonByte Technologies Ltd.
Copyright 1999-2014 diyAudio

Content Relevant URLs by vBSEO 3.3.2