Beyond the Ariel

This goes back to the phasing plug sections and it was just easier to use straight walls on each section when the tooling is made than having to produce a curved surface, approximations of expansion rates were just accepted as good enough. Time has not changed these things, we still see the same methods used today that were used when Western Electric and others started the entire process, minor changes of materials, magnets and other parts have been the only real improvements to speak of. Today with modern CNC machinery there really is no excuse for the present situation.

Hello Kindhornman

Not all companies use phase plugs are straight sided walls. JBL has been using their Coherent Wave phase plug since the 90's starting in the 2446. Attached is a cross section of a driver that also uses it.

Rob:)
 

Attachments

  • EDS%20JBL%20476Mg_Page_02.jpg
    EDS%20JBL%20476Mg_Page_02.jpg
    142.7 KB · Views: 573
Last edited:
Rob,
Thanks for that, it's about time! With modern plastic molding techniques and even die casting these is no reason not to get this right.
myhrrhleine,
It wouldn't take much to knock the center section out of say a Radian driver if you wanted to experiment. One of the biggest problems I have with 3D printed plastics and why I can't use it in certain instances is that the heat distortion values of all but one of the materials is so low that they have such low melt points that they will fail under load at fairly pedestrian values. I am taking less than 2oo degree F, more like a max of 180F for short term use only. If you are only using say 1 watt input on the compression driver no problem but in many other instances the ambient temp can go high enough to destroy a 3D produced plastic part. The only alternative I know of today is the Ultem material but I have been quoted as high as $1,000 per pound for that material made into a model, not exactly a cost effective solution at that cost!
 
Rob,
Thanks for that, it's about time! With modern plastic molding techniques and even die casting these is no reason not to get this right.
myhrrhleine,
It wouldn't take much to knock the center section out of say a Radian driver if you wanted to experiment. One of the biggest problems I have with 3D printed plastics and why I can't use it in certain instances is that the heat distortion values of all but one of the materials is so low that they have such low melt points that they will fail under load at fairly pedestrian values. I am taking less than 2oo degree F, more like a max of 180F for short term use only. If you are only using say 1 watt input on the compression driver no problem but in many other instances the ambient temp can go high enough to destroy a 3D produced plastic part. The only alternative I know of today is the Ultem material but I have been quoted as high as $1,000 per pound for that material made into a model, not exactly a cost effective solution at that cost!

I would try a polyimide resin system with kevlar or carbon fibre with suitable fibre orientation. There are many many other possibilities. Yes it can be made into diaphragms. These firms play at materials science.There is nobody pressing them. If it was a war effort we would have made drivers and speakers which are better than the sound material they are reproducing. We came out of the war effort with tremendous potential. We ran out of steam after vinyl, with low grade CD players that could not get near to vinyl in reproducing what was recorded.It is only just catching up now as Bluray SACD and FLAC are trying to take over.
 
Boldname,
I am unaware of any solution impregnated polyimide/ carbon type material available but I haven't tried to keep up with the current thermoplastic composite materials available today, it would be very expensive if it is out there. Kevlar is another story all together, it is a very hard material to work with and does not bond well with most polymers. One of my areas of interest has been the development of new cone materials and I do use a combination of carbon and Kevlar, it is not a simple thing to do or to work with. I could make a composite diaphragm for a compression driver application but the precision needed to match weights and stiffness on that small scale for repeatability would be a real challenge, not saying it couldn't be done but it wouldn't be an easy task. You would be looking for no more than milligram differences before you would see output differences between devices. You would also need to have a way to consistently align the fiber orientation to get two identical parts with the same physical properties in a dome shape.
 
Last edited:
Hello Kindhornman

Not all companies use phase plugs are straight sided walls. JBL has been using their Coherent Wave phase plug since the 90's starting in the 2446. Attached is a cross section of a driver that also uses it.

Rob:)

As an aside, I can't see why the ends of the channels toward the diaphragm in these phase plugs have sharp edges. Given that the air movement has to negotiate a 90º turn before it can travel either way (compression or rarefaction) through the conduit, wouldn't radiused edges make sense to prevent needless turbulence? Is this just a manufacturing convenience?
 
Boldname,
I am unaware of any solution impregnated polyimide/ carbon type material available but I haven't tried to keep up with the current thermoplastic composite materials available today, it would be very expensive if it is out there. Kevlar is another story all together, it is a very hard material to work with and does not bond well with most polymers. One of my areas of interest has been the development of new cone materials and I do use a combination of carbon and Kevlar, it is not a simple thing to do or to work with. I could make a composite diaphragm for a compression driver application but the precision needed to match weights and stiffness on that small scale for repeatability would be a real challenge, not saying it couldn't be done but it wouldn't be an easy task. You would be looking for no more than milligram differences before you would see output differences between devices. You would also need to have a way to consistently align the fiber orientation to get two identical parts with the same physical properties in a dome shape.

Have you had samples of what is made for Boeing or GE just to make some trial diaphragms. I mean, perhaps I am naive but I took it for granted any speaker manufacturer would already have researched all the fabricating materials available from aerospace satellite industries. They cannot really all have been unsuitable. Do we know what diaphragms today are made from current state of art fabric and structural resin binders. You know I cannot understand why more use is not made of hard anodising with Al diaphragms or duplex or multiplex metal/non metal before getting to very costly and fussy pure beryllium. Something is missing.
 
Boldname
The materials we use in aerospace are basically carbon/epoxy matrix and are quite unsuitable for the purpose really. The density per unit area is high on purpose and to get where we are trying to go is in the opposite direction, we want very low mass with high initial stiffness but that doesn't have high Q resonances inherent in the material. Believe me you can make the carbon/epoxy so dense that it rings like a piece of metal. There are many different fiber types used for composites but none are really designed to work with a diaphragm having perfectly oriented fiber for this purpose. It can all be done it would just be very expensive and very few would have a clue how to do this on a production basis let alone at a high volume for production.
 
Russel,
Inadvertently you have answered your own question. Compressions drivers as we know them with high compression ratios, typical is 10:1 compression ratio, work on the rarefaction principal, that is what I have been told by a friend a compression driver designer. One of those simple principals that is most often overlooked when trying to understand these devices. These designs require extremely small air volume between the diaphragm and the front of the phasing plug to work, if it was only as simple as smoothing the edges of the annular rings that would be done.
 
I found the explanation here, Compression driver - Wikipedia, the free encyclopedia, immediately made sense.
In 1924 Hanna, C. R. and Slepian, J. [1] were the first to discuss the benefits of using a large radiating diaphragm with a horn of smaller throat area as a means of increasing the efficiency of horn loudspeaker drivers. They correctly surmised that this arrangement results in a significant increase in the radiation resistance (and therefore increased efficiency), because the loading mismatch between the vibrating transducer surface and air is largely corrected, thus allowing for much better energy transfer.
 
Just a thought Lynn, but have you consider these knock-down boxes (3cuft each) so that you could get a prototype up and running before committing to a permanent arrangement of the system?

Audio Knock-Down MDF 3.0 cu. ft. Subwoofer Cabinet for Dayton Audio 15" Ultimax Subwoofer

You could stack two each side for the dual GPA driver arrangement and place the subs elsewhere for the time being.

I like the prototype idea, especially since you have spent so much time contemplating various tradeoffs. This would allow you to atleast build and start measuring.
 
Just a thought Lynn, but have you consider these knock-down boxes (3cuft each) so that you could get a prototype up and running before committing to a permanent arrangement of the system?

Audio Knock-Down MDF 3.0 cu. ft. Subwoofer Cabinet for Dayton Audio 15" Ultimax Subwoofer

A pair of these might do fine for closed-box Altec/GPA 416's on the floor, and the upper driver in a to-be-determined format ... open-baffle, short horn (180~700 Hz), or whatever. I have a stash of four GPA 416 Alnico's (16 ohms), and a pair of GPA 515 Alnico's (16 ohms), so I have plenty to experiment with. The existing REL Strata II subwoofer will do just fine for evaluation purposes.

As for timing ... Karna and I have been deluged with offspring staying in our house, and it may be a while until we can induce them to move out. Our garage and basement are filled up with their junk (well, I think it's junk, but treasures to them). After the big move-out, Karna and I need to do some house updates ... it's getting on 10 years old, and the inevitable repairs and replacements await. New windows here and there, new washer/dryer, that kind of thing.

My Christmas present is a Marantz UD-7007, suggested by Gary Dahl, that serves as a CD/DVD-A (S/PDIF) transport for the 2-channel system, a direct analog-output source for 2-channel SACD's and high-res Blu-Ray discs, and an HDMI source for multichannel SACD's, DVD-A's, and Blu-Rays into the Marantz AV-8003 (which supports SACD over HDMI). It also has two HDMI outputs, so I can feed the Panasonic plasma display directly, and reserve the other output for the Marantz AV8003.

Since Blu-Ray is probably the end of the line for optical media, this should take of all physical-media digital sources. As for computer downloads, an Audio-GD DI-2014 takes care of translation chores from USB to S/PDIF, as well as reclocking S/PDIF to S/PDIF. A phonograph also awaits ... either the Technics SL-1210 I own already, or maybe the new turntable that my neighbor, Thom Mackris, is working on.
 
Last edited:
Member
Joined 2012
Paid Member
“DACless” Stereo Bass Management?


First, can someone completely define what is meant by stereo bass management?
Obviously, it’s not quite the same as the much more often discussed home theatre bass management (i.e. LFE), since home theatre implies movies, which implies multi-channel audio. My system build will be two channel stereo for music only, as follows:

My desktop pc will feed my DAC. Speakers are a pair of HF and MF direct radiator drivers, run by a First Watt J2 amp; and a pair of midbass drivers run by a First Watt F4 amp. The F4 amp is a voltage follower so a line stage with stereo attenuators must go between it and my DAC’s unbalanced outputs. And a passive stereo attenuator must go between the input of the J2 amp and my DAC’s other pair of unbalanced outputs. Because I’m not tri-amping my mains, I can’t use an active or passive line level crossover, so it has to be a speaker level type like this XM47 passive crossover for loudspeaker drivers Stereo balance for the mains is obviously achieved by adjusting the attenuators between the DAC and both amplifiers. When done, volume/mute control is had by the DAC’s wireless remote.

The sealed boxes for my GPA Altec 414 midbass drivers will cut their low end off below ~ 70Hz. Below this a pair of Rythmik 12 or 15” active servo subs take over. The subs’ plate amps have balanced differential I/Os for master/slave daisy chaining; the master sub’s plate amp’s unbalanced inputs are then connected to the line stage’s outputs. The input impedance of the plate amp and F4 are 30K ohms and 47k ohms, so I wouldn’t expect them to excessively load the line stage. Yes?

Thanks to the midbass drivers’ 2 or 3 cu. ft. (interior) sealed boxes, no crossover is needed between them and the subs. Instead, with the help of two others-one person at the controls of each sub’s plate amp-the sub’s low pass filters are adjusted to blend it with the midbass drivers. Thus, so achieved is one of the three (or more?) goals needed to successfully integrate the subs with the mains.

Time Alignment: If the subs are placed directly under the mains (and assuming their wiring relative to the amplifiers’ outputs are in phase), then all drivers should be time-aligned?

However, as the two subs will be placed where they ought to work best against room modes-nearly as well as four subs-regardless of listener position (
( http://www.harman.com/EN-US/OurCompany/Innovation/Documents/White Papers/multsubs.pdf , figures 48 & 50 ), delay adjustment between the subs and the mains is required. And that can be done with me listening, while two people adjust the sub’s delay (phase control)? http://www.rythmikaudio.com/images/XLR2.jpg


And the third obvious integration goal is to balance the volume of the two subs with that of the mains. Again, with the two people helping to set the volume between two identical subs and the mains shouldn’t be too difficult.

Please correct any misconceptions and/or omissions in my understanding.

This whole process may or not be a trivial pursuit, but what I want to know is if can I accomplish good “stereo bass management” or “bass integration” by using the above “manual” techniques and the subs’
plate amp’s analog control circuitry, instead of a DSP device-or at
least without leaving it running in my system. My music source DAC
will definitely exceed $4.5K. Therefore, I’d prefer to avoid having any further digital conversion in my system if at all possible.


 
Signal manipulation of any kind ... equalization, crossover high or lowpass, or time delay ... implies digital processing, or DSP. As used in the home theater world, "bass management" uses all of these techniques. If the HT receiver or pre/pro has bass management engaged, or Audyssey EQ, you will be hearing the sound of the built-in DACs. If the receiver or pre/pro is fed a 2 or 5-channel analog signal, it will be digitized through an inexpensive ADC (not studio grade), processed by the on-board DSP (with on-chip algorithms that are not studio quality), and converted back to analog with the built-in DACs.

Some receivers offer a "Pure Direct" mode that claims to bypass all bass management and EQ functions, and offers an all-analog signal path through the receiver. If that's really true, there will be no tone controls, no Audyssey, no bass management, and no subwoofer output. All you get is a volume control, and the same number of channels going in and out.

I should mention that is not possible to manipulate SACD or DSD signals; you can't even level-shift, much less the functions mentioned above. Professionals convert the 1-bit DSD signal to DXD (352.8/32 PCM) or DSD-Wide (2.8 or 5.6 MHz at 8-bit resolution) to apply the usual range of studio modifications, and then convert it back to 1-bit DSD for commercial release. This level of conversion is beyond the capabilities of consumer-grade home theater receivers; they automatically convert 1-bit DSD content to 176.4/24 PCM, apply the DSP processing, and leave it in PCM format for conversion with the inbuilt DACs.

So ... if you want to hear what an expensive external DAC sounds like, or hear SACD or DSD in native format, all signal processing by the receiver or pre/pro must be disabled, and hope that internal ADCs and digital signal processing have been removed from the signal path. If you have test gear, sending a 50 kHz square wave through the receiver should reveal the truth, since that will not survive the brickwall lowpass filter that an internal ADC requires.

Referring back to my previous post, when I listen to multichannel SACD's transmitted through HDMI to the receiver, I use "Pure Direct" mode to (hopefully) bypass all the PCM-based signal processing. As for the DACs inside the receiver, it's a good question whether the signal remains in 1-bit DSD format. Delta-sigma DACs live in a world halfway between straight PCM and low-bit, high-speed DSD signals.
 
Last edited:
With reference to "oltos" post, yes, what you want can be done with passive line-level methods (with RC filters). I would be careful not to load down the output of the DAC too badly; it'll tolerate 10K without trouble, maybe even 5K if it's a good solid-state design. The passive-filter approach also means you have to watch cable lengths; audiophile cables in particular have substantial capacitance with long cable runs. Even good solid-state gear becomes unhappy when it has to drive 500 pF or more of cable capacitance.

Gary Pimm found that low-quality electrolytics used for AC-coupling the inputs of mid-fi studio gear (like plate amps or EQs) also affected sound of the main channel, even though they were not in the direct signal path. A good-quality direct-coupled buffer stage (between the mid-fi and good system) takes care of the problem.
 
Last edited:
A pair of these might do fine for closed-box Altec/GPA 416's on the floor, and the upper driver in a to-be-determined format ... open-baffle, short horn (180~700 Hz), or whatever. I have a stash of four GPA 416 Alnico's (16 ohms), and a pair of GPA 515 Alnico's (16 ohms), so I have plenty to experiment with. The existing REL Strata II subwoofer will do just fine for evaluation purposes.

Don't know if this horn is short enough for you, but John Inlow just introduced a 135Hz midbass horn design that might work for you.

inlowsound.com
 
Beautiful sim results, but the 34" length for the 135 Hz midbass horn is a show-stopper.

The time-alignment of the current system puts the mouth of the AH425 is about 1/2" in front of the 416 frame (as shown in the drawing). Adding another 34" to the depth of the current system makes the total depth of a time-aligned version about 46" deep; no possible way can that fit anywhere in my house. Not interested in DSP processing.

If I go with the direct-radiator version, here's a drawing of what I've been thinking of, using a bentwood bass enclosure:
 

Attachments

  • Aspen_Ellipse.jpg
    Aspen_Ellipse.jpg
    155.6 KB · Views: 627
Last edited: