A few questions about speaker design

[Consul]

Okay, really, what I'm interested in is no-compromise speaker design and building.

I have seen two schools of thought when it comes to making a set of near-field monitors:

1) passive crossovers with single amp per channel, and
2) active crossovers biamping (or triamping) to each speaker.

I guess the question is, in your opinion (as in, whoever is reading this), which is the way YOU would go?

I don't really like the idea of building amps into the speakers. For amps, I was thinking of going with nothing but discrete class A designs (like Nelson Pass discusses on his pages). I suppose I could build four amps and an external active crossover all in one box to go out to the speakers, though.

Truthfully, I'm too new to this to even think about attempting this kind of a project, but hey, if you're gonna start, start big. I do know my fair share of physics and electronics (and can wield a soldering iron to good effect), but speaker design is an area I'm just starting to explore.

If you wanted a no-compromise design, from the amps to the drivers, how would you do it?

[barefoot]

If you wanted a no-compromise design, from the amps to the drivers, how would you do it?

I would do it somewhat like you stated:

".... build four amps and an external active crossover"

They don't necessarily all need to be in the same cassis though.

Electronics and physics are great to know (I'm a physicist), but it takes a lot of practice and failure to master the art of loudspeaker design. I wouldn't recommend such an adventurous and expensive project until you've successfully designed several less complex systems. Although, you might find a designer who could help you with your project.

Thomas

[Consul]

Practice and failure are two things I am well-accustomed to.

I suppose I could make the design I have in mind, and try a set as a pair of hi-fi speakers to gain some confidence. I just started a new job in a new town, so I need to get myself established a bit, then it's on to DIY pro audio-ville for me.

I'll have to study some crossover designs and class-a amps designs and see if anything strikes my fancy. Thank you for the reply.

[barefoot]

Colorado Springs? You don't happen to work at Intel do you?

Standard electronic crossover designs won't do you much good. You have to account for the frequency and phase response of the drivers themselves. So, you never wind up designing a simple Linkwitz-Riley crossover, for example. And you can fool around with exotic Gyrator circuits and such, but I've never found any sonic advantage over a straightforward Sallen-Key topography.

As far as components, I personally like to the use the AD797 opamp. It's one of the quietest and most linear opamps ever made. But I guess you could go way out and use something like a Jensen 990C. The specs are somewhat worse, but it has that cache of being all discrete.

Thomas

[Consul]

Actually, I just moved to Montrose from Colorado Springs. Thanks for reminding me to change that. I never worked for Intel, though. I now work for the County of Montrose as their Linux admin.

And I don't know any of the terms you just threw at me (except for Linkwitz-Riley), so I think it's Google time.

So I guess the first thing I need to do is select which drivers I want to use. There are a ton of options out there, and no real way to know for sure which ones will work well. The only clues I have to work with involve the basic types the expensive pro companies use. Avalon seems to like the ceramic woofers, for example. I was looking at aluminum drivers myself, but I keep finding most of them designed for ported speakers, and I want to do a closed-box design.

Time to do more research, I guess. Thanks for your input, yet again.

[barefoot]

And I don't know any of the terms you just threw at me (except for Linkwitz-Riley), so I think it's Google time.

You definitely have a long way to go then. First thing to do is pick a copy of "The Loudspeaker Design Cookbook" by Vance Dickason. This will get your feet wet. You can find it and other good books at audioXpress.

Another thing is, I don't see how anyone can even begin to design a modern high performance loudspeaker without design software and a measurement system. You really need a design package like LEAP, Calsod, or LspCAD. I use LEAP, plus I even supplement it with some of my own algorithms written in Matlab and Excel. A good measurement system is also a must. MLSSA and LMS are the standard tools these days, but there are less expensive systems like Praxis and CLIO.

Thomas

[Consul]

Then that may be going too far for what I'm actually after, then. What I want is to build a good set of nearfield studio monitors. I only asked about modern no-compromise designs to see if it might be possible.

How about this: A closed-box speaker with two elements: an aluminum mid-woofer and a ribbon tweeter. Build and tune up a good passive crossover network and then build a nice, clean class-a amp in front of it all. I'm really just after something that will perform somewhat better than the current crop of inexpensive monitors on the market, like the M-Audios or the Events. I think it can be done.

Well, thanks a lot for your help. I think the next step will definitely need to be selecting drivers.

[Consul]

Actually, I found this just recently:

http://www.speakerbuilding.com/content/1044/

The mid-woofer/tweeter half of this might make for a good near-field monitor design. Then I could combine it with a good subwoofer.

Any thoughts?

[Consul]

I jusy had an odd idea. I don't know if it could be done, but I figure it doesn't hurt to ask.

What if you built your speaker elements (woofer, tweeter, etc.) straight into the soffits and bypassed having a cabinet? Think "open-baffle speakers" which are soffit-mounted. Then you stuff fibreglass behind the soffits and call it good. (It would look like the speaker elements themselves were just screwed into the wall.)

Could it work?

[barefoot]

It could work, but you would need to find drivers that are optimally designed for "free air" applications. This has the disadvantage of putting the entire alignment burden on the driver suspension itself - i.e. no "air spring" from the cabinet. The advantage would be that you could end up with a more efficient system, which might yield lower power compression.

Thomas

[Consul]

(snip...) This has the disadvantage of putting the entire alignment burden on the driver suspension itself - i.e. no "air spring" from the cabinet.

This sounds like a mechanical issue with the drivers themselves. Could it be possible to modify the chosen drivers to align themselves properly?

At least, I hope I understand what you're saying here...

[barefoot]

No, it's not a "problem" per se. It's a potential design weakness. Most drivers rely on a significant level of "acoustic suspension" to control the cone motion, whether it's the air spring in a closed box or the reactance in a ported box. Without this, the driver suspension must supply all of this control. And with the inherent limitations of material properties this could perhaps be more prone to degradation over a shorter time period. I'm not saying it can't be overcome. I'm just pointing it out as a potential design issue.

Thomas

[Consul]

Well, people are making open-baffle speakers, which implies that there is a solution to the issue. I'll look into it a little further and see what I find.

[barefoot]

Morel might be fruitful ground here. Their drivers tend to have fairly high Qts (> 0.5). Putting one of these in an infinite baffle would yield a well-damped, but not over-damped alignment. You may also need multiple units, since you'll have fairly significant excursion requirements with a single driver.

Thomas

[Consul]

Thanks for the information. I'll have to go take a look at what they have to offer.

I think I'm going to seek out some of those books you recommend. That'll give me a better basis for discussion. The problem is, my local library is small and has zero speaker design books, so I may have to venture an hour away to Grand Junction to find them.