JNorth1178
JNorth1178's picture
Offline
Last seen: Never ago
Joined: Jul 2 2007 - 7:17am
Are B&W speakers happy with Tube amp and preamp?
Jan Vigne
Jan Vigne's picture
Offline
Last seen: Never ago
Joined: Mar 18 2006 - 12:57pm

It depends on the amplifier, the speaker in question and what you mean by "enhance".

Electrically many tube amps aren't well suited to B&W's recent speaker line due to the higher current draw many of the speakers demand. Traditionally, reasonably priced tube power amps have no great ability to deliver the sort of amperage on demand which can be found in most solid state amplifiers in the same price range. Additionally, a transformer coupled amplifier, which includes any reasonably priced tube amplifier, will have a slightly higher output impedance than a direct coupled solid state amplifier. The higher the amplifier's output impedance (tube, solid state or MOSFET), the more stable the speaker's impedance curve should be. The higher the output impedance of the amplifier the more it will react to the wandering impedance load of the speaker and the frequency response of the pair will rise and fall with the speaker's impedance curve. Most modern push pull tube amplifiers can now be designed with reasonable output impedance values, so check the specs of the amplifier and speaker before making a purchase. Output impedance of the amplifier changes with frequency just as the speaker's actual impedance will. A spec'd nominal 1 Ohm output impedance on the amplifier will probably be sufficiently low to drive a speaker that doesn't dip beneath 6 Ohms at any point with a still good damping factor for "tight" tube-type bass. Pick an amplifier which offers the provision for numerous (at least two) impedance taps off the output transformers. Depending on the speakers you choose to pair the amp with, eight and four Ohm taps should be sufficient. Try both or all taps with the speakers since output impedance will be different from each tap and the resulting sound will be slightly different from each tap with any given speaker.

However, if you want the "enhanced" sound to be more classically "tubey", the more the frequency response wanders with the impedance, the more "tube-like" (not necessarilly in a good way) the final sound. It's just that this is inherently a wrong approach to system matching since the response aberrations aren't controllable. (It's a bit like having someone arbitrarilly set all the positions for you on a graphic EQ device and then locking them so you can't alter the settings.) But, depending on your preferences, the right combination of speaker and amplifier might be a sound you like though not technically correct.

Read any review you can find on the prospective speaker match and look for the electrical phase angle of the speaker at various frequencies. High values here, particularly on the negative (capacitive) side at lower frequencies, would generally rule out a tube amp as a partner. As I said, a tube amplifier usually cannot develop sufficient current in most cases (Quad-like current dumping designs aside) and the sound will be sluggish and rather sloppy overall (most noticeably in the lower frequencies) as the speaker begins to drive the amplifier with Back EMF.

Possibly someone with more technical skills may want to comment further but quite simply I would definitely recommend against a no feedback design in a tube amplifier paired with B&W speakers. Moderate amounts of mostly local feedback will be much better despite what you might think or have read about feedback circuits. Also do not buy an amplifier that sells the ability to switch between pentode and triode operation. Look for a well executed pentode or beam power type output tube in a traditional - either UltraLinear or Unity Coupled - push pull amplifier. No frills. (I prefer a soft start circuit on tubes to extend tube life.)

If you choose the correct tube amplifier to pair with your B&W's, the sound will be more neutral while still retaining the sonic benefits of tubes. If you choose incorrectly, the pairing will (at best) soften the frequency extremes and probably destroy much of what you like in the B&W sound.

I would recommend against listening for "tube sound", i.e. a rolled off high end is typical of how many people think of "tube sound". Many people think this mitigates the "brightness" of a system's sound, particularly those speakers with metal dome tweeters, but I would have to say that is the wrong way to implement a good tube amplifier within a system. It becomes a BandAid approach to poor system matching rather than trying for the most realistic sound quality. Keep the system sound as neutral as possible and you should have the benefits of tubes to mate with any future speaker purchases you might wish.

http://www.symphonysound.com/articles/tubefriendly.html

JNorth1178
JNorth1178's picture
Offline
Last seen: Never ago
Joined: Jul 2 2007 - 7:17am

Hi Jan: Just wanted to thank you for the thoughtful response to my posting. I shall spend some time digesting all of the info. One thing became apparent to me; namely, when I start seriously (money in hand) to make a change, I will need a lot of help. Thanks a lot. JBNorth

Elk
Elk's picture
Offline
Last seen: 3 years 7 months ago
Joined: Dec 26 2006 - 6:32am


Quote:
Also do not buy an amplifier that sells the ability to switch between pentode and triode operation. Look for a well executed pentode or beam power type output tube in a traditional - either UltraLinear or Unity Coupled - push pull amplifier.

Jan, excellent, thoughtful post.

I am curious why you recommend against a switchable amp? Do you dislike the additional complexity or do you see another downside?

Jan Vigne
Jan Vigne's picture
Offline
Last seen: Never ago
Joined: Mar 18 2006 - 12:57pm

A pentode is not a triode. You can disconnect two sections of a pentode to run in quasi-triode operation, but you do not have a triode by doing so. You are not switching from a push pull pentode to a single ended triode, so I see this all as mere marketing. The manufacturers are playing on the ignorance of most buyers to be unfamilair enough with tubes to the point they believe they suddenly have the "magic" of triodes. Single ended triodes are a whole different ball game and suggesting you can make something be what it is not is dishonest if not outright ridiculous, IMO.

A local dealer recently picked up the PrimaLuna line where the new Dialogue amps are switchable. Despite the fact the dealer has carried Audio Research for decades, no one in the shop knew enough to be able to explain why a client might want a triode amplifier and what the ramifications of triode amplifiers might be. They don't sell a speaker that would run well on a SET. So what are they selling? A gimmick! They really didn't know how to effectively demonstrate triode operation to sell it as a benefit/advantage, just as a feature.

I don't really think there is a benefit/advantage to disconnecting sections of a tube. A KT88 is never going to sound like a 300B and PP is never going to sound like SE. But it is another example of the sliding scale of "accuracy" we have come to expect in high end audio. Change things until you like them, even if it is just between songs on an album. No one listens to live music any longer so any attempt at recreating the live event falls victim to making it sound "pretty" or just different.

That's really all I can see as the benefit from switching between pentode and triode or switching out negative feedback. Buyers might have read just enough to think this is desirable and that is not; so sell features, the faceplate and the remote. That's how you sell a $200 HT receiver. If there's no benefit or advantage other than you can, just what is the manufacturer selling?

JNorth1178
JNorth1178's picture
Offline
Last seen: Never ago
Joined: Jul 2 2007 - 7:17am

I appreciate all that has gone before. Just got an offer on a prima Luna 6 amp to go with my prima luna 3 pre and my B&Ws speakers. Is this combo feasible? I got a bit lost in the previous post. Thanks. JBN

CECE
CECE's picture
Offline
Last seen: 5 years 3 months ago
Joined: Sep 17 2005 - 8:16am

Get a HYBRID from AVA. www.avahifi.com Sound great, priced for mortals, super rliable, not made in China, so you know it won't kill or maim pets, etc. Made in teh "heartland" of America, rah rah rah. Since 4th of July just passed, buy real hi fi, from the hi fi guru, has been for almost 40 years, how many audio companies exist after 40? a few classis, hmm, they too are U.S. Marantz, (weel now Japanese former Dutch, started in Long Island N.Y.) McIntosh, same thing, not Dutch owned though, now part of same comapny as Marantz, all good. AVA is STILL by teh same guy Frank, no buyouts needed to survive, hmmm, must be doing something right, great stuff, priced right, without teh nonsense. www.avahifi.com

Elk
Elk's picture
Offline
Last seen: 3 years 7 months ago
Joined: Dec 26 2006 - 6:32am

Jan, good point. I wasn't thinking of it this way; I was making the issue much more complicated thinking that your objection may be much more technical than this.

JBN, I don't know the two pieces of equipment well enough to have an opinion. Jan's initial advice to you is excellent, but doesn't address your current specific question.

Give Kevin Deal at Upscale Audio and ask him. He is the importer of PrimaLuna. He specializes in tubes but carries solid state as well. He has a great pair of ears and is happy to share his opinions. Be prepared to tell him what you are looking for in your system; that is, do you wnat mid-range warmth, high end extension, harmonic accuracy, low end slam, etc. - what's most important to you and what are you willing to compromise?

Jan Vigne
Jan Vigne's picture
Offline
Last seen: Never ago
Joined: Mar 18 2006 - 12:57pm

Sorry, I don't know the amplifier or speakers well enough. I cannot give you a fish, you will have to learn to fish for yourself.

Find reviews of the amplifier and speakers.

Look at the highest output impedance of the amplifier.

If it is higher than 1 Ohm, the amplifier will very likely react with some degree of frequency response errors as the speaker's impedance drops. Most tube amplifiers will have an output impedance that rises above 1 Ohm at some frequency. Over how much of its bandwidth it remains above 1 Ohm is what's important when mating to a loudspeaker. If the amplifier's output impedance remains above 1 Ohm for most of its bandwidth, the simpler the speaker load must be. This is partially the reasoning behind matching SET's (typically high output impedance) with single driver full range speakers (very consistent impedance with little or no intervening capacitors or inductors to create a dramatic phase angle).

How much error occurs depends on the actual impedance and phase angle of the speaker at any similar frequency (for example 100Hz & 100Hz, or, 5kHz & 5kHz).

Look at the speaker's measurements and find the lowest impedance point and the worst case phase angle point.

If the corresponding impedance of the speaker is not high at the same frequency where the phase angle is low, then pass on a tube amplifier that has a high output impedance. In fact, sell the speaker and buy something more competently designed. This high/low combination is difficult for any amplifier.

Remember, the output impedance of the amplifier varies from tap to tap. If the speaker has a lowish impedance, you should be judging the amplifier's output impedance taken from its lowest impedance tap, usually the four Ohm tap.

If the speaker's impedance and phase angle are not relatively benign or it has a high phase angle where the impedance is low, it is probably not a match to a tube amplifier. Again, sell the speakers.

The speaker's overall average impedance does not have to be high or low, but a stable impedance is desirable with any tube amplifier - or any amplifier for that matter. Your solid state amplifier will sound better with a stable impedance load and a low electrical phase angle. Because solid state, direct coupled amplifiers tend to have a low output impedance, they will not have the same degree of frequency response errors as a transformer coupled amplifier (tubes), but the solid state unit may not be able to deliver the current on demand that difficult load speakers require.

If the speaker's electrical phase angle is mostly capacitive (a negative phase angle), sell the speaker. Forgoing that, do not match it to a tube amplifier.

Jan Vigne
Jan Vigne's picture
Offline
Last seen: Never ago
Joined: Mar 18 2006 - 12:57pm

Yes, my advice refers only to electrically matching the speakers to the amplifier. It does not guarantee the sound quality of either component is a synergistic mate to the other. You will need to have your priorities in mind when you go about mating two or more component's sound qualities. It is never a good idea to buy a pig in a poke. Buying an amplifier without hearing it paired with your speakers, and preferrably in your system with your favorite music, is likely to lead to disappointment.

pbarach
pbarach's picture
Offline
Last seen: 3 weeks 1 day ago
Joined: Sep 8 2005 - 3:10am

Jan: Can you explain exactly what phase angle is, and how to read the graphs that appear in Atkinson's lab reports concerning this variable?

Jan Vigne
Jan Vigne's picture
Offline
Last seen: Never ago
Joined: Mar 18 2006 - 12:57pm

"Since the voltage across a capacitor is proportional to the integral of the current, as shown above, with sine waves in AC or signal circuits this results in a phase difference of 90 degrees, the current leading the voltage phase angle. It can be shown that the AC voltage across the capacitor is in quadrature with the alternating current through the capacitor. That is, the voltage and current are 'out-of-phase' by a quarter cycle. The amplitude of the voltage depends on the amplitude of the current divided by the product of the frequency of the current with the capacitance, C." http://en.wikipedia.org/wiki/Capacitor

Place "capacitor phase angle" and "inductor phase angle" in a search engine for more complete explanations or, click here; http://search.yahoo.com/bin/search?fr=ybr_sbc&p=capacitor%20phase%20angle. Or, possibly you will do better to try "speaker crossover phase angle" to understand exactly what happens in a speaker rather than a generic circuit.

Passing current/voltage across a capacitor or inductor causes a lead/lag situation where one of the two components of "work" leads the other and thus creates the electrical phase angle which results in less work being done.

A capacitor will cause a negative 90 degree shift with current leading voltage. An inductor will have a positive 90 degree shift and voltage will lead current. The most efficient "work" is done when voltage and current are delivered to the motor simultaneously (in the case of a speaker's driver[s]).

Most speaker crossovers, though not all, are constructed or managed by combining capacitors and inductors (and often resistors) to form the filter order desired. Filter order being first, second, third, fourth, etc. and each order increasing the filter roll off by an additional six decibels per octave. So, a first order filter is a fairly simple affair that can be no more than a capacitor hung on a tweeter to create a high pass filter (in this application it blocks low frequencies and passes high frequencies - a high pass filter). This first order, high pass filter applied to a tweeter will result in six dB of roll off to the frequency response of the driver(s) beginning at the "knee" of the filter (the frequency where it begins to roll in or out, in this case in). When you see a crossover frequency stated, say, 2.5kHz, you are seeing the point where both drivers being crossed are down 3dB resulting in a nominally flat response at the spec'd "crossover" frequency. Look at the response graphs in Stereophile to see this action.

The electrical phase angle of the combined crossover components (capacitors and inductors) reduces the work being done by the amplifier to the point, as the angle increases, where less and less work is being accomplished as voltage and current are thrown further and further out of phase. This voltage/current lag can increase to the point where, at the most extreme phase angle, no work can be accomplished. Obviously, this severe "no work" phase angle is never reached in a speaker's crossover but the steeper the phase angle, the harder the amplifier must work in order to drive the speaker system.

More current will be drawn from the amplifier if the phase angle and the impedance are both too low at the same frequency - indicating a predominantly capacitive load in the phase angle (which is most commonly the case in speaker crossovers). But, inductive loads are just as troublesome as capacitive loads since an amplifier prefers to work into a "resistive" load. A complex capacitive (negative) phase angle at one frequency combined with a highly inductive (positive) angle at another frequency (not too uncommon) makes for a difficult speaker load on any amplifier unless the impedance of the system is fairly stable. The more severe the phase angle changes along with impedance, the more "reactive" the speaker is said to be. The further the speaker strays from a purely resistive load (which cannot be accomplished since the voice coil of even a single driver is inductive to some degree) the harder the speaker will be for any amplifier to drive and the more current will be required from the amplifier. Available current- well done - is (generally) expensive in audio amplifiers.

So, when you see the impedance dipping to 4 Ohms or lower and the phase angle dropping to (negative) 45 degrees or greater, the speaker will be difficult to drive (reactive) and will require a sturdy power supply in the amplifier and likely lots of heat sink area. Receivers need not apply.

In Sterophile's measurement pages, the negative phase angle is a capacitive load and the positive phase angle is an inductive load. The solid line is impedance and is read from the left hand axis of the graph. Frequency is the baseline. The dashed line is phase and is read from the right hand axis. The impedance rises above the base line of the graph since there will be no negative impedance in a speaker while the phase is charted from the center of the graph and divided to reflect both negative and positive angles. Notice the "0" on the right hand side of the graph, negative phase (capacitive) is beneath that point and positive (inductive) phase angle is above that point.

It is the cumulative phase angle of the combined inductors (including the driver's motor though this measurement varies with temperature as the speaker is asked to play louder for a longer period of time) and capacitors at any one frequency that are important to consider when judging whether a speaker will be efficient (easy to drive) or a current hog (difficult to drive). The "sensitivity" spec, IMO, of a speaker tells you very little about the actual volume potential of the speaker system as higher phase angles and more crossover components will rob the drivers of workable power leading to wasted wattage that literally goes up in the air as heat. Speakers also have differing "sensitivity" at different frequencies so one number is not going to tell the entire story.

It is important to note that electrical phase angle is not the same as absolute phase of the (acoustic) audio signal and that AC circuits work differently than DC circuits. (An active crossover can, if well designed, do minimal damage to the phase angle and/or the absolute phase of the audio signal while passive crossovers will do some damage.) When you read the measurements page and see the drivers have been connected in reverse acoustic phase, this refers to the absolute phase of the audio signal and is the result of the filter action of the crossover. Each additional order of filter will throw the audio signal another 90 degrees out of absolute phase, so the steeper the filter order, the more the speaker looses absolute phase from one driver to the next. (This is reflected in the step response graphs Stereophile includes.) This too is not the same as time alignment in a speaker. Among other things, this lack of absolute acoustic phase is one reason for arguing for single driver speakers, there are no crossover components to affect acoustic phase and most single driver speakers are both time and phase coherent. However, single drivers have their own problems.

As I've pointed out in another thread, complex crossovers which suck current due to severe phase angles will be very inefficient. A Klipschorn, with a simple, near constant impedance, low phase angle and high sensitivity will be approximately 10% efficient while a speaker with multiple drivers, complex crossovers and high phase angle combined with low impedance will drop down to 1% efficiency. While "sensitivity" and "efficiency" are typically used to describe the volume potential of a speaker and how easy it will be to drive or how many watts are required for "X" SPL level, both are distinct measurements that do not combine to make a single statement. I can probably get some disagreeement on this, but my experience says that two company's speakers, each spec'd at the same "sensitivity", will not play at the same level for the same input voltage and will certainly not always play well on the same amplifier. The former may be just a matter of fudging on paper specs while the latter is typically due to impedance and phase angle.

I hope that all makes sense. Possibly someone with more engineering acumen will fill in the holes and expand on what I've posted here.

JA, you're on.

smejias
smejias's picture
Offline
Last seen: 13 years 4 months ago
Joined: Aug 25 2005 - 10:29am


Quote:
Jan: Can you explain exactly what phase angle is, and how to read the graphs that appear in Atkinson's lab reports concerning this variable?

These three essays might help:

Measuring Loudspeakers, Part One
Part Two
Part Three

Log in or register to post comments
-->
  • X