Identifying Legitimately High Fidelity Loudspeaker: The Drivers

[admin]

This article focuses attention on loudspeaker drivers while continuing the series on Identifying Legitimately High Fidelity Loudspeakers. We take you on a tour of all the various aspects in driver design and the trade-offs associated with each type. We discuss why budget minded designs often employ stamped frame baskets as opposed to better but more costly cast frame alternatives. We also go into more detail on driver mechanics to gain a better understanding of what's inside the loudspeaker driver to make it work. Getting a peek at the guts inside a speaker system you are considering purchasing can tell you a lot about the budget allocated towards the drivers in the design. If the manufacturer doesn't supply such images, it doesn't hurt to ask. Better parts truly can yield better performance in the hands of a competent designer which are more common these days with the advent of inexpensive measurement equipment and knowledge of the basics in loudspeaker mechanics 101.


Discuss "Identifying Legitimately High Fidelity Loudspeaker: The Drivers" here. Read the article.

[shadyJ]

Terrific article! Very informative, thanks for writing it. So far in the series for "Identifying Legitimately High Fidelity Loudspeakers", we have articles on cabinets, cross overs, drivers, and cost-cutting, will there be more in this series? Drivers, cross overs, and cabinets are mostly what a loud speakers is composed of. Whatever the case, I enjoy these articles and am looking forward to more like them.

I have a question though, pertaining to the tweeters, what is the mechanical difference between the compression drivers used in high sensitivity speakers and normal tweeters. I take it it is a matter of design more than material?

[gene]

Terrific article! Very informative, thanks for writing it. So far in the series for "Identifying Legitimately High Fidelity Loudspeakers", we have articles on cabinets, cross overs, drivers, and cost-cutting, will there be more in this series? Drivers, cross overs, and cabinets are mostly what a loud speakers is composed of. Whatever the case, I enjoy these articles and am looking forward to more like them.

I have a question though, pertaining to the tweeters, what is the mechanical difference between the compression drivers used in high sensitivity speakers and normal tweeters. I take it it is a matter of design more than material?

thanks. As you could imagine, these articles take quite a lot of effort and time to write. Each time it went through a peer process from the folks I selectively chose to peer it based on their expertise, it caused me to refine and add stuff. Designers all have their own preferences, and biases (myself included). I tried my best to find a happy medium for all. I plan on writing 2 more articles in the series (one dealing with audio theories, and the other as a wrap up). Don't hold you're breath as I have a lot on my plate and I don't particularly look forward to writing about such a broad topic like this.

I didn't get too into detail on various tweeter types such as horns but the principles are still the same. A horn basically provides a passive method of providing a more efficient coupling path between the speaker and air. It's also useful in controlling both vertical and horizontal directivity. I may expand on this Driver article in the future or get one of our writers to do it for me if I can't find the energy to do so myself

[PENG]

thanks. Each time it went through a peer process from the folks I selectively chose to peer it based on their expertise, it caused me to refine and add stuff. Designers all have their own preferences, and biases (myself included). I tried my best to find a happy medium for all.

I think peer process is great, for reasons you cited. Sometimes we see forum members quoting things they read on the www as though they are facts while quite often those are opinions expressed/regurgitated by others, often mixed with some facts and many are naturally biased. Peer review process certain helps to present a more balanced view and minimize the chance of misunderstood, or over exaggerated "facts and opinions.

I plan on writing 2 more articles in the series (one dealing with audio theories, and the other as a wrap up). Don't hold you're breath as I have a lot on my plate and I don't particularly look forward to writing about such a broad topic like this.

I hope it will still be more on loudspeakers as aside from the quality of the media source itself they are the weak link of today's hifi equipment in my opinion.

[gtpsuper24]

Both drivers are 5.25" the one on the left is a more expensive speaker and the cast alloy driver on the right is a cheaper speakers. I think when some say theres no difference in stamped or cast, I think that has to do more with the driver technology not being good enough to take advantage of cast baskets. The top frame on the stamped basket actually bent as I was screwing it back in to place with just a hand screwdriver Not very good for a $500 bookshelf.

[slipperybidness]

Good article. I really would have liked some info on ribbon tweeters, planar transducers, maybe even electro-stats. Perhaps an update to the article down the line.

[GranteedEV]

Nice article. Would have liked a little bit more on flux modulation and also the benefits on custom-woven diaphragm materials, if you ever do update it

I have a question though, pertaining to the tweeters, what is the mechanical difference between the compression drivers used in high sensitivity speakers and normal tweeters. I take it it is a matter of design more than material?

Think of a compression tweeter, as being basically a large dome (or even a ring radiator) with an oversized magnet / voice coil with all the output being "compressed" into a smaller exit along with a phase plug (when I say phase plug, think of KEF's "tangerine waveguide", not seas' copper pokey thing)

Because a 1" exit compression tweeter uses a larger, probably 1.75" or 2" diaphragm, it's a lot more prone to audible issues from diaphragm material highlighted in this article. The few smaller diaphragm compression tweeters (IE 1.4" diaphragm) don't normally play well below ~1.8khz or so but might be cleaner in the upper treble; all else the same.

A 1.5" exit compression tweeter will often have a 3" diaphragm... Likewise, a 2" exit compression tweeter will probably use a 4" dome. As you probably can figure, a 4" dome is probably not going to be pistonic up to tweeter frequencies (with the exception of TAD or JBL's beryllium stuff). So these are best used with a crossover to a super tweeter! But having one driver cover 300hz to 8khz, does have its advantages.

The horn in front of the compression tweeter will be a big contributor in terms of impulse response and sound power response. There're horns optimized for high SPLs, and others that provide little "boost" but plenty of directivity control.

Basically - compression drivers are probably more sensitive than 1" domes to the material used simply because they're bigger. what's more, many are made strictly for pro audio use, so diaphragm durability can be favored over sound quality. Custom versions of these drivers for home use IMO is a lot of untapped potential.

[GranteedEV]

Gene, I just read the last page on tweeters, and I'm a bit unclear on something.

However, there are no free lunches and the designer needs to be mindful that doing so does NOT lower the resonance frequency of the tweeter. Thus, care must be taken NOT to lower the crossover point within the region of the drivers Fs.

The tweeter resonance frequency is a critical metric in determining where to set a loudspeaker’s crossover frequency.

It seems to me that the argument being made here, is that the resonant frequency in and of itself, is "bad".

As far as I can tell, a tweeter that is 8db down in level at resonance acoustically, would "physically" only be asked to be "18db" down assuming, say, the waveguide/horn adds 10db of acoustic gain. So you're not really compromising power handling or linearity.

It seems to me, that the resonance frequency, assuming the driver has a sufficiently low Q at resonance, does not contribute negatively in such a manner that "avoiding it" is the goal. I think using the tweeter fs to determine crossover frequency is flawed... the fs is almost an arbritrary number as far as I see, as far as the final loudspeaker is concerned. I'd say the article itself showed how you can "artificially" lower fs of a sub-par tweeter. While there's some correlation, i think it's a red herring to imply causation of the idea that fs is itself the source of distortion.

[gene]

Gene, I just read the last page on tweeters, and I'm a bit unclear on something.



It seems to me that the argument being made here, is that the resonant frequency in and of itself, is "bad".

As far as I can tell, a tweeter that is 8db down in level at resonance acoustically, would "physically" only be asked to be "18db" down assuming, say, the waveguide/horn adds 10db of acoustic gain. So you're not really compromising power handling or linearity.

It seems to me, that the resonance frequency, assuming the driver has a sufficiently low Q at resonance, does not contribute negatively in such a manner that "avoiding it" is the goal. I think using the tweeter fs to determine crossover frequency is flawed... the fs is almost an arbritrary number as far as I see, as far as the final loudspeaker is concerned. I'd say the article itself showed how you can "artificially" lower fs of a sub-par tweeter. While there's some correlation, i think it's a red herring to imply causation of the idea that fs is itself the source of distortion.

Every major loudspeaker engineer that has peer reviewed this article have all agreed that you typically want to crossover a tweeter one octave above Fs. Using a shallow horn does add efficiency, but it doesn't change Fs so you would still have to cross it over pretty close to the same frequency you would have had to without the horn loading. I am sure there are designers successfully using the Fs bump as part of their crossover design, but now you're operating a tweeter down to its lowest usable limit and potentially straining it and increasing distortion while lowering power handling.

This tweeter for example is arguably one of the best 1" domes ever made and used in countless high end designs:
http://www.scan-speak.dk/datasheet/pdf/d2905-950000.pdf

It's Fs is 550Hz and nobody would ever use it down that low. Most speakers I've seen implementing this tweeter, cross it over at 2kHz (2 octaves above Fs).

[AmFuzzy]

Every major loudspeaker engineer that has peer reviewed this article have all agreed that you typically want to crossover a tweeter one octave above Fs. Using a shallow horn does add efficiency, but it doesn't change Fs so you would still have to cross it over pretty close to the same frequency you would have had to without the horn loading. I am sure there are designers successfully using the Fs bump as part of their crossover design, but now you're operating a tweeter down to its lowest usable limit and potentially straining it and increasing distortion while lowering power handling.

This tweeter for example is arguably one of the best 1" domes ever made and used in countless high end designs:


It's Fs is 550Hz and nobody would ever use it down that low. Most speakers I've seen implementing this tweeter, cross it over at 2kHz (2 octaves above Fs).

Thanks for pointing out Scan-Speak. I have a couple of speakers using this tweeter and it's phenomenal. I'm amazed at how many "high-end" speaker companies use inferior actual drivers and use bizarre cabinet designs to offset their inability to build proper drivers or buy from Scan-Speak or Vifa. I'd love to see real world tests of drivers in some of these so called "high-end" speakers. Driver vs. driver capabilities. I know crossovers and cabinet design all come to play here...but with simpler designs and high quality drivers can a more budget speaker compete with some of these ridiculous high priced speakers? Here's a good example of quality without the price. norh_com/Norh_Loudspeaker/Norh_9.html