Monster Massive

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
In this thread I'm going to take another stab at building a Synergy Home for my home. I've built some Unity horns for my car, but have never finished one for home use. I had another project that died on the vine about a year back, and this project will attempt to address some of it's shortcomings.

Here's some things I hope to accomplish with this project:

1) From my car projects, I've learned that using existing boundaries can allow one to make horns much much smaller than what would be "ideal." The typical solution is to use the corner of a home. I have tried corner horns, and I do not like the 'headphone' effect. Basically corner horns have no reflections, and that makes them sound a bit uninvolving, like a giant set of headphones. But there's another way to use boundaries which will be utilized in this project.

2) I really think that Geddes is on to something with his research that indicates that diffraction and higher order modes are two of the things that make horns sound bad. At the same time, these things can be problems in Synergy horns, because of the midrange taps and a lack of termination on the horn.
An externally hosted image should be here but it was not working when we last tested it.

For instance, the old Lambda Unity horns may suffer from diffraction at the edge of the horn, and in the horn throat at the midrange taps

3) The last thing is cosmetics; my reference speakers are currently banished to my bedroom due to their size. So I'm hoping that this project will have a smaller footprint and also a look that's avantgarde enough to make it back into my living room.
 
monster-massive1.jpg


Here's a picture of the proposed speaker.
It is based on the original Lambda Unity horn, but with a few improvements:

1) The main one is that the slope of the woofer enclosure can be used to reduce diffraction from the horn. You'll see this in a moment; basically with some manipulation of the horn shape we can create a seamless blend of the curve from the horn to the rest of the enclosure. A seamless transition is important because it reduces diffraction. Reducing diffraction makes the speaker more listenable, reduces horn honk, and makes it image better.

gradient.jpg

2) Another innovation is stolen from the Gradient Helsinki. I haven't heard many speakers that throw a bigger soundstage than the Helsinki, and I think part of their soundstaging ability is that the main lobe of the loudspeaker is aimed at the ceiling above the listener. By doing this, early reflections off of the floor and ceiling are delayed in time. Gradient does a better job of explaining this than I do, but the basic idea is that a regular speaker has a symmetrical floor and ceiling bounce in FRONT of the listener. The Helsinki has a ceiling bounce, but the ceiling bounce happens later due to the angle of the drivers.
So my Synergy horn does the same thing; the main lobe is aimed right at the ceiling, and we're basically listening off axis.
A side benefit of this is that it makes the speaker very short, which looks good, and it allows me to put loudspeakers beneath my home theater screen instead of behind it. By doing that I avoid the 'headphone' effect that happens with speakers behind a home theater screen.
 
monster-massive2.jpg


Here's the predicted response of the compression driver and midranges of this horn.
At the moment, the horn curve is LeCleach.

Normally I would avoid LeCleach because the beamwidth is very narrow. But what I've found from my experiments in diffraction reduction is that the beamwidth of the horn grows broader when you use a very very large roundover. For instance, when you put a 6" roundover on the horn there's a significant amount of energy that's radiated all the way out to even 180 degrees, even if the horn throat is only forty degrees. So we can use the roundover to both reduce diffraction and widen the directivity. Basically the way that this works is that diffraction off of an abrupt horn termination basically stops radiation at the edge of the horn mouth. But add a big roundover and the sound exiting the horn mouth is gently and slowly radiated into a wider beamwidth.
 
monster-massive3.jpg


Here's a picture illustrating what I mean, about using the enclosure shape to extend the horn curve.
In this case, there's a gi-gan-tic roundover at the mouth of the horn, a full six inches on the top *and the bottom.
The roundover shape mimics the existing curve of the LeCleach horn; basically the idea is to shorten the horn, then bond it to the enclosure so that the enclosure itself extends the curve.

1220588002.jpg

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


While it's true that I could simply build the LeCleach horn, and set it on top of the enclosure, I think that looks a bit clunky. I'm hoping to make this enclosure look sculptural while still being functional, similar to the Jadis Eurythmie and also the Synergy proposal that was posted earlier in this thread.

The big roundovers are made with plain ol' sonotube.

I would consider a roundover that's even larger, but Home Depot only goes up to 12" :)

monster-massive4.jpg


Here's a pic of the same speaker, but with transparency on the horn body, to better illustrate how the speaker enclosure extends the curve of the horn.
 
Here's a tentative summary of the frequency ranges, crossover slopes, drivers, etc

1) tweeter: Celestion CDX1-1425, 1500hz - 20khz
2) midrange: Dayton ND91 x 3, 180hz - 1500hz. This is the driver most likely to change, as I'm not sure if I can get them to cover a bandwidth that's this wide.
3) Woofer: Alpine SWS-10D2, 38hz-150hz. I'm going to run this off of my sub amp. I already have two subwoofers in the room, so this will basically act like a subwoofer that happens to be nearby the mains. The reason I can have the driver aimed off axis is that the wavelengths are so long at 150hz, it doesn't matter if it's facing forward or sideways.

I'm hoping to do first order slopes on the midrange and tweeter using MiniDSP; while this would normally be quite aggressive for such a small compression driver, I've found that the CDX1-1425 can take quite a bit of abuse and I've run it with first order slopes before. The efficiency on the horn doesn't hurt either.
 
I'm not really getting the ceiling reflection thing though. You're going to build highly directive speakers that mostly avoid the floor and ceiling reflection problems but you are going to aim them to get a reflection above you? Did you try it with the Summa's placed on the floor and aimed at the ceiling above you?

Seems to work for Gradient.
Their papers explain it better than I can - you really have to hear them.
Gradient admittedly uses a waveguide with a much broader pattern than mine.

Tipping the waveguide back also does a few other things:

1) The main thing is just making the speaker shorter. I have a 100' projection screen, so I'd really like to keep the loudspeakers below the screen
2) I've been screwing around with Beolab type lenses this year, and it's remarkable how far back you can tip a speaker, and it will still sound good. I don't like the sound of having a speaker firing straight up, but even 45 degrees backwards sounds quite good. (The Beolab uses a lens to redirect the very top octave into the forward lob.)

Haven't tried it with the Summas, but I should.
 
The pathlength from tweeter -ceiling-ear is the same with tilt or without. Tilting reduces floor reflections effectively (larger angle) but sidewall reflections are almost same. The relative proprotions of reflections from different surfaces changes.

Floor bounce is often the most annoying and it's cancellation F is usually at range of 200-400Hz Floor/Ceiling Reflection Calculator

Generally I like the idea of tilted midrange horn!
 
The pathlength from tweeter -ceiling-ear is the same with tilt or without. Tilting reduces floor reflections effectively (larger angle) but sidewall reflections are almost same. The relative proprotions of reflections from different surfaces changes.

Floor bounce is often the most annoying and it's cancellation F is usually at range of 200-400Hz Floor/Ceiling Reflection Calculator

Generally I like the idea of tilted midrange horn!

The gradient midrange is a cardioid, so that definitely has an effect too.

Here's a funny story about cardioids:

I've raved a couple of times about how amazing some of the prosound sub setups are these days. For instance, Excision toured with PK Sound last year, and I've never heard bass like that in my life. But nearly every show that I've been to in 2013 has had the same crummy bass that I hate. Just a dull undefined roar, with no 'kick' and no 'slam.' The shows I've been to in 2013 just sounded like a bunch of mush.

Just a short list of shows with mushy bass:
1) Mad Decent Block Party
2) Datsik
3) Flux Pavillion

But last Friday I saw Placebo at the Wiltern, and *boom*, the show brought the bass! Tight, defined, slammin' bass.

And lo and behold, what did I notice on stage?

Yup, a cardioid array.

I'm starting to think that's what basically spells the difference between crummy bass and great bass is just use a cardioid. It's odd, because cardioid arrays seem to be semi-ignored by the diyaudio set.

Both Bassnectar and Placebo were using the same cardioid setup, with an array of subs in front, and a second array behind the first.

Like this:

How To......Cardioid Sub Arrays - Speakerplans.com Forums - Page 1
 
Maybe you could get rid of diffraction at the midrange taps
by not having holes. Make the throat of foam or cardboard
with some bending mode radiators around the outside...

You said you were crossing with dsp anyway, so can you
blend the required mid delay without the phase shift of a
frustrum's cavity and taphole?
 
Last edited:
Maybe you could get rid of diffraction at the midrange taps
by not having holes. Make the throat of foam or cardboard
with some bending mode radiators around the outside...

You said you were crossing with dsp anyway, so can you
blend the required mid delay without the phase shift of a
frustrum's cavity and taphole?

I am going to deal with diffraction from the midrange taps by filling everything with foam. I have 30ppi reticulated foam but I may try higher densities also. We have so much midrange power to burn, I don't think it would be crazy to use foam that's even higher density than 30ppi if necessary.

BMRs would probably work too though. I heard the Linn speaker at CES with a big BMR and it sounded fantastic.
 
Here's a tentative summary of the frequency ranges, crossover slopes, drivers, etc

1) tweeter: Celestion CDX1-1425, 1500hz - 20khz
2) midrange: Dayton ND91 x 3, 180hz - 1500hz. This is the driver most likely to change, as I'm not sure if I can get them to cover a bandwidth that's this wide.

If you're going to use a mid with rubber surround, then you might want to look at the Faital 3FE25 in 16 ohms. Three of them can be wired in parallel and a 2.7 ohm series resistor can serve to bring them back to a nominal 8 ohms and give some attenuation to match the sensitivity of the compression driver. If you want to go bigger, then the 4FE35 is also available in 16 ohms.

FaitalPRO 3FE25 3" Professional Woofer 16 Ohm 294-1105
 
Regarding faital 3" drivers: the 4 ohm are superior EBP, sensitivity, No%..
I don't know why 16's are so much worse, I'm guessing dead weight of
excess coils outside the gap.

Patrick wants a silly high Q, and has his formula. But then complains
mids don't play high enough. I think time alignment should be in the
dsp crossover, and not by sandbagging mechanicals to make delay...

The 16ohm have higher Q and narrower EBP (given no cup on back
yet, they may not be all that different). but I think this is the wrong
direction to purposely go...
 
I will be getting 4 of the 4ohm Neos for a different experiment.
We can see then if the published specs are accurate. But just
look to the half space efficiency, Q, EBP. And I ask you why
the 4 is so different from 16? In both the 3 and 4 inch series,
both the ferrite and neo. Makes mis-measurement of the spec
by Faital seem unlikely...
 
Regarding faital 3" drivers: the 4 ohm are superior EBP, sensitivity, No%..
I don't know why 16's are so much worse, I'm guessing dead weight of
excess coils outside the gap.

Patrick wants a silly high Q, and has his formula. But then complains
mids don't play high enough. I think time alignment should be in the
dsp crossover, and not by sandbagging mechanicals to make delay...

The 16ohm have higher Q and narrower EBP (given no cup on back
yet, they may not be all that different). but I think this is the wrong
direction to purposely go...

Once you get those drivers on a Synergy horn, the EBP difference between them doesn't matter. Drivers with relative lower No% respond with higher horn gain over that of a higher No% driver. Saying there is a difference between the 4 ohm / 8 ohm / and 16 ohm model of the same driver would be like splitting hairs. Feel free to use what ever model gets you to the desired impedance. I prefer to wire drivers in a parallel when ever possible. You must always consider the effect of the other driver's Le on the response of the other drivers when you wire them in series. Sometimes it can help, sometimes it hurts you.
 
Since you are saying you want to use three mids, I'm guessing you are planning on a round horn with the mids mounted each at 120 degrees?

I am approximating the LeCleach curve by using square conical for the first half of the horn, mated to a monster roundover for the second half.

Here's how I'm doing it:

http://www.diyaudio.com/forums/multi-way/140190-jean-michel-lecleach-horns-158.html#post3676419

The only reason I'm considering three drivers is because I'm not certain I can fit four.
 
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.