Tube gear obsolete??

[Ric Schultz]

I think what this whole enchilada comes down to is: What kind of sound gets you off?

Years ago I worked for Audible Illusions so I had a Modulus preamp at home. I decided to do a straight wire bypass test on it. I had two of the jacks on the back of the Modulus wired together for the straight wire and I used my transistor phono stage directly into my amp (no volume control). There were simply two interconnects between the phono stage and the amp. Another set of jacks on the Modulus had two resistors on the input that padded the signal the same amount as the tube gain. I ran that resistor pad directly into the grid of the tube line stage (no switches, pots or anything). The output went through a Wondercap and then right to the output jacks. The transistor phono stage going directly through the two interconnects sounded good......but going through the single tube line stage sounded outrageously better. It plain sounded bigger, warmer, more real, more everything. No way anyone would have preferred the "straight wire". However, all this great sound was just the added coloration of the tube line stage.....plain and simple.

There was an article in the old FI magazine where they described a listening session: First they had someone stand in between the speakers and sing. Then they played a CD of him singing the same song through a system that had a solid stage amp (Aleph 3?). The CD sounded good and pretty close to him singing live. Then they inserted a single ended Class A tube amp for the solid state amp and everyone came in their shorts! Wow, you could now hear his tonsils....its so alive....its so pretty....its so big....it sooo.....everything. Then they realized that this sound was better than when he sang live! Yes, better than live!!!! (I might have some details wrong...its been a long time since I read that article.....but you get the idea).

It is not that all tube equipment makes the sound better (more liguid, sweeter, smoother, creamier, bigger, more space, etc.). Some tube stuff is quite transparent. However, tube gear is really easy to make sound good whereas transistor gear is unforgiving. If you do not pay attention to every little thing in a transistor amp then you will hear brightness, closed down sound, graininess, glassiness, peakiness, etc. You need to upgrade the fuse, damp the heatsinks (how many transistor amps have damped heatsinks?.....not many....if you run your fingernail down a heatsink and it goes zing...then you are adding zinginess to the sound of the amp. Transistors, regulators and diodes need to be damped if they are sitting up in the air on leads and not held to something. Binding posts are veiling and some are downright horrible. Some diodes sound really bad (fast, soft recover are nice). Then of course there is feedback......too much and you get a closed down sound relative to what you can get without feedback or with a small amount. If you have a really fast tweaky circuit like the Job then a small amount of feedback is OK. You have to get the square wave really steep and clean with no overshoot....then you get sweet extended highs. High open loop bandwidth is really cool.

If we all had access to amazing master tapes then it would not matter so much what we played them through (great solid state, tubes or hybrid). But we don't. There is almost always some harmonic richness and openness that is missing from our recordings. So it makes sense to have at least one tube piece in the chain to sweeten it up. Check out Srajen on 6 moons. He is always talking about the synergy you need to get the sound that "he likes". Usually it is a combo of passive or transisitor pre and a tube amp or a tube preamp with a solid state amp.

Do you really want to know how transparent your amp is? Very simple. Just do a straight wire bypass test on it. My friend Sieg Modes was doing this back in the 70s. You simply insert your amp in between a preamp and another amp and turn the preamp down the amount of gain of your amp. You need to mount rca jacks on the output of the amp. Remember, an amp is simply a gain stage. If it cannot drive the high impedance of an amp without coloration it certainly is not going to do any better driving a 2-8 ohm speaker. All arguments about transparency would end if someone were to do this. Transparency means.....no sound of its own. A great "sounding" amp is not necessarily transparent. How come no manufacturer in the history of audio has ever done this? Could it be that we don't really want to know? I mean....."I paid $50,000 for that color!?!"

I personally feel that if you tweak each technology to its max then both solid stage and tubes will get you dang close to "no sound". It used to be that transistor amps were considered a no no.....but in the last 10 years transistor amps have made great strides: Dartzeel, Soulution, Constellation, Technical Brain, Goldmund/Job, Ncore class D, etc. etc.

So, we are stuck with our "colors".....since all amps sound different then only one amp could be totally "transparent". I know, its the one you own....he he. We play at the highest level we can afford (or tweak to) and we pick the colors that match our other colors and if we are smart and keen we get some synergy and sound that is amazing. I think this is the best we can do right now. Enjoy you color!

 

[MylesBAstor]

I think what this whole enchilada comes down to is: What kind of sound gets you off?

Years ago I worked for Audible Illusions so I had a Modulus preamp at home. I decided to do a straight wire bypass test on it. I had two of the jacks on the back of the Modulus wired together for the straight wire and I used my transistor phono stage directly into my amp (no volume control). There were simply two interconnects between the phono stage and the amp. Another set of jacks on the Modulus had two resistors on the input that padded the signal the same amount as the tube gain. I ran that resistor pad directly into the grid of the tube line stage (no switches, pots or anything). The output went through a Wondercap and then right to the output jacks. The transistor phono stage going directly through the two interconnects sounded good......but going through the single tube line stage sounded outrageously better. It plain sounded bigger, warmer, more real, more everything. No way anyone would have preferred the "straight wire". However, all this great sound was just the added coloration of the tube line stage.....plain and simple.

There was an article in the old FI magazine where they described a listening session: First they had someone stand in between the speakers and sing. Then they played a CD of him singing the same song through a system that had a solid stage amp (Aleph 3?). The CD sounded good and pretty close to him singing live. Then they inserted a single ended Class A tube amp for the solid state amp and everyone came in their shorts! Wow, you could now hear his tonsils....its so alive....its so pretty....its so big....it sooo.....everything. Then they realized that this sound was better than when he sang live! Yes, better than live!!!! (I might have some details wrong...its been a long time since I read that article.....but you get the idea).

It is not that all tube equipment makes the sound better (more liguid, sweeter, smoother, creamier, bigger, more space, etc.). Some tube stuff is quite transparent. However, tube gear is really easy to make sound good whereas transistor gear is unforgiving. If you do not pay attention to every little thing in a transistor amp then you will hear brightness, closed down sound, graininess, glassiness, peakiness, etc. You need to upgrade the fuse, damp the heatsinks (how many transistor amps have damped heatsinks?.....not many....if you run your fingernail down a heatsink and it goes zing...then you are adding zinginess to the sound of the amp. Transistors, regulators and diodes need to be damped if they are sitting up in the air on leads and not held to something. Binding posts are veiling and some are downright horrible. Some diodes sound really bad (fast, soft recover are nice). Then of course there is feedback......too much and you get a closed down sound relative to what you can get without feedback or with a small amount. If you have a really fast tweaky circuit like the Job then a small amount of feedback is OK. You have to get the square wave really steep and clean with no overshoot....then you get sweet extended highs. High open loop bandwidth is really cool.

If we all had access to amazing master tapes then it would not matter so much what we played them through (great solid state, tubes or hybrid). But we don't. There is almost always some harmonic richness and openness that is missing from our recordings. So it makes sense to have at least one tube piece in the chain to sweeten it up. Check out Srajen on 6 moons. He is always talking about the synergy you need to get the sound that "he likes". Usually it is a combo of passive or transisitor pre and a tube amp or a tube preamp with a solid state amp.

Do you really want to know how transparent your amp is? Very simple. Just do a straight wire bypass test on it. My friend Sieg Modes was doing this back in the 70s. You simply insert your amp in between a preamp and another amp and turn the preamp down the amount of gain of your amp. You need to mount rca jacks on the output of the amp. Remember, an amp is simply a gain stage. If it cannot drive the high impedance of an amp without coloration it certainly is not going to do any better driving a 2-8 ohm speaker. All arguments about transparency would end if someone were to do this. Transparency means.....no sound of its own. A great "sounding" amp is not necessarily transparent. How come no manufacturer in the history of audio has ever done this? Could it be that we don't really want to know? I mean....."I paid $50,000 for that color!?!"

I personally feel that if you tweak each technology to its max then both solid stage and tubes will get you dang close to "no sound". It used to be that transistor amps were considered a no no.....but in the last 10 years transistor amps have made great strides: Dartzeel, Soulution, Constellation, Technical Brain, Goldmund/Job, Ncore class D, etc. etc.

So, we are stuck with our "colors".....since all amps sound different then only one amp could be totally "transparent". I know, its the one you own....he he. We play at the highest level we can afford (or tweak to) and we pick the colors that match our other colors and if we are smart and keen we get some synergy and sound that is amazing. I think this is the best we can do right now. Enjoy you color!

Actually there's a few WBF members (including myself) with master and second gen tapes who would most certainly take issue with your statement. It most certainly makes a difference [ss or tube] what you play the tapes through! (and the same might be said of the best turntables today also.) And that includes the deck's electronics as well.

And like SS, tubes (and for that matter analog) have made great strides in the same time period too. Fact is that the two technologies sound very different and in my listening experience, much of one's final choice comes down to what aspects of musical reproduction one values most eg harmonics, detail, noise floor, imaging, spatial resolution, timbre, etc for none of them do it all.

 

[Ric Schultz]

I did not say that it "would not matter" which you played them on.....I said it "would not matter so much". I bet you could enjoy both the big Dartzeels and whatever great tubes you choose when playing your first generation 30 inch master tapes. I would be grinning, smiling, crying, dancing, goosebumping and grooving either way. Maybe you could only listen to one amp......Heck, I am moved to tears using my high feedback tweaky solid state custom amp playing CDs on my modded Oppo 105. I would probably die and go to heaven with any of the super heavyweights amps and 30 inch master tapes!!!! I think we spoiled.....he he.

 

[MylesBAstor]

I did not say that it "would not matter" which you played them on.....I said it "would not matter so much". I bet you could enjoy both the big Dartzeels and whatever great tubes you choose when playing your first generation 30 inch master tapes. I would be grinning, smiling, crying, dancing, goosebumping and grooving either way. Maybe you could only listen to one amp......Heck, I am moved to tears using my high feedback tweaky solid state custom amp playing CDs on my modded Oppo 105. I would probably die and go to heaven with any of the super heavyweights amps and 30 inch master tapes!!!! I think we spoiled.....he he.

OK the amplifier matters even more because of the master tapes. These tapes contain more musical information than one ever imagined could exist on an LP (much less CD) and reveals the smallest nuances about the amplifier's (much less system's) performance. Sound of the room, timbre, sense of space, low end extension (disregarding head bumps), instrumental body, macro and micro-dynamics, low level information. Take something like the Tape Project's release of Bill Evans Waltz for Debby and listen to the audience chatting in the background while his trio performs. This recording will separate the men from the boys in a New York City minute.

No, if there's any part of the reproduction chain that seems to be benefit from using tapes as a source, it's the speakers. Tapes can make a $500 JBL sound like a million dollars!

I also don't think we should confuse terminology. Your definition of transparency is what most of us would call colorations. What at least I call transparency is how the noise floor and other things allow us to hear and perceive in the mind's eye instruments that are situated in the farthest reaches of the hall or back of the soundstage. Drums, that are often found on the best jazz recordings like Coltrane's Ballads in the back of the stage, are a great test for transparency.

Finally a side note about Wondercaps. After Moncrief's first generation, they were far from Wonder (ful). In fact, so unlistenable that I ripped them out of my system They were another cap that used ferromagnetic leads too.

 

[microstrip]

Fransisco-Please define "constant power" as it relates to an amplifier playing music through a speaker.

Ralph Karsteen explains it pretty well in his site:

http://www.atma-sphere.com/Resources/Paradigms_in_Amplifier_Design.php

Mep,

You should read this page considering Ralph Karsteen statement "It is my contention that if we are to make serious performance increases in the world of audio, the rules of human hearing can be our only guide." I have to say I take it seriously because I know from long experience how good his amplifiers can sound, when used with adequate speakers. Surely, once we enter the rules of human hearing in audio we enter the subjective zone and many people will not agree with his points, e.g. just stating that any speaker not obeying the voltage paradigm is a defective design and should not exist!

 

[mep]

Fransisco-I have read Ralph's article, and I have a hard time taking it seriously. Ralph refers to a "variety of tests " performed by GE to study distortion without footnotes so one can find the actual paper and read it. But according to Ralph's paraphrased analysis of the GE tests, and I quote from Ralph: "In the 1960s, General Electric's conducted a variety of tests, confirming that amounts of barely hundredths of a percent distortion were not only audible but also irritating to the human ear (conversely, they also found that the ear is quite tolerant of lower ordered harmonic distortion)."

From there Ralph goes on to 'build' his case that only amplifiers that put out the same power into 4, 8, or 16 ohms are true power amplifiers and of course this distinction applies mainly to tube amplifiers. The reality is that tube amps don't put out exactly the same power into different loads. More importantly, amps that have zero feedback are going to have more distortion than amps that use feedback regardless if they are tube or SS. So if GE's tests that Ralph refers to are correct and super low amounts of distortion are audible and irritating, it would appear logical that amplifiers with the highest measured distortion levels would be more irritating to listeners than amplifiers with very low measured distortion. It appears to me that Ralph is picking and choosing which parts of the GE tests he wants to align with. We are now down to the GE 'facts' that very low levels of distortion are audible and irritating and Ralph is trying to minimize the most irritating odd order harmonics so that his amps will have the low order harmonic distortion that our ears are "more tolerant of" while trying to ignore the "fact" of the GE tests which proved that all distortion is irritable and therefore not desirable.

The talk of speakers wanting "constant power" regardless of impedance can't be achieved with tube amplifiers with output transformers. You have to choose a tap from your amplifier to hook up to your speakers and with most modern tube amps, that will be either a 4 ohm or 8 ohm tap. Some tube amps actually have a 2 ohm tap and most have done away with 16 ohm taps. Most people understand that the rated impedance of their speakers is known as the 'nominal' impedance. They aren't purely 4 ohm speakers or purely 8 ohm speakers. As the impedance of the speakers changes dynamically as you are playing music, you can't jump up and switch output taps on the fly which means as the impedance changes, you don't have constant power. And I seriously doubt if OTL amps can put out the same power regardless of speaker impedance either so neither tube amps with output transformers or OTL amps are true "constant power" amps.

Finally, has Ralph's paper ever been published in a scientific journal? I highly doubt it as I don't think it could withstand the scrutiny of a review by peers meaning Electrical Engineers.

 

[FrantzM]

Fransisco-I have read Ralph's article, and I have a hard time taking it seriously. Ralph refers to a "variety of tests " performed by GE to study distortion without footnotes so one can find the actual paper and read it. But according to Ralph's paraphrased analysis of the GE tests, and I quote from Ralph: In the 1960s, General Electric's conducted a variety of tests, confirming that amounts of barely hundredths of a percent distortion were not only audible but also irritating to the human ear (conversely, they also found that the ear is quite tolerant of lower ordered harmonic distortion)."

From there Ralph goes on to 'build' his case that only amplifiers that put out the same power into 4, 8, or 16 ohms are true power amplifiers and of course this distinction applies mainly to tube amplifiers. The reality is that tube amps don't put out exactly the same power into different loads. More importantly, amps that have zero feedback are going to have more distortion than amps that use feedback regardless if they are tube or SS. So if GE's tests that Ralph refers to are correct and super low amounts of distortion are audible and irritating, it would appear logical that amplifiers with the highest measured distortion levels would be more irritating to listeners than amplifiers with very low measured distortion. It appears to me that Ralph is picking and choosing which parts of the GE tests he wants to align with. We are now down to the GE 'facts' that very low levels of distortion are audible and irritating and Ralph is trying to minimize the most irritating odd order harmonics so that his amps will have the low order harmonic distortion that our ears are "more tolerant of" while trying to ignore the "fact" of the GE tests which proved that all distortion is irritable and therefore not desirable.

The talk of speakers wanting "constant power" regardless of impedance can't be achieved with tube amplifiers with output transformers. You have to choose a tap from your amplifier to hook up to your speakers and with most modern tube amps, that will be either a 4 ohm or 8 ohm tap. Some tube amps actually have a 2 ohm tap and most have done away with 16 ohm taps. Most people understand that the rated impedance of their speakers is known as the 'nominal' impedance. They aren't purely 4 ohm speakers or purely 8 ohm speakers. As the impedance of the speakers changes dynamically as you are playing music, you can't jump up and switch output taps on the fly which means as the impedance changes, you don't have constant power. And I seriously doubt if OTL amps can put out the same power regardless of speaker impedance either so neither tube amps with output transformers or OTL amps are true "constant power" amps.

Finally, has Ralph's paper ever been published in a scientific journal? I highly doubt it as I don't think it could withstand the scrutiny of a review by peers meaning Electrical Engineers.

+1 here.

Read the reference to the Ralph's article and came to the same conclusions. I still don't get what is a Constant Power Loudspeaker.

 

[puroagave]

...So if GE's tests that Ralph refers to are correct and super low amounts of distortion are audible and irritating, it would appear logical that amplifiers with the highest measured distortion levels would be more irritating to listeners than amplifiers with very low measured distortion. It appears to me that Ralph is picking and choosing which parts of the GE tests he wants to align with...

given that tube amps might be 'dirtier' than SS its 'distortions' are rather innocuous when riding with the signal as its primarily 2nd order and unoffensive.

what i gather from his statements and what has been widely publicized re even order distortions, even at high levels human hearing is more tolerant to it than odd order distortions, regardless if its small percentage of the total distortion. my own practical example is when you clip a tube amp vs. a SS amp, the onset of clipping in SS is often signaled well before it happens as the sounds get harder and brighter rather quickly. you typically don’t experience that with tubes as you often have to overdrive a tube amp to know its clipping and even then its not that obvious.

 

[sickophant]

+1 here.

Read the reference to the Ralph's article and came to the same conclusions. I still don't get what is a Constant Power Loudspeaker.

I think I am going to be less heavy weight than the players currently circling this thread but I have read some well considered reasonings behind some developer's criteria and I think Robert Koch from Robert Koda probably said something akin to an electrics truth that I have seen when he details the power supply's behaviour to a speaker load being ideal should all speakers be a simple resistor, therefore he builds around the concept of the power supply seeing speaker loads, he mentions as 'reactive' as they are, as a resistor. God I hope I got that right, otherwise I'm in for a slamming; or a sensoring.

There are beautiful designers and builders out there and I think many that have lost their ways, SS and Tube, but on this thread I would still need to vote that Tubes have NOT had their day just yet. There is fabulous gear out there; for anyone interested in listening instead of incessantly posting, I think you've all gone mad.

 

[Atmasphere]

Ralph-I have nothing to recant. I don't know how old you are and you may be older than I am. I will turn 58 in August so I while I was born at night, it wasn't last night. You mentioned the Radiotron Designer's Handbook. I'm familiar with it as I own a copy. You alluded to me not really reading your article several times, but I did. Do any of your amps come with a dampening control? I didn't think so. Dampening controls typically increase or decrease the amount of negative feedback. There are several tube amps on the market that come with feedback controls where you can dial in less or more feedback to suit your tastes with all "purists" desiring zero negative feedback while most push-pull tube amp designers use some amount of negative feedback. Although some Lamm amps come with a feedback control, if memory serves me correctly, Lamm doesn't really recommend monkeying around with it.

There is no way around the fact that tubes are high-impedance/low current devices compared to SS. That is why the majority of all tube amps on the market use output transformers to step down the impedance and raise the output current. OTL amps of course have no output transformer so you find yourself in a different box where you are now forced to connect a bunch of output tubes together in order to try and lower impedance and raise the current. There is a reason why OTL amps like higher impedance speakers and why power falls off as impedance drops. Please tell me one tube amp on the market whether it's OTL or transformer coupled that is going to measure "flat" into the majority of all loudspeakers currently available. To my knowledge, there are none. We can argue whether we like the sound of watts created from high voltage and low current better than we like the sound of watts created from low voltage and high current, but we can't argue that P = I x E. You can argue that you like the sound of tubes better than the sound of SS (and I have owned tube gear all of my life), but you can't argue that a tube amp will measure flat into a loudspeaker load. And no Fisher A-80 is going to change that.

Why don't you present some measurements of your amps taken by someone reputable like JA from Stereophile and show us how flat they measure into simulated speaker loads? If your amps measure flat into simulated speaker loads and the measurements are taken by someone that we have all heard of and respect like JA, I will recant. Until then, I stand by everything I said.

MEP, you simply have no idea of what you are talking about. And you are a few months older than me, so I have to assume audio is not your first interest.

It appears you read the article but did not understand it despite it being written with the minimum amount of math and layman's terms so it would be easy to grasp. And you still persist with the Strawman. I will point out what the Strawman/weakness of your arguement is: I never said anywhere that an amp with high output impedance could have flat frequency response on any speaker!!! What I did say is that using a speaker designed for such an amp, flat frequency response is possible. It is a matter of intention with the designer. This has to do with the two design camps, and if the two technologies are mixed you get tonal anomolies (due to frequency response errors). I **did** say that in the article- therefore you are engaging in a Strawman, and since that is a logical fallacy, your argument is false by definition. Read the article again, but drop your filters. This has more to do with feedback than it does SS vs tubes.

I *did* also point out that Soundstage.com did some measurements and confirmed what I am saying. Look for the MA-1 review published on their site. Bascomb King did the measurements. Then look for my response. You might then see what I am talking about. However I have added some examples below.

Fransisco-I have read Ralph's article, and I have a hard time taking it seriously. Ralph refers to a "variety of tests " performed by GE to study distortion without footnotes so one can find the actual paper and read it. But according to Ralph's paraphrased analysis of the GE tests, and I quote from Ralph: In the 1960s, General Electric's conducted a variety of tests, confirming that amounts of barely hundredths of a percent distortion were not only audible but also irritating to the human ear (conversely, they also found that the ear is quite tolerant of lower ordered harmonic distortion)."

From there Ralph goes on to 'build' his case that only amplifiers that put out the same power into 4, 8, or 16 ohms are true power amplifiers and of course this distinction applies mainly to tube amplifiers. The reality is that tube amps don't put out exactly the same power into different loads. More importantly, amps that have zero feedback are going to have more distortion than amps that use feedback regardless if they are tube or SS. So if GE's tests that Ralph refers to are correct and super low amounts of distortion are audible and irritating, it would appear logical that amplifiers with the highest measured distortion levels would be more irritating to listeners than amplifiers with very low measured distortion. It appears to me that Ralph is picking and choosing which parts of the GE tests he wants to align with. We are now down to the GE 'facts' that very low levels of distortion are audible and irritating and Ralph is trying to minimize the most irritating odd order harmonics so that his amps will have the low order harmonic distortion that our ears are "more tolerant of" while trying to ignore the "fact" of the GE tests which proved that all distortion is irritable and therefore not desirable.

The talk of speakers wanting "constant power" regardless of impedance can't be achieved with tube amplifiers with output transformers. You have to choose a tap from your amplifier to hook up to your speakers and with most modern tube amps, that will be either a 4 ohm or 8 ohm tap. Some tube amps actually have a 2 ohm tap and most have done away with 16 ohm taps. Most people understand that the rated impedance of their speakers is known as the 'nominal' impedance. They aren't purely 4 ohm speakers or purely 8 ohm speakers. As the impedance of the speakers changes dynamically as you are playing music, you can't jump up and switch output taps on the fly which means as the impedance changes, you don't have constant power. And I seriously doubt if OTL amps can put out the same power regardless of speaker impedance either so neither tube amps with output transformers or OTL amps are true "constant power" amps.

Finally, has Ralph's paper ever been published in a scientific journal? I highly doubt it as I don't think it could withstand the scrutiny of a review by peers meaning Electrical Engineers.

Again here you are engaging in a several fallacies, the Stawman (here is what a logical fallacy is, if you want to engage in intelligent debate it is handy to know how they get used:
http://www.nizkor.org/features/fallacies/)
FWIW this is not a scientific article but that does not make it false. That is another Logical Fallacy known as Personal Attack, You have no idea what peers might think, the comment is simply intended to discredit without proof of any sort.

I did not 'build' the case as you put it. What is happening is you don't understand how amplifiers work, and thus don't understand the distortion-generation mechanisms. To give you an idea of how a tube amp without feedback will behave (including OTLs), what happens is the amplifier has a power curve that looks a lot like the cross-section of an aircraft airfoil. It rises quickly from zero ohms, reaches a peak power at a certain impedance, and decreases very gradually at impedances above that. That area above the maximum power is the 'constant power' response (as I mentioned ad nauseum, in the article; no amp is perfect at this just as no SS amp is a perfect voltage source). For example, our MA-2 puts out maximum power of 220 watts into 5 ohms. At 16 ohms it is still making 218 watts (into 3 ohms it makes 205 watts). When you look at this from a perspective of decibels, that is pretty well constant power over that range.

Now this is a matter of the amp having no feedback. It behaves quite differently if feedback is applied (we have built the amps with feedback switches. Spare me the comment about the damping control). The problem is that if feedback is applied, odd ordered harmonic distortion is also added. See Nelson Pass (https://passlabs.com/articles/audio-distortion-and-feedback ) for a nice article about this; however this phenomena has been known for over 60 years (usually accepted as a necassary evil), see the writings of Norman Crowhurst from the late 1950s and early 1960s. In a triode amplifier that is operating class A, the odd ordered harmonic distortion is much lower than the amounts seen in a SS amp. That is why they are not bright. Add feedback and they get bright. So how does one eliminate distortion without feedback? Class A operation, triode operation, single stage of gain, no output transformers, fully differential operation, star grounding, anything that will reduce intermodulations and harmonics IOW. The price paid is that the amplifier will not work with all speakers, not due to low impedance, but due to impedance variation wherein the designer might expect that a high impedance will cause the amp to cut power. For example, a set of our MA-2s has plenty of power to drive a set of mbl 101s. The problem is that the amp sounds bright on them because there is an impedance peak at 8KHz of about 8 ohms. The designer is in the voltage camp and so is expecting the amp to cut its power in half at this frequency. But the amp is constant power so it does not. Thus the brightness.

Here's another example. The old Wilson Watt/Puppy had a reputation for brightness in the 2KHz range. There was a tweeter used that had a 2KHz resonance. So what Wilson put a 2KHz trap in parallel with the tweeter. This caused the impedance to be 2 ohms at 2Khz. Now if you put a SS amp on that, it made 4X more power than it did at the rest of the range of the speaker which was 8 ohms. The result was a tonal anomoly- brightness, because we are mixing the two technologies. But if you put a tube amp on the speaker, it could not make power at that frequency and so the resonance was controlled. Kapish?? It has to do with the intention of the designer!

+1 here.

Read the reference to the Ralph's article and came to the same conclusions. I still don't get what is a Constant Power Loudspeaker.

To understand this you simply have to know that there are two kinds of speakers- those designed for a Constant Voltage response in the amp (example is a SS amp with lots of feedback that can double its output power as impedance is cut in half) and speakers designed for a Constant Power response. Examples of the latter: Quad ESL63, Sound Lab, Merlins, Wilson Audio, most horns, most single-driver full-range speakers, Audiokinesis, Coincident Technology, Rogers LS35A, Acoustic Research AR-1 (the first acoustic suspension loudspeaker) and of course many more. I think the misunderstanding here is thinking that somehow the amp is making constant power all the time- that would be a steady-state signal like a sine wave. What the term refers to is that the amp will attempt have the same amount of power output regardless of the load impedance. It will not be perfect at this but there is a range of speaker impedance where this can be true. The Wilson Loudspeaker in my previous paragraph is an example.

 

[sickophant]

Yep, I definitely think you've all gone mad.

 

[microstrip]

(...) Finally, has Ralph's paper ever been published in a scientific journal? I highly doubt it as I don't think it could withstand the scrutiny of a review by peers meaning Electrical Engineers.

As far as I know, it was never published. I find curious that now you want to debate audio matters in terms of " scrutiny of a review by peers meaning Electrical Engineers." I am sure that these peers would tell you that your Krell KSA 250 sounds the same or better with an idle current of 200mA than with 2.5A .

IMHO, the concept of constant power loudspeaker is just a model to explain some sonic attributes , simple to understand for any one knowing about signal transfer in voltage, current and power modes. I think I could figure what Ralph means with it. Perhaps he will explain it to you, but I will not enter such task. As far as I can understand he does not want to prove anything, just explain what is behind his designs.

Look at the power curve of the Soundlab A1 PX that sounds great with his OTLs. It can help understanding his arguments.

 

[sickophant]

We need Raul to bring a little sanity back to this thread.

 

[Atmasphere]

Look at the power curve of the Soundlab A1 PX that sounds great with his OTLs. It can help understanding his arguments.

The Sound Lab like many ESLs has a high impedance at low frequency. It has a bit of a bump in the otherwise 8 ohm midrange as it transitions from the bass transformer to the HF transformer, with about 1.5 ohms to 3 ohms at 20KHz depending on the position of the brilliance control. If the amp can't make power into 30 ohms you will get no bass authority. If the amp has lots of feedback, the result is that a 600 watt transistor amp will seem to have the same power as a 150 watt tube amp on this speaker (do the math: if 600 watts into 6 ohms, will be 300 into 16 and 150 into 32 ohms). Additionally it is hard for SS amps to not sound bright on the speaker due to the lower impedance (producing more power from the amp- there is over a 10:1 impedance variation with this speaker; it flies in the face of the Voltage Paradigm tradition as it seems to challenge transistor amps more than tube amps due to this variation) at high frequencies. This speaker seems to sound its best if the amp can make the same power at 30Hz (30 ohm load) as it does at 15Khz (3 ohm load). So this makes it an example of Power Paradigm technology. Note that we are not talking about a smooth impedance curve nor are we talking about one that is lacking in low impedances. The older Quads- the ESL 57 and 63 behave in much the same way.

FWIW what my article was about was really two things: the issue of equipment matching and the flaws that occur if feedback is added to obtain a Constant Voltage response. Its been a conversation in high end audio for decades. The article just puts some predictability to the results.

 

[Ric Schultz]

Myles,
I have no idea of what you mean by "my definition of transparency". Transparency is "no sound".....not a coloration. The most transparent gear has the lowest noise floor...of course. It allows us to hear all the subtle details.....farther into the soundstage, micro detail, etc. However, it does not add warmth, space, juiciness, distortions (no matter how pleasant) and anything at all.

RE Wondercaps: The first generation were simply Relcaps. the second generation were steel leaded caps made by IMB (I am the person that turned Peter on to IMB...I lived a block from Peter in Berkeley), the third generation had copper leads and after that he developed his own leads (Wonderwire) that he has used since. I think his latest "Stealth Caps" are used in the Audio Research Reference gear. Those second generation Wonders were really bad, as you said....after that they got much better. The Wondercaps in the Modulus preamp I tested had copper leads. There have been at least 10 generations of Wonder, Infiniti, Dynamicap and Stealthcaps. All sounded different from each other. There have been at least 7 generations of Wonder solder too (and all sounded different from one another).

 

[DaveyF]

MEP, you simply have no idea of what you are talking about. And you are a few months older than me, so I have to assume audio is not your first interest.




That is another Logical Fallacy known as Personal Attack, You have no idea what peers might think, the comment is simply intended to discredit without proof of any sort.

[/B][/I][/U]

+1

 

[KeithR]

Mep- have you tried a no feedback amplifier? Ayre for instance

 

[ack]

What the term refers to is that the amp will attempt have the same amount of power output regardless of the load impedance. It will not be perfect at this but there is a range of speaker impedance where this can be true.

Just went through all most recent posts... I am not sure what the disconnect between you and mep is... I understand that a lot of SS designers tout their amps' higher output as impedance drops, but frankly, I don't believe many of those claims. Reason being that, IFF a SS amp's output impedance is low - say 0.1 or lower - then by Ohm's law of parallel resistances, a speaker connected in parallel to such an amp may have varying impedance, say from 1 to 16 ohms, w/o any significant consequence, because per Ohms law, the overall impedance of the system will vary only from 0.090909 to 0.099379 ohms (given the numbers I just used). Therefore, I fail to see how such amps could possibly be putting out widely varying power with such a widely varying load when the overall system impedance remains so linear (exactly because of the amp's low output impedance), and how they cannot be branded Constant Power? Perhaps I am not understanding the impact of feedback here, so perhaps you are postulating that feedback somehow widely and wildly affects the output impedance of an amp - thus, of the overall system's impedance - and, in this case, an SS amp??? If that's the case, what are the formulas that govern amp-speaker impedance when you factor in feedback?

 

[mep]

I thought the crux of Ralph's position with using feedback is that it adds odd order harmonic distortion. I don't 'get' his use of the term 'strawman argument' for anything I said either.

 

[Atmasphere]

I thought the crux of Ralph's position with using feedback is that it adds odd order harmonic distortion. I don't 'get' his use of the term 'strawman argument' for anything I said either.

Why don't you read the website at the link I posted then??

How the Strawman works in your case: I pose that you can get flat frequency response if you use a high impedance amp with certain speakers designed for it. You translate that to say *all speakers* and then easily knock that argument down. Problem is, you didn't address my position at all, instead created one of your own liking. That is how a Strawman works.