Aethers: budget OB from Iowa 2007

[Undefinition]

I've wanted to put this up here for a while. Anyway, I see a bevy of dipole and OB projects around here, so it seems like a good place. I think Andy might have mentioned it here and there, but here's an "official" thread.

quick breakdown:
vented bass cab using Dayton DVC 8"
Open-back mids: Tang Band W5-704D
Tweeter: Seas 27TDFC
The XO is parallel. The woofer is 2nd order electrical, but everything else is first order electrical.

Originally, they were meant to be clones of Andy G's Blackwoods, using a linked series XO. Unfortunately, there just aren't any reasonably-priced bass drivers to pull it off in the USA. So I went for one of Dayton's great bang-for-the-buck bass drivers. This forced me to change the XO topology, and the design took on a life of its own.

Not to brag, but they did win the Budget competition at Iowa 2007, using Vifa P13 mids. As those mids became unobtanium, people asked for other possible mid drivers. Andy pointed out that the TB W5-704D would probably work well. Then, as luck would have it, a few weeks ago, Parts Express had a super $15 deal of the day on them. Now, they're back to their regular price of $23 or so... a good deal.

http://undefinition.googlepages.com/diy-aethers

There's everything you need to know. Enjoy.
Paul Carmody

An externally hosted image should be here but it was not working when we last tested it.

 

[NurEinTier]

Good looking speakers for a very small price.
Not tempted to add another woofer?
Or sub?

 

[F1 FAN]

Thanks for sharing this desgn Paul,at some point I want to try a dipole and your MTM section looks like a great place to start and cost is minimal.I would likely actively crossover to an already existant woofer section.I snagged one of those W5 704D's during the sale for use in a WMTW center channel.They certainly look to be well made and a great bargain.

 

[gtforme00]

Paul,

Clever design! It looks like you successfully bridged the gap between a typically complicated full dipole and the traditional box speaker. I may have to try this type of speaker out as a budget build for my brother.

Regards,
David

 

[Undefinition]

Good looking speakers for a very small price.
Not tempted to add another woofer?
Or sub?

Nah. Not that it wouldn't look more intimidating... but I'll just live with the sensitivity loss due to BSC. It would have also required a larger bass cabinet, and I was trying to keep things practical. The Dayton 8" DVC sub has quite a bit of extension; according to Unibox, the F3 is around 26 Hz. I tested it with sine waves, and this is about right. Unfortunately, I used a 2" vent, so with the sine waves, it was chuffing. With any sort of music, movie, or program material, I can't hear the chuffing.

 

[Andy G]

There is however no reason why you couldn't use a pair of 8" or a pair of 6.5" woofers. It all depends on what final balance you are looking for in terms of bass/mid tweeter. And as Paul say, it will be doing the BSC duty as well, so a few extra dB isn't going to hurt, unless you go too far.

You do not want to be padding down the bass !!!

 

[critofur]

Can you give some more subjective (and/or objective) comments/opinions about the Peerless buyout woofers (830341?) compared to the TangBand W5-704 woofers?

 

[Undefinition]

Can you give some more subjective (and/or objective) comments/opinions about the Peerless buyout woofers (830341?) compared to the TangBand W5-704 woofers?

In terms of the Aethers design? Well, if money is a concern, I'd go with the Peerless. They totally accomplish the design goal at a ridiculously low price. The only thing to contend with would be the pincushion frame. (Really, though, I'd like to see someone make a set of grilles for the Aethers. They'd look like a Mirage tower.) The Danish make good drivers, in my opinion, and aside from the stamped frame, this one is no exception.
The Tang Band W5-704 also accomplish the design goal. It took me a lot longer to GET them there because TB's FR graphs are so awful. Where the Vifa and Peerless versions just sort of "fell together" using mfr's graphs, I had to actually measure everything to get the TB version to work. But it certainly does work now, and the system integration is "just right" to my ears.

Sorry, but I couldn't really pick one driver over the other. And if you're asking for input about them in terms of OTHER designs, I'm probably even less help. I will reiterate, however, that TB's measurements are terrible, so you must measure yourself (or have someone else measure for you). But once you have an accurate FR graph for either of them, I think most people would find these to be VERY easy drivers to work with. Being poly cones, they MIGHT get muddy down in the low bass, but I've never run them as such, so that's just a wild guess at best.

 

[critofur]

I've already got about 6 of the Tang Band W5-704s from when they were on sale at $15.

I was thinking of picking up some of those Peerless drivers for a couple reasons: more efficient, a pair is about 8 ohms, higher power handling, and I like Peerless/Vifa stuff, particularly when it's not made in China (Just bias against Chinese products, I guess).

I was thinking of building some box MTMs and your Aethers to compare the difference. (I've never actually heard MTMs OR Open Baffle speakers (heard various other dipoles, such as my dad's Magnapan Tymapn 1Ds though).

I'm going to build some HT speakers for my friend, and a stereo pair of speakers for my living room. Wife will want speakers which go against the wall for downstairs, if I have Aethers I'll those'll have to be another pair that live in my decent sized 3rd floor room.

If you weren't considering cost OR appearance, (imagine you have some nice looking grills?) do you have a preference as to the sound of either the Peerless or TB drivers?

edit: Oh, and also compared to the Vifa drivers? Comparing all three only on sound quality alone...

I want speakers where you can FEEL the midrange vibrate you without making your ears hurt. Maybe it's the bass, and seeming that it's the low midrange is just an illusion? I like how a cello playing in the room makes the arms of your chair vibrate and you can really feel the energy all around you, in addition to what you hear with your ears, yet, you don't feel like you're loosing your hearing, to listen to something with that much musical energy.

 

[Undefinition]

I was thinking of picking up some of those Peerless drivers for a couple reasons: more efficient, a pair is about 8 ohms, higher power handling, and I like Peerless/Vifa stuff, particularly when it's not made in China (Just bias against Chinese products, I guess).

I don't think you'll run into problems with power handling. Because they just run midrange (200-2500 Hz), you won't have much issues with excursion. However, I honestly do understand what it's like to have an unexplainable desire to not use Chinese mfr'd products.

I was thinking of building some box MTMs and your Aethers to compare the difference. (I've never actually heard MTMs OR Open Baffle speakers (heard various other dipoles, such as my dad's Magnapan Tymapn 1Ds though).

Funny you should say that. There were two reasons I wanted to build a set of dipoles. 1. I heard a pair of NaO minis at a DIY meet and loved the "live" sound of them and 2. My dad has a pair of Martin Logan electrostatic dipoles which I've always been envious of.

I want speakers where you can FEEL the midrange vibrate you without making your ears hurt. Maybe it's the bass, and seeming that it's the low midrange is just an illusion? I like how a cello playing in the room makes the arms of your chair vibrate and you can really feel the energy all around you, in addition to what you hear with your ears, yet, you don't feel like you're loosing your hearing, to listen to something with that much musical energy.

I'm not exactly sure I know the sensation you're talking about, but this might be the right design for you, since it was basically tuned from the midrange out. I wanted male vocals, guitars, snare drums, cellos to sound REAL. If you measured them, you'd probably see a bump around 300 Hz. I tried a crossover where it was flat all the way up, but I found that I really preferred the speakers with the bump in there. As a musician, I felt it made the snare drums sound more like snare drums, the violas sound more like violas, the snozzberries taste like snozzberries... oops! :-D

In terms of comparing the three midrange drivers, it's hard at this point. But in a nutshell, the Vifas just had a killer midrange. They were just rock solid in the crucial 200-500 Hz area. I can understand why the P13 gets a reputation for being one of the best midrange drivers.
The Tang Band design will definitely provide the listener with the most detail (you know, the ability to pick things out of a recording, hearing room cues, etc). This is partly due to driver design and also due to my evolution as a crossover designer.
The Peerless fall somewhere in the middle of the two.

 

[Moondog55]

Hi paul, would you mind sharing the crossover you did for the Vifa version of this project??
I thought i had it but when i looked for it today i can't find it.
I will be using that quad of KEF B-139's in a sealed box as the bass units in this design as I can't be bothered at this stage building a TL.
Haven't made up my mind on the front and rear aspect of the design yet.
Regards Ted

 

[Undefinition]

Hi paul, would you mind sharing the crossover you did for the Vifa version of this project??
Regards Ted

Fixed it. Googlepages has been pretty good to me--especially for the price. But they are not being so nice about gifs anymore.
http://undefinition.googlepages.com/diy-aethers
By the way, if one wanted to increase the level of the mids and tweeter, you could increase the shunt resistors on those filters. You could go as high as 30-35 Ohms, depending on the sound you like and the room they're playing in.

 

[Moondog55]

Ok thanks Paul, I'll have to work up my own cross-over we have different versions of the driver

Mine are P13 WH-00-08

Description: 130 mm polycone bass / midrange with alloy frame. Very smooth roll off and neutral response.

Nominal impedance Zn (Ohm) 8 Reference voltage sensitivity (dB) 88
Voice coil diameter d (mm) 25
Voice coil length h (mm) 14
DC resistance Re (Ohm) 5.7 Voice coil inductance Le (mH) 0.7
Resonance frequency fs (Hz) 60
Mechanical Q factor Qms 1.38 Height of the gap hg (mm) 6
Electrical Q factor Qes 0.43 Total Q factor Qts 0.33
Long term power (IEC) (W) 40
Xmax peak (mm) 4
Moving mass Mms (g) 7.5 Nominal power (W)

Effective piston area Sd (cm²) 86
Equivalent volume Vas (ltrs) 10
Force factor BI (Tm) 6

 

[critofur]

Fun With PCD

Alright, here's the result of tonights fiddling w/PCD (note the phase alignment!):

 

[Andy G]

Hi paul, would you mind sharing the crossover you did for the Vifa version of this project??
I thought i had it but when i looked for it today i can't find it.
I will be using that quad of KEF B-139's in a sealed box as the bass units in this design as I can't be bothered at this stage building a TL.
Haven't made up my mind on the front and rear aspect of the design yet.
Regards Ted

Kef B139 should do nice bass for you ! The P13 WH should meet it well at around 200-220.
I would try to use a solid tweeter so you don't need to push the P13 much above say 2.4k. I have found they can get a bit overpowering if used much higher.

And as I think I've said before, I'd make the top baffle just a bit taller, and make sure the U-frame has a top part on it. I can take pics of the back of Blackwood if you are unsure what I mean.

Have fun !!

 

[Moondog55]

Hi Andy, thanks for all that info, now and earlier! Now that the wedding is over I can get back into it, i'm working on a cheaper 3.5 way at the moment just to get some idea of what I need to do for the "Eather" project, should have the experiment finished in a week or three ( busy time at work ) Will be going with the KEF drivers but my mind isn't yet made up on using 2 drivers front firing or geting some more and using them as Di-pole.
I think thats the correct term for drivers firing front and rear from a box, B-139's should be used in a TL but these will probably be sealed to get a smaller box with decent WAF, the TL is just too big for the room

 

[Undefinition]

Re: Two things

Alright, here's the result of tonights fiddling w/PCD (note the phase alignment!):

Critofur,
Two things:
1. You need to extract the minimum phase of the .FRD and .ZMA files for the drivers you are using. There are some tools at the FRD Consortium that can do this, as well as Jeff Bagby's response modeler. Until you do this, you cannot accurately model a crossover. That is why your phase looks so smooth... it's an estimation of the electrical phase inherent in the crossover design--but does not take into account the driver's phase.
When you trace the FR or Impedance curve of a driver, you don't have the phase--YET. You need to run the .FRD or .ZMA file through a minimum phase extractor: a very complex algorithm that figures out where the driver's phase is at each plot on the curve. You re-save the file, then bring that into your simulation software. YOu'll notice the driver's phase will be far more ragged; that's normal.
For instance, take a look at the Peerless version of the Aethers. These graphs were all traced from mfr's specs, then minimum phase extracted. Had I not extracted minimum phase, it would have all looked smooth--which is not accurate.

An externally hosted image should be here but it was not working when we last tested it.

2. Your design does not take into account baffle step loss. That is, on a normal 6-8" wide baffle, you will lose approx 6 dB as the frequency gets into the low hundreds of Hz.

An externally hosted image should be here but it was not working when we last tested it.

The only case this wouldn't be true is if this were an in-wall design. Otherwise, the speaker will sound very bass-shy. Try messing around with either Paul Verdone's Baffle Diffraction Simulator or (my personal favorite) Jeff Bagby's Response Modeler 2.0 to see what I mean.
You need to do this step (and the minimum phase extraction) before you take the files into PCD or any other XO design software.

Hope that helps.
-Paul

 

[Andy G]

Hi Andy, thanks for all that info, now and earlier! Now that the wedding is over I can get back into it, i'm working on a cheaper 3.5 way at the moment just to get some idea of what I need to do for the "Eather" project, should have the experiment finished in a week or three ( busy time at work ) Will be going with the KEF drivers but my mind isn't yet made up on using 2 drivers front firing or geting some more and using them as Di-pole.
I think thats the correct term for drivers firing front and rear from a box, B-139's should be used in a TL but these will probably be sealed to get a smaller box with decent WAF, the TL is just too big for the room

The TL for a B139 would certainly be too big for the Aethers design.

The other thing too watch is what spl the bass is relative to the mids, you want a bit to spare, otherwise you will have to pad the heck out of the mids and tweeter.
That's why I like the old Jaycar 10", sensitivity is 91dB and you can use a 30L box. If you can find a pair, they really are a great bass driver for the Aether/Blackwood style of speaker.

 

[critofur]

Critofur,
Two things:
1. You need to extract the minimum phase of the .FRD and .ZMA files for the drivers you are using. There are some tools at the FRD Consortium that can do this, as well as Jeff Bagby's response modeler. Until you do this, you cannot accurately model a crossover. That is why your phase looks so smooth...

Right after I posted that I was wondering why the phase wasn't included in the .FRD files you sent me and realized that was why it looked rediculously good in the picture I posted above...

It's easy to miss some things when you're newly learning speaker CAD tools at 2 AM...

2. Your design does not take into account baffle step loss. That is, on a normal 6-8" wide baffle, you will lose approx 6 dB as the frequency gets into the low hundreds of Hz.

That "design" is only 2 way, I was thinking of using the 8" woofer (which is not taken into acount there yet) for part of the BSC, and, haven't built/modeled a dipole yet - was thinking there'll be some difference in the BS w/a DP...

Also, running the midwoofers in parallel (w/ ~ 2.5 xover) instead of in series will give plenty of extra output for BSC - just have to see if I have an amp that doesn't mind impedences that low...

The main point was to see how the DC28F would matchup, which, of course, won't be complete until phase is added into the equation...

With the "Walsh" style designs I was working on earlier, BSC isn't really an issue, though, there are other issues unique to them.

Where is the BSC in your PCD model of the Peerless Aethers version posted above?

 

[Undefinition]

That "design" is only 2 way, I was thinking of using the 8" woofer (which is not taken into acount there yet) for part of the BSC, and, haven't built/modeled a dipole yet - was thinking there'll be some difference in the BS w/a DP...

When I was first designing the Aethers, I was told by very reputable sources that dipole roll-off would be an issue around 150-200 Hz. I tested the dipole section of the Aethers on its own, and I found that what they told me was dead-on.
I took this into consideration when choosing a crossover point between the woofer and mids.

Where is the BSC in your PCD model of the Peerless Aethers version posted above?

The Baffle Step is compensated for in the design. As you can see, the system has a sensitivity around 85 dB. The Dayton SD215 has an 89 dB sensitivity--however you must keep in mind that when they take those measurements, that's on a flat (sometimes infinite) baffle. Once you put that woofer in a cabinet of normal width, you lose 4-6 dB, depending on how far you are from the wall.
So, using simple math, I have about 4 dB of BSC in that version of the Aethers, same as the first version. They're really made for music, not special effects. With the 3rd version (the TB version), I put in a bit more BSC, so I'd estimate the sensitivity of that system to be 83-84 dB. That time, I thought I'd shoot for something that was just rock-solid all the way down into the 20 Hz range, and I think they hit that mark.

In other words, the COMPENSATION for Baffle Step loss is to attenuate frequencies above 200 Hz to bring them down to the now-reduced level of the woofer. There are many ways to contour the upper frequencies so that everything comes out flat to match the volumes the lowest frequencies of the woofer.