A Progress Report (Audio Tech Talk)

 by Bryan Geyer

Today, at a time when so many of us have experienced the audible benefit of properly implemented subwoofers, it seems archaic to assign a single driver as the sole bass source in a serious full-range loudspeaker. The spread from the bottom of the low bass passband (20Hz to 80Hz) to the top of the mid-bass passband (80Hz to 400Hz) is simply too wide for one transducer to handle, and the logical way to address that limitation is to split the 20Hz to 400Hz span, and apply separate power amplifiers to drive separate loudspeakers. Each of the assigned drivers can then be optimized as needed, with stiff, long-throw, piston-like air pumps for the ultra-low bass, and flatter, faster woofers for the middle bass. Implicit differences in efficiency would be of no concern because, with each leg driven by a captive amplifier, the respective levels can easily be balanced—or not—as desired. Overall bass levels could readily be adjusted to compensate for prevailing acoustics, as well as for the whim of the listener.

Giant single-driver bass design was initially popularized at the 1939-’40 New York World’s Fair, but we can do better today, and make everything look less cluttered. Compact class D power amplifiers, utilizing multi-layer boards and surface-mount components, can now be buried inside the loudspeaker enclosure. More efficient (also more precise, also less expensive) high impedance analog crossover* and equalization networks can be tailored to extract peak performance from a designated driver, and then blend smoothly with the ensuing upper bass and treble stage. Of course, these same desirable assets can also be implemented by digital means, and that cost-effective alternative is very popular today, despite the need to introduce yet another cycle of analog-to-digital and digital-to-analog conversion. This further complexity is avoided with the all-analog approach, so I personally favor pure analog simplicity, where less is more. But my personal bias is warped by the fact that I can’t test and verify what’s happening when the digital processing is implemented. Indeed, some of the digital simulation jargon seems artificially contrived**, but I’m an “old school” technician in every sense. I don’t own the kind of instruments (or the smarts) needed to appraise digital manipulation of an analog signal.

So who’s at the forefront of this improvement trend? Who is leading the charge to provide electronically augmented loudspeakers? Well, one niche that quickly embraced the integration concept is the mini-monitor makers. They commonly integrate customized power amplifiers with their speakers, but forego the lowest bass. That’s implicit with desktop expectations. The companies that produce hi-end full range loudspeaker systems can best mirror these improvements. Some already offer floorstanders with integrated subwoofers and multiple internal amplifiers. Some utilize digital signal processing (DSP); others might stick with classic analog design. Staffer Bill Heck has previously reported on his upscale speaker system from Legacy Audio; he uses it with his Wavelet 2 digital processor and says the sound is sensational. (So is the the price.) Editor Karl Nehring also favors hi-end Legacy Audio products. It’s obvious that this trend is active and evolving. More entrants should follow, maybe with models for modest budgets and smaller rooms. 

Stay tuned—check reliable reviewers, e.g., sites like Audio Science Review (https://www.audiosciencereview.com/forum/index.php) and Audioholics (https://www.audioholics.com/) — and stash away some savings; Nirvana may be near!

------------------------------------------------------------------------------------------------------------------------------------------------

*A Linkwitz–Riley filter ( https://en.wikipedia.org/wiki/Linkwitz–Riley_filter) that provides -24dB/octave attenuation is not particularly complex or costly, but tight precision is essential to ensure that the critical - 6dB down locus coincides precisely for both (low-pass and high-pass) of the filter sections, and that it is positioned at the desired crossover frequency. In practice, this is best achieved by “cherry-picking” the critical R/C components, so production gets messy when accuracy is vital. Digital control apps will tend to artificially obscure such production inaccuracies, but the output will reflect the full extent of any error.

**A DSP control application that I once monitored responded to my inputs by simulating graphic bar charts to mimic the impact of each command, just as if reporting a test result. Pure pseudo-science.

Addendum (by KN):

Bryan raises some important issues here that are well worth consideration. Indeed, it is a really tall order to ask a single driver to handle the frequency spectrum ranging all the way from deep bass up through the midrange – something’s gotta give. As he points out, one viable alternative approach is a good pair of two-way speakers augmented by one or better yet a pair of carefully matched subwoofers. There are many subwoofer manufacturers who offer models with not only built-in amplification and control settings, but also apps that allow tailoring of the system setting through your mobile phone or tablet.

 

Bill Heck and I have taken another route, opting for large multi-way loudspeaker systems that assign the lowest frequencies to what are in effect built-in subwoofer systems (a pair of 10” drivers crossed over at 180Hz in each channel of Bill’s system, a pair of 12” woofers crossed over at 120Hz in mine). To clarify, neither Bill’s Legacy Signature SEs nor my Focus SEs employ any internal digital amplification or crossover circuitry; they are both standard passive speakers with normal analog crossovers. Bill does employ the Wavelet 2 DAC /Preamp / Processor; however, that is a device entirely separate from the speaker. Moreover, it can be used with virtually any speaker, not just those from Legacy Audio. You can learn more about the Wavelet 2 in Bill’s review, which was in two parts: Part 1 is here and Part 2 is here.

 

Finally, I’d like to mention that there are several loudspeaker manufacturers that are employing advances in both digital amplification and signal processing to offer loudspeaker systems with amazing capabilities. One such example is the British manufacturer KEF, who offers several speaker models that are not only powered, but also incorporate Bluetooth, wi-fi, streaming services, etc. For example, the KEF LS60 Wireless (pictured) is triamped (100 watts Class AB for the tweeter, 100 watts Class D for the midrange, 500 watts for the woofers). It incorporates a 0.75” tweeter coincidentally mounted inside a 4” midrange driver, plus four 5.25” woofers (two on each side of the cabinet). The speakers can be connected together wirelessly or wired for higher-resolution connectivity. Each speaker has an RCA socket for connect an external powered subwoofer. There is an available KEF app that can be used for all manner of control and tailoring for frequency response, room settings, listener profiles. etc. You can also control the LS60s with Roon or Apple AirPlay – it’s a whole new world, folks. 

I’m still amazed that I can walk into my listening room, sit down in my listening chair, open my iPad, pull up Amazon Music or Qobuz, and control my NAD C658 streamer/preamp right from where I sit and listen to hi-res music through my stereo system with so little fuss. Or that I can prop my feet up in the recliner in my living room, turn on the TV, grab the Roku remote, and watch a performance by the Frankfurt Symphony Orchestra or Belle and Sebastian whenever I feel the urge. Yep, it’s a whole new world…

On LED Flashlights…

By Bryan Geyer

The innate connection between a flashlight and “Audio Tech Talk” becomes apparent as soon as you try to inspect any part of your sound system that lies beyond the front panel. In recognition, here’s some personal guidance about the most vital tool you’ll ever need: A modern single-cell LED flashlight.

First, learn why the light emitting diode (LED), a semiconductor device that dates to the late 1960s, didn’t become abundant in flashlights until some 30 years later. What a curious delay! What happened? And what changed? Well, a modern “white light” LED won’t fire until it sees a threshold bias of about 3.2 volts minimum, and 4 volts or more works better. But a fresh single cell battery produces only ~ 1.2 to 1.7 volts, dependent on its chemistry (see footnote*), so multiple batteries would be needed to fire a solitary white light LED. Clearly, using more batteries than before didn’t advance the state of flashlight design, so miniature incandescent lamps endured. Then, barely before year 2k arrived, integrated circuit “boost chips” appeared. These new silicon ICs could efficiently elevate the energy from a single storage cell to the operating level needed to fire a white light LED. The basic design concept was old, but implementing it in IC form was breakthrough genius. The entire boost function, consisting of the surface-mount chip plus some tiny passive peripherals, could now be packed into a bundle smaller than the size of two stacked dimes. A practical single AA cell LED flashlight was suddenly viable, and its launch assured immense advantage over anything using an incandescent bulb. A lone LED could now provide more light (with selectable lower levels), draw less current, offer high resistance to shock and vibration, and assure decades (not hours) of maintenance-free service life—all while selling at competitive mass-market prices.

Product development was intense, with many players engaged. Some failed, some flourished, and several smart trans-Pacific sources became dominant. There’s now a wide array of LED flashlights and multi-LED lanterns on the market. Many designs serve niche applications, but my intent here is to feature a basic everyday flashlight—a modestly priced model that’s appropriate for general purpose use and practical in-pocket carry. A flashlight that…

…is short (≤ 3.1" long), skinny (≤ 0.75" diameter), and uses only a single AA size cell.

…is blindingly bright at full output, with multiple lower levels to conserve current drain.

…has long throw capability, with enough side spill to see the borders of your path at night.

…is obvious in operation, with a single butt-end control button for all switching functions.

…offers instinctive battery exchange: unscrew head and replace AA cell, button end up.

…has solid mechanical integrity; built to withstand the rigors of moderate abuse.

…if supplied with a pocket clip, such clip must be readily removable and leave no evidence.

… meets the IP68 waterproof standard (up to 30 minutes at depths of ≤ 2 meters).

…is cheap enough stash extras (toolboxes/cars/bikes/garage) and give as gifts.

There aren’t a lot of candidates that meet all of my objectives. Many come close, then fail the “side spill” criteria. That’s because flashlight marketers always hype throw (range), and inexperienced buyers focus on that feature; they haven’t yet grasped the importance of spill (beam spread). The two parameters (throw and spill) are mutually exclusive; more of one means less of the other. A very tight beam will deliver impressive reach, but hot spot fixation invites personal risk when you can’t see the edges of a narrow path from just one step away.

My pick as a good “everyday” single AA cell flashlight is the Fenix E12 Version 2.0. It meets all of my criteria, and you can buy it here: https://www.fenix-store.com, or here: https://www.batteryjunction.com The Fenix E12 V2.0 flashlight offers three selectable outputs: 5, 30, and 160 Lumens. It uses a cool white LED, and has an ultra-thin (3mm) optical front lens—all very desirable. The lowest level is appropriate for checking your watch in a darkened theater or finding a keyhole at night. The mid-level setting, 30 Lumens, assures a practical balance between output and capacity. It’s great for all of your “fix it” tasks, also for walking at night. The full 160 Lumen output provides a brilliant spotlight, but keep it brief! Apply only in 10 to 20 second transient bursts. At this level, supply current is consumed at a rapid rate, and there’s significant internal dissipation (heating). As with many LED flashlights, these output levels are all electronically regulated, so the lighting intensity stays constant; it won’t droop as the AA cell ages. Full expiration becomes suddenly apparent; it’s dependent on the drain that’s required for the level you’ve selected. When the supply current is insufficient for a given level, the output will drop to the next lower level. When it’s too low to maintain the regulated 5 Lumen minimum, the output drops to a useless glow, then goes out.

The Fenix E12 V2.0 flashlight comes with an attached (by compression) pocket clip that I dislike, so I trash it. This is easy to do with the aid of pliers—just grab the clip and pull it off. It will readily detach and leave no evidence that it was once mounted on the flashlight body. This flashlight is so small and light that a pocket clip isn’t of practical benefit; however, removing the clip will then expose a perfectly cylindrical body, and round stuff that’s not restrained can roll off flat surfaces. If this annoys you, consider mating a wrist strap to a rear lug. Straps that have a small pigtail loop at the end will pass through the lug without impeding the control button.

BATTERY OPTIONS: AA cells are readily available in the variants (3) noted below*, but you’d do well to reject the use of alkaline (manganese dioxide) LR6 cells. These conventional cells are very popular, but too risky to merit informed consideration. When an alkaline AA cell approaches expiration (output ≤ 0.7 volts), it will progressively destabilize and form a caustic compound (KOH, potassium hydroxide) that will penetrate the outer shell and attack anything that’s within reach. The consequent damage can be severe. This is an inherent flaw—all alkaline cells present this same inevitable hazard—so don’t use alkaline AA cells. Instead, select either of the other options. I use the non-rechargeable lithium-iron disulfide Class FR6 cells, e.g., Energizer’s “Ultimate”**. Others prefer the rechargeable nickel metal-hydride (NiMH) Class HR6 cells. Both options are safe, don’t leak, and work well. Non-rechargeable lithium AA cells exhibit extended operating life, perform well in cold ambients (alkaline cells don’t), weigh about 40% less than the alkaline equivalent, and can tolerate long term (10 years) passive storage with minimal (5%) net self-discharge. Buy in bulk, stow for use as needed.

* The nominal AA cell output voltage, when fresh (or freshly charged), is…

…1.5 volt for a non-rechargeable alkaline (manganese dioxide) IEC Class LR6 cell.

…1.2 volt for a rechargeable nickel metal-hydride IEC Class HR6 cell.

…1.6 to 1.7 volt for a non-rechargeable lithium-iron disulfide IEC Class FR6 cell, e.g., Energizer “Ultimate” AA cells.

IEC = International Electrotechnical Commission

 ** I buy on-line, from Battery Junction; refer… https://www.batteryjunction.com/energizer-aa-ln91-24pk-battery.html

BG (December 2022)

My “Battery” Advice…

By Bryan Geyer

Re. “battery,” let’s nail the definition first. A single AA or AAA voltaic storage cell is actually not a “battery”; it’s a cell. A “battery” is a combination of two or more cells. So the sign at your local hardware store that directs you to the Energizer and Duracell AA and AAA “battery” display is technically in error. (That’s excusable, but the prices they post are not. You can do lots better at…https://www.batteryjunction.com.)

Yes, I realize that ramblings about AA and AAA (size) DC storage cells is distinctly outside my assigned specialty (audio nitty-gritty) here, but this is my 30th tech paper for this site, so host John Puccio (he of the chocolate digressions) has kindly allowed me some off-topic slack. The outcome of this liberty might well be more pertinent for you than anything in my previous 29 posts, so I say hey, listen up—this is stuff you need to know!

I’m an especially heavy user of both AA and AAA voltaic storage cells. However, I have not purchased any standard alkaline (LR6 class) AA or AAA cells since the late 1970s. Instead, I use only lithium (FR6 class) non-rechargeable AA and AAA cells (aka L91 & L92 cells). I do so because these lithium iron disulfide (LiFeS2) cells present a stack of compelling advantages over archaic alkaline (manganese dioxide) cells, not the least of which is the fact that their extended life generally makes them cost-competitive in all but a few benign applications. The service life advantage of a lithium AA cell—when compared to its alkaline equivalent—varies, dependent on how it will be utilized. In very low drain service (e.g., a TV remote control) the lithium battery will outlast an alkaline cell, but not by much. However, lithium longevity becomes appreciable when used in higher drain service; e.g. in a flashlight or lantern, or in photo strobes, RF transmitters, and in some portable audio equipment. There’s a real 2X to 3X service life advantage, as compared to an alkaline equivalent, in a great many common applications. For example, I’m seriously into photo-macrography (close-ups) and also long lens wildlife photography (mostly birds). These endeavors often involve using two or more electronic strobes that can support rapid and repetitive high-power bursts, and the related current drain is normally more than wimpy alkaline cells can sustain. This is especially so in colder weather because alkaline cell output declines rapidly when ambient temperatures approach +40˚F, whereas lithium AA or AAA cells will still retain most of their native room-temp. punch.

While long service life is nice, maybe more important is the fact that lithium L91 AA cells will NEVER LEAK, even when discharged and left in your equipment for ≥ 5 years. (Been there, done that.) Conversely, all alkaline cells (of any brand) will eventually expel their noxious corrosive electrolyte (it’s KOH, potassium hydroxide) when they’re at or near a depleted state.

Further, lithium AA cells are ~ 40% lighter than alkaline equivalents, based on size AA comparison. That’s a helpful benefit, especially when you consider packing sufficient back-up spares.

The best place to buy Energizer non-rechargeable AA-size lithium (L91 type) cells is Battery Junction; see…https://www.batteryjunction.com/energizer-l91aa.html, where they’re $1.60 ea. + tax and shipping. The current Ace Hardware “walk in” price for Energizer Max alkaline AA cells (in an 8 pack) equates to $1.25 each + tax.

FR6-class lithium cells will retain ~ 94% of their capacity after 15+ years of storage, so it’s feasible to maintain extensive backup stock without concern. Alkaline cells degrade more rapidly. They’re specified to be storable for ~ 5 years if kept in cool dry environs, and then be down by only ~ 10%. However, I think that’s a rather idealistic projection.

The story is the same for AAA size cells. Energizer’s lithium (FR6 class) AAA cells are known as type L92, and their listing at Battery Junction is…https://www.batteryjunction.com/energizer-l92-bulk.html. The lithium cell benefits and characteristics that are noted above also apply for the smaller AAA (L92) size lithium cells.

Regarding Travel & Safety: The perilous warnings that you’ve heard about lithium battery fire hazards are real, but they trace entirely to the use of multi-cell rechargeable-type lithium-ion batteries, as commonly used today in cellular smart ’phones, digital cameras, and portable computers (also Teslas). A non-rechargeable (single use) L91/L92 type lithium (LiFeS2) cell does not pose the same fire threat. However, to assure that the federal safety mandate has clear and effective public compliance, ALL lithium power sources are now lumped as posing such fire hazard. As a consequence, you are not permitted to ship lithium cells without a related exterior warning sticker, and shipping by air (or by sea) is banned within the USA. Regardless, when you are traveling by commercial air service, you can then carry on equipment that utilizes lithium cells. This means that all of your battery-powered “carry on” gear (your cell ’phones, computers, digital cameras, strobes, flashlights, et al), whether using lithium cells or not, are welcome aboard the passenger compartment of the aircraft. Packing such devices in your checked luggage is permissible only if they are “…completely powered off and protected to prevent unintentional activation or damage.” This latter restriction is subject to the whim of the TSA agent on duty, so you’d best “carry on” your battery-operated devices. Further, all back-up stocks of uninstalled “…lithium metal and lithium ion batteries are always prohibited in checked baggage and must be placed in carry-on.” And: “When a carry-on bag is checked at the gate or at plane side, any spare lithium batteries must be removed from the bag and kept with the passenger in the aircraft cabin.” (On your person, not in your shoulder bag or briefcase.) Obviously, these last requirements don’t pertain to L91/L92 cells, but they directly apply to any backup rechargeable-type li-ion camera body batteries that you might have in your possession. The relevant TSA publication is here…https://www.faa.gov/hazmat/packsafe/more_info/?hazmat=20 (Issued 5/31/2018).

Re. NiMH Rechargeable Cells: In closing, please take note that many AA and AAA cell users rely on rechargeable nickel metal-hydride (NiMH) cells as their preferred alternate to alkaline cells. NiMH rechargeable storage cells work quite nicely for many applications (far, far better than the older nickel-cadmium rechargeable alternative), but the self-discharge rate for NiMH cells still makes them suspect for general usage. A fully charged NiMH cell reads about 1.2 Vdc when under load at the start (versus ~ 1.55 Vdc for a fresh alkaline cell), but it will sag to ~ 0.9 Vdc after just 6 to 8 weeks of passive storage. Conversely, a stock non-rechargeable L91 (AA) or L92 (AAA) lithium battery will read ~ 1.6 to 1.7 Vdc initially when under load regardless of whether it’s new or has been in storage for several years.

--BG (May 20, 2021)

On Upgrading What’s Analog…

By Bryan Geyer

THE POSIT—Recent technology advances make it practical to upgrade select analog functions in many conventional stereo systems. It’s now possible to…
…dramatically extend the range and enhance the linearity of low bass response.
…materially reduce the mass of the main speaker enclosures.
…clarify the midrange response by isolating it from low bass modulation.
…eliminate archaic high-level passive crossovers that reflect only low order filter slopes.
…enable precise line-level active crossovers with full 4th order (-24dB/octave) attenuation.

BACKGROUND—The pursuit of good sound is evolutionary, but the conviction that high fidelity = big speakers persists. Low frequencies have long wavelengths (28 feet at 40 Hz), so big transducers, in big enclosures, have always been considered essential to reproduce, propel, and propagate big bass. Regardless, monstrous monkey coffins that require prominent front-and-center placement look absurd in domestic living rooms, so reasonable compromise is vital. The usual course is: (a) downsize the speakers, or (b) always entertain outside, or (c) cram the system into a “man-cave”, even though the main LR offers superior acoustics (lower Schroeder frequency), or (d) use headphones only. These are odious options; let’s pursue better.

WHAT’S CHANGED—In the mid ’90s, the push to popularize home theater became dependent on finding a practical way to recreate Godzilla’s footfall while squeezing the mass out of the additional speaker enclosures. This paradox looked hopeless until design guru Bob Carver, then at Sunfire, proposed his astonishing new (1997) self-powered “True Subwoofer” as a viable solution. Market acceptance was slow until Sunfire’s patent grip was softened; then the trade piled on. The new subs could pump out pure (low harmonic content) 30Hz sine wave bursts at pressure levels exceeding anything previously envisioned. Today, relatively compact subwoofers with only a single 10" or 12" diameter driver (e.g., https://www.jlaudio.com/collections/home-audio-e-sub/products/e112-gloss-home-audio-e-sub-powered-subwoofers-96279) can produce lower, louder, and cleaner bottom bass (the two bass octaves below 80Hz) than the very best, biggest, and most expensive full range floor-standers ever built by anybody, anywhere.

Tempering this marvel is the reality that these new self-powered subwoofers excel at just one thing: They can handle those two bass octaves better than any conventional woofer. They’re able to do that because they’re expressly designed for a single, solitary mission. They have extremely deep, ultra stiff cones, long and compliant surrounds with extended x-mass capability, and they’re driven by high power class D amps that apply carefully contoured equalization. Unlike the woofers in a costly full range floor-stander, the “sub” doesn’t have to produce any appreciable output above ~ 140Hz. Conversely, a conventional full range floor-stander is designed with the expectation that it must also cover the full low-to-middle bass range, commonly extending to frequencies approaching ~ 800Hz, and do so with good linearity. That’s a stringent additional assignment.

TODAY—If the intent is to achieve optimal fidelity, we need to reconsider how stereo systems should be configured. Clearly, paired stereo subwoofers* should handle the bottom bass octaves. The main stereo loudspeakers can then be reduced in size to reflect the fact that they will no longer have to reproduce those subterranean sounds. Stand-mount main speakers with 6" to 8" drivers have proved absolutely optimum. Excellent bookshelf-size speakers, like Harman’s Revel Performa M126Be (https://www.audioholics.com/bookshelf-speaker-reviews/revel-performa-m126be), or the Revel M106 (https://hometheaterhifi.com/reviews/speaker/bookshelf/revel-performa3-m106-2-way-bookshelf-monitor-loudspeaker-review-part-one/), come to mind, and they confer welcome decor relief. The related subwoofers are best placed near the front corners of the room, at or near floor level, where they’re generally easy to accommodate. There’s simply no need, anymore, for monstrous four to five foot tall main speakers that are packed with multiple 10", 12", or 15" woofers. At least, not unless you’re personally committed to high-end electrostatic-type speakers, such as those championed by Sanders Sound Systems. (In that event, please see https://classicalcandor.blogspot.com/2019/09/on-elephants-in-room.html.)

In addition to the cited decor benefit, separating the low bass from the upper bass presents a compelling aural upside. In a conventional setup, the main woofer and its power amplifier handle both signals, so there’s often some mild modulation blur of the upper bass or lower mids when the score calls for a sudden burst of big bass thunder. Routing that sound directly to the low-pass amplifiers, inside the self-powered subs, means the main speakers will now be unaffected. The main power amplifier will never see those low bass signals when they’re appreciably below the crossover notch. Eliminating this potential midrange modulation is one of the most vital rewards that you can claim when you add subwoofers and install an external active crossover. Some listeners feel that the midrange benefit is even more audible than the subwoofers’ bass extension. Your own impression might be a bit program-dependent in sensing these complementary improvements, but it’s clearly evident when the score fits**.

MODERNIZING CROSSOVERS—A further concern is that conventional full-range speakers impose the need for accurate driver-level crossover networks. Such filters are fussy to design and awkward to assemble. They’re wholly dependent on precisely defined R/L/C components that must pass high current signals to low impedance loads. Tight tolerance passive parts of this unique nature are rare—most have no other fundamental application in the scope of the electronics trade, so they’re often difficult to obtain in the exact values that are needed. The inevitable consequence of this difficulty is compromise, and compromise breeds inaccuracy†. Even when they’re effectively designed and constructed, these high level crossovers often prove inadequate because they’re typically just first order or second order (-6dB or -12dB/octave) filters with gentle attenuation slopes. Full fourth order (-24dB/octave) passive crossover networks would be more suitable, but they become impractical (too complex, physically prohibitive) for high level (low impedance) applications.

The solution to this crossover dilemma is to move that function from its classic position, between the power amplifier outputs and the various driver inputs (where it must process high level signals), to an earlier point, prior to the power amplifier, where the components will see only lower line level signals.This preferred location assures ready access to lots of stable, tight tolerance R/L/C components of virtually any desired value. When combined with active op-amp circuitry it’s then possible to form highly selective Linkwitz-Riley type filters with full 4th order attenuation slopes. The penalty implicit with inserting the crossover in this line-level position is that a separate power amplifier will now be needed to connect each crossover output (high pass and low pass) directly to its designated driver. That means operating in “bi-amplifier” mode, a form of connection that confers useful advantages; e.g., better load (driver) damping; also the ability to independently dictate the signal amplitude sent to each power amplifier.

A fully external (meaning separately enclosed and powered) active crossover controller is invaluable when adding supplementary self-powered subwoofers. The active crossover’s high-pass outputs can feed the main stereo power amplifier + main speaker system, while its low-pass outputs feed the line inputs on the self-powered subs. (The active crossover thereby supersedes any built-in low-pass filtering that might be a part of the sub’s internal circuit, so the subs should then be operated in their selectable “bypass mode”.) This setup facilitates convenient adjustment of the “mains-to-subs” dominance (the relative output levels of the main speakers compared to the subs) from a single, central location. Without this crossover controller you’d need to crawl to each individual subwoofer and separately adjust their input sensitivity controls—and repeat that crawl every time you want to tweak the mains-to-subs settings or alter the relative channel balance.

An external active crossover controller with variable frequency capability will allow the user to position the crossover notch at a point that’s optimum for the main speakers in use. For most subwoofer setups, this will involve a setting that’s somewhere between ~ 76Hz and 100Hz. (Lower crossover frequencies are not advisable, regardless of the main speakers’ perceived bass capability.)

Loudspeaker systems with internal line-level active crossovers + self-powered class D amps are already in common use as near-field monitors; also as desk-top setups. Designs of this nature offering higher power output have also appeared, e.g., the “Kii Three”. Clearly, this trend will accelerate. We’re going to see lots of analog circuitry integrate with the loudspeaker, and some of the speaker makers are likely to merge with those that currently make power amplifiers. Of course, vacuum tube power amps—already a moribund species—will then fade forever. Ditto for big solid-state quasi-class A power amps and heavyweights that resemble engine blocks. Niche products will always exist, but it’s doubtful that those who make them will prosper and survive, so be wary about where you hitch your wagon.

THE PREAMPLIFIER—Refer https://classicalcandor.blogspot.com/2019/12/on-controlling-volume.html. Amen.

Footnotes…
*Paired stereo subwoofers (not a solo “shared bass” subwoofer) are essential for acoustic advantage; refer https://classicalcandor.blogspot.com/2019/03/basics-about-bass_20.html.

**Re. “when the score fits”…is a phrase that brings to mind a subwoofer quirk that you need to know. Many (all?) of the companies making self-powered subs provide the useful ability to select automatic “signal sensing” as a means of activating the sub’s internal power amplifier; i.e., to auto-awaken the sub from its passive sleep state. In sleep mode, the sub stays on, but it dissipates minimal power until it perceives that it’s time to rumble. The sub then goes into active mode, and remains active until there’s no input signal for an appreciable duration, e.g., 20 or 30 minutes. An initial problem is sometimes evident with respect to the maker’s signal-sensing level. Bear in mind that the sub can only hear very low frequency signals because its line-level input path is through a low-pass filter. When there’s persistent heavy bass (as with virtually all pop music today), the sub will awaken quickly. But when the bass content is evident only where Mozart scored it, that sub might not awaken (go active) until you’re a third through an extended concerto. Classical music just doesn’t exhibit the same pounding bass line that persists with pop. Unfortunately, the folks that make subs think classical music is extinct, so they normally set the signal sensing to trigger at a higher (less sensitive) level than what’s optimum for classical format. In many cases the original factory setting is as much as 14dB to 16dB too high (too insensitive). So—when you buy a sub, communicate with your supplier. Have the maker set your subs to auto-activate when The Lark [is barely] Ascending. (I bow to Mr. Vaughan Williams.) If they fail to do this you’ll then need to return your subs to the source, and have a factory guy readjust the trip point (warranty return, maker’s expense). With all of the subs that I’ve seen, this setting is always an internal adjustment. Some subs might now make this tweak externally accessible, but internal-only is more logical because it’s a “set once” consideration, and external access invites misuse. Home-based DIY readjustment is definitely not feasible. The internal class D “plate amp” has complex multi-layer boards + tiny surface-mount (not “through-hole”) components, and that makes this simple task too risky when no schematic is at hand.

Be assured that there’s no downside related to increasing the subs’ activation sensitivity. When your external active crossover controller’s low pass outputs (to the subs’ inputs) cease sending a source signal (for the specified time lapse) to the subs, they’ll automatically revert to their dormant sleep mode, exactly as intended. The only instance in which this might not happen is if your crossover controller suddenly became intolerably noisy—with noise so annoying that you’d be fully absorbed with that fix before noting that the subs didn’t go to sleep.

†It’s only natural that those who own big full-range floor-standing speakers would assume that the internal high level crossovers buried inside their costly purchase are both optimum and accurate. Regardless, their faith is likely misplaced. Accuracy is certain to be deficient in the case of passive networks that have accumulated appreciable use because passive high level crossover components drift. High level operation = high level stress. (Refer paper headed “On Crossing Over,” at https://classicalcandor.blogspot.com/2020/01/on-crossing-over.html.)

BG (April 2021)

On Rational Selection--Part II…

By Bryan Geyer

In Part I, I listed the most important specifications that should be reviewed when selecting an appropriate power amplifier. These truly critical parameters included…

            …size and appearance.

            …input impedance (Zin).

            …output impedance (Zout).

            …power output capability.

            …voltage gain; i.e., input sensitivity.

Now let’s look at some other specs that merit attention. These aren’t quite so vital as the essential considerations that were listed in Part I, where appropriate compatibility was the main objective. Instead, these specs reflect basic quality-related issues, such as…

SIGNAL-TO-NOISE (s/n) RATIO: Freedom from hum will be highly dependent on power supply design and grounding issues. Freedom from noise will generally trace to circuit design and component selection variables. It’s desirable to exhibit a wide spread (e.g., a difference of 100dB or more) between the value of a given output signal (the modern reference is 1Vrms) and the residual detected when there’s no input signal (input shorted). This represents the amplifier’s innate signal-to-noise (s/n) ratio. It’s commonly measured through a filter (it’s “A-weighted”) to emphasize audible frequencies rather than ambient hum or response beyond the audible spectrum. If the consequent noise reading is 10µVrms, the related s/n would be -100dB relative to a 1Vrms reference. A fine quality solid-state power amplifier of current design will meet -110dB (re 1Vrms) s/n ratio. A 90dB s/n ratio is certainly not state-of-the-art, but it’s relatively acceptable. A s/n ratio that dips to -80dB (0.1mVrms re. 1Vrms) means that you’ll start to hear some annoying background hiss from your loudspeakers, dependent on their efficiency. Of course, all vacuum tube power amplifiers produce poor signal-to-noise ratios—it’s an inherent limitation.

QUIESCENT DC OFFSET VOLTAGE: Well, it’s just nuts to neglect checking your power amplifier for DC offset. It’s easily measured with almost any DC voltmeter, and it’s vital that offset be minimized. Do it! Check this column for background info.: https://classicalcandor.blogspot.com/2019/10/on-power-amplifier-dc-offset.html

FREQUENCY RESPONSE, DISTORTION, & DAMPING FACTOR: These specs were once of vital concern in expressing the efficacy of a power amplifier. That’s no longer the case today because virtually every solid-state power amplifier exhibits near-perfection with respect to its innate frequency response and distortion performance. Ditto for damping factor, as that parameter is directly related to a solid-state amplifier’s ultra-low output impedance; refer Part I of this paper. As a consequence, you can safely skip these three specifications when ranking the merits of solid-state power amplifiers. Any divergence will be too small to justify analysis. Conversely, there are operating characteristics that you should research, such as…

INPUT OPTIONS: If it’s your intent to apply balanced interconnects, make certain that the required XLR sockets are provided. I don’t personally recommend the use of XLR balanced interconnects in a normal home stereo installation (refer “On Noise, Coax, and Control” at https://classicalcandor.blogspot.com/2020/08/on-noise-coax-and-control.html); however, you might want that option. Balanced in/out connections are of great benefit when operating in noisy environs, but typically present no measurable advantage in benign home environs. XLR plugs and sockets derive from an archaic vintage (see https://classicalcandor.blogspot.com/2019/10/on-equipment-interface-options.html), so they’re grossly oversized and generally awkward to accommodate.

Most modern power amplifiers now offer both XLR (balanced) and RCA-type (unbalanced) inputs, although some older units might lack the XLR option. Confirm (view back panel photo) that you’ll get what you want. To the best of my knowledge, there’s just one power amplifier on the commercial market that provides XLR balanced stereo inputs but NO unbalanced (RCA-type) stereo inputs, and that’s the Benchmark AHB2. They offer short RCA-to-XLR adapter cables to accommodate users that require unbalanced input compatibility.

POWER LINE PROVISIONS: Verify that any power amplifier of interest will be compatible with the power line provisions available at the site where the amplifier is to be installed. This implies more than merely confirming that the AC supply voltage and power line frequency are compatible—it should also encompass the AC current drain (AC line load) compatibility. Refer “On Assuring Adequate AC Power” at https://classicalcandor.blogspot.com/2020/05/audio-tech-talk.html.

TURN-ON and TURN-OFF PROVISIONS: While it’s always possible to turn your power amplifier on/off manually, via a switch that’s wired in series with the power supply, there are potentially more convenient options to consider. Here are some popular automatic control options…

(1)  Turn on/off via a low voltage (12-15V) DC control signal: This is a popular turn-on option. It’s commonly offered on many modern power amplifiers and on other peripheral components.

(2)  Turn on/off via signal sensing: The amp senses an audio input, then turns on. The turn on sensitivity setting is generally variable, via a back-panel switch or pot, so that background noise won’t be sufficient to falsely trigger operation. The off function is generally based on a time delay; i.e., when no input signal is detected for a specified period of time (e.g., 5 minutes) the amplifier then turns off.

In either case, it’s helpful to have a power amplifier that connects the input signal path only after a momentary (relay-controlled) sequential delay. That permits the initial DC power-up surge to “settle out” first, so that there’s no audible turn-on “thump”; i.e, no sound output until the amplifier’s internal DC power supply has stabilized.

IN THE FUTURE: As you consider these issues, do bear in mind that the days of a stand-alone stereo power amplifier look limited. The current design trend is to fully integrate the power amplifier with the loudspeaker system. Future audio systems will reflect that preference, especially when listeners learn to appreciate that…

…it’s more accurate to use active—rather than passive—crossover networks to feed the drivers in a speaker system with their intended passbands.

…low bass frequencies (those < 60 Hz) can be more accurately produced by separate self-powered subwoofers than by an integrated driver that must also handle all of the mid-bass frequencies (to 600 Hz) as well. The best top quality speaker systems will no longer include multiple large diameter woofers because we’ve learned that the classic all-in-one-box approach is an inferior and outmoded way to implement a full spectrum sound system.

BG (February 2021)