Planning for my some-day reference system

[RockLeeEV]

Disclaimer: I don't know much about speaker building, and I don't plan on building this anytime soon for financial reasons, but I want to get started on some general plans to make a modified design based on an RBH T-30LSE. Why rip it off? because it looks amazing; but it's limited edition! By the time I ever have money to buy one of those, they'll be long gone! If I ever do DIY this, then I can't make cost an object. I mean if a pair of T-30LSEs retail for around 15000 a pair, i guess that can be the "budget" for a 3-speaker front stage, give or take.

By modified, the differences I'm leaning towards are

-taller, wider cabinet, maybe like the T-3 configuration over T-2 configuration. I still plan to use the same height from the ground for all the drivers, just add some extra cabinet space, and maybe passive radiators up there. The dimensions of the T3 are

91.75" H
13.25" W
18" D

and the dimensions of the T30LSE are

60.5" H
15.5" W
18" D

So given these factors, I think the dimensions for my box, with 3/4 HDF as the cabinet material, will likely be somewhere around

92" H
16" W
20" D

-Dual 12" subwoofer drivers instead of a pair of 10s, and maybe passive radiators instead of a port - with the larger cabinet i think this could work

-a front stage of 3 matched speakers instead of a L/R pair + Horizontal center (note: I'm not sure how to go about this, since the L/R are a mirror pair.


General configuration

Tweeters - The T30-LSE uses a trio of scanspeak 9500 lined up on the inner baffle. Hey, it's DIY, why not go with a step-up from that!? 9x[3 per speaker] ScanSpeak D2905/9700 in a truncated line array (ahem "dispersion averaging array). Who knows, I might even go with the scanspeak revelator, although the price for 9(possibly 15 if you factor in surrounds) of those sounds like they're overkill - is it?

M/T Crossover:

I'm leaning towards using a close or same crossover design to what the T30lSE already does, which means I need a woofer that'll blend really well around the 2.2 khz range). I'm also considering active crossovers and bi (er... tri?) amping.

What should I look into? The key is being well spaced, and with HQ parts all over the place for max linearity. How was DIY crossover making the first time you did it?

MidWoofer:

12x [4 per speaker] 6.5" woofer with genuine phase plugs. I don't neccessarily need the aluminum that RBH used in theirs, but comparable speed, power, precision, and power handling are a must. What kind of impedance woofers should i be looking at in this scenario (let's assume the amplifier to be a trio of Emotiva XPA-1, one for each speaker's tweeter/mid sections. Or should I go with a built in class D amplification?)

M/W crossover

On the T-30LSE, there is no m/w crossover - it basically relies on the natural rolloffs but I'll probably go with a true crossover of some sort. maybe active.

SubWoofer section

I'm also unsure on this section. A couple general thoughts i've got include:

-Would a subwoofer section on the center compromise the sound? Most people seem to go with a horizontal T-1 to pair with the surrounds which makes sense in a lot of situations, but in my planned scenario it'll be behind an acoustically transparent screen

4x or 6x 12 inch subwoofer drivers (2 per speaker)
8x or 12x Passive Radiators at 12" (4 per speaker)

I'd have to model some drivers to see which 12 would work best. Undistorted extension down to 15 hz maybe, with a HPF at around 12 hz? Or is this simply too low?

And powered either by a plate-amp or pro-amp.

Again, the biggest difference between my DIY and the real thing is probably ported vs PR. I'm not sure if that's the way i'll ultimately go but it's just the way i'm leaning.

So on that note, give me a ton of recommendations for woofers to consider among the other questions i had. I have all the time in the world to make changes, this is more of like a dream project or something and it probably won't be my first speaker build by that time either on that note. I just want to have it written down what i'm aiming for. Any suggested design modifications are welcome as well including other tweeters i might want to look into. One other idea i've got floating around up there is a transmission line instead of a port or PR.

[vinylkid58]

Originally Posted by RockLeeEV

Disclaimer: I don't know much about speaker building, and I don't plan on building this anytime soon for financial reasons, but I want to get started on some general plans to make a modified design based on an RBH T-30LSE. Why rip it off? because it looks amazing; but it's limited edition! By the time I ever have money to buy one of those, they'll be long gone! If I ever do DIY this, then I can't make cost an object. I mean if a pair of T-30LSEs retail for around 15000 a pair, i guess that can be the "budget" for a 3-speaker front stage, give or take.

Have you heard them? My guess is probably not.

Just because something looks amazing, doesn't mean you should build it.

jeff

[InclinedPlane]

Open and shut case eh vinyl? The guy is obviously got the interest and must have spent an hour or more on the post there. He'd benefit from some constructive advice; ya know kind of like what this site is for. But if a one liner shrug-off is called for, I guess you know best.

[vinylkid58]

Ok, how about starting with something you could actually lift by yourself. Try cloning a Von Schweikert UniField 3.

All those drivers in a 160 pound box rendered me speechless.

jeff

[planet10]

With a speaker like that this thread should be in the multi-way forum.



One would need to listen to make a proper judgement, but based on appearances one can make some general inferences.

0/ a big huge monkey coffin with lots of shiny drivers. That has a definite appeal to a certain part of the market.
1/ 3 widely spaced tweeters. If you use multiple tweeters, they should be VERY closely spaced, and as close to filling the space between floor & ceiling as possible. A distant listening position is required to get integration. It also needs to be accompanied by floor-to ceiling mids.
2/ optimum bass usually requires the (sub)woofers sited in a different position than the higher frequency components.
3/ given that you have 2 big (sub)woofers XOed at 80 Hz, below which everything is omni, one could get big advantage with mounting them push-push... this would really help with problems caused by ...
4/ a speaker this big made with 3/4" HDF and with minimal bracing and not well executed
5/ a passive 80 Hz XO... with the components needed for an XO this low, and all the issues with XOing where the driver impedances look like a roller coaster, it would probably be cheaper and certainly sound way better with active woofers.
6/ even with just a single tweeter the array of mid-basses is too widely spaced to support a cross-over as high as is used.

The way to approach this kind of project is piecemeal. Look at it as 2 powered subwoofers + 2 sattilites.

Since you are financially constrained, you start small and work up. What you learn during the journey will inform where you get in the end. And i bet it won't be where you think you'd like to be today. We all learn so much on the journey.

dave

[mayhem13]

While i agree with MOST of what you've posted Dave, a clever designer COULD pull off the top section with 2.5 way topology on the mid drivers and power tapering in the way of series wiring on the top and bottom tweeters. Lots of XO components and design time but it just might work.......really well.

Passively working in those two sub drivers.....that's not so good.

Either way, i wouldn't want to have to move those around at all!

[vinylkid58]

Originally Posted by InclinedPlane

The guy is obviously got the interest and must have spent an hour or more on the post there. He'd benefit from some constructive advice.

You're right. My apologies to RockLeeEV. It was a knee-jerk reaction, and I shouldn't criticize someone for something that doesn't appeal to me personally.

jeff

[Lynn Olson]

First, you need to clarify your goals: do you want good sound, good looks, or to impress your friends? This might sound facetious, but it's a serious question. What do you want? If you don't know, you're unlikely to get it.

What sounds good to you - based not on looks, but extended listening sessions with your favorite music - might come from some rather unlikely systems. The more you listen, the more you should be ready for preconceptions to be destroyed. I often hear speakers at hifi shows that sound much better than they should, based on looks. And I also hear many speakers that have great reputations and top reviews, and sound just dreadful.

Second, I have to warn you: the complexity, difficulty, time, and expense is a square (or even cube) law of the number of crossovers in the speaker system. Translated to English, a full-range speaker is the easiest, consisting of driver selection and the right enclosure (and you could do a lot worse than check out the Planet10 site for some very interesting designs).

It takes anywhere from several months to several years to master designing a two-way system, depending on how good you want the final result to be. This requires getting some kind of PC-based measurement system, by the way. Off-the-shelf (or "textbook") crossovers do not work. You have to measure, and learn how to make an accurate and repeatable measurements. If you are not willing to do this, stick to kits or pre-designed systems with good reputations. If you go this route, build them exactly as specified - most important of all, do not swap drivers from the specified models.

Three-way systems: This is where the deep waters begin. I started designing speakers in 1975, and never really liked any of the 3-way systems I designed. My current system under design is essentially an augmented 2-way, with a subwoofer and a supertweeter. I find three-ways to be very, very hard to design - the big problem is a lack of coherency. By that, I mean the system sounds OK at one volume level, but the sound disintegrates when it is quiet or when it goes loud, drawing attention to all the different drivers, instead of sounding like a single, large driver. Just getting a three-way to sound as good as a single-driver table radio at quiet listening levels is not a trivial task - very few succeed.

Four-way systems: Speaking only for myself, I've yet to hear a four-way system that sounds coherent. They must exist - but I haven't heard one yet. Going the full-active route, with a stack of power amplifiers, a flexible active crossover, and a good measurement system is probably the only way to make this work. This is so ambitious that some people give up speaker design entirely after attempting a 4-way project.

Moving on, a passive crossover at 80 Hz is a very bad idea, particularly with the great choice of self-powered subwoofers we have today. Power amplifiers are not at their best below 100 Hz, and putting a crossover in that region creates ripples in the impedance that creates a very difficult load for nearly any power amplifier. If the impedance dips are severe enough (4 ohms or lower), it can threaten the survival of the power amplifier. You shouldn't have to buy a 1-kilowatt arc-welder amplifier just to compensate for poor crossover design. (A rather famous speaker, well-reviewed in the audiophile magazines, has just this design error, with an impedance dipping down to 2.5 ohms in the 80 Hz region.)

A self-powered subwoofer solves several problems at once; it relieves the main amps of delivering power at the lower edge of their working range, and removes the need for a very expensive, very large, and very heavy passive crossover. (Large inductors and large caps cost big money - more than the drivers, which is a hint there is a conceptual design error.)

[Lynn Olson]

Don't mean to sound negative, but crossovers have a steep learning curve. There are many well-reviewed audiophile speakers selling at quite high prices that have badly-designed crossovers, resulting in uneven frequency response, high driver distortion, and ragged polar patterns. A good crossover, by contrast, controls driver excursion (which lowers distortion), has flat response at the listening position, and smooth control of off-axis frequency response.

Referring to this forum, a single-driver system gets you from build to good sound the quickest way possible. Coherency is never an issue, but heavy-metal or full-concert sound levels require a sophisticated approach. The folks on this forum are here to provide guidance; this is probably one of the best sites in the world for single-driver speakers.

If you choose to go to the dark side and go for a multiway system, there is of course the "multi-way" forum. The hardest task isn't driver selection, even though that's the most fun part, but the hard slog of crossover design, diving into the theory of filter slopes, phase shifts, driver spacing, and the effect on polar pattern. The waters are pretty deep here, and the rarity of well-designed speakers, even at the high end of the industry, is a testament to the difficulty involved. Consumers don't look at the ugly crossover hidden deep inside the speaker cabinet, but it's the difference between a bad speaker and a good one.

These days, there are very powerful digital crossovers, combined with measurement systems, that let you twiddle with all kinds of crossovers, and hear the result for yourself in real-time. You do need a stack of amplifiers, but the power requirements are much more modest in a system with active crossovers, since each amplifier is only handling a limited frequency range, and clipping in one frequency range is much less audible than a full-range amplifier. 60 watts per driver is plenty in the real world, unless it's a PA system.

[a.wayne]

Lynn,

Which brand digital x-over would you recommend for such an endeavor?


regards,