Return of the Vacuum Tube

 

Platinum Member
Username: Jan_b_vigne

Dallas, TX

Post Number: 17265
Registered: May-04
Peer inside an antique radio and you'll find what look like small light bulbs. They're actually vacuum tubes—the predecessors of the silicon transistor. Vacuum tubes went the way of the dinosaurs in the 1960s, but researchers have now brought them back to life, creating a nano-sized version that's faster and hardier than the transistor. It's even able to survive the harsh radiation of outer space.

Developed early last century, vacuum tubes offered the first easy way to amplify electric signals. Like light bulbs, they are glass bulbs containing a heated filament. But above the filament are two additional electrodes: a metal grid and, at the top of the bulb, a positively charged plate. The heated filament emits a steady flow of electrons, which are attracted to the plate's positive charge. The rate of electron flow can be controlled by the charge on the intervening grid, which means a small electric signal applied to the grid—say, the tiny output of a gramophone—is reproduced in the much stronger electron flow from filament to plate. As a result, the signal is amplified and can be sent to a loudspeaker.

Vacuum tubes suffered a slow death during the 1950s and '60s thanks to the invention of the transistor—specifically, the ability to mass-produce transistors by chemically engraving, or etching, pieces of silicon. Transistors were smaller, cheaper, and longer lasting. They could also be packed into microchips to switch on and off according to different, complex inputs, paving the way for smaller, more powerful computers.

But transistors weren't better in all respects. Electrons move more slowly in a solid than in a vacuum, which means transistors are generally slower than vacuum tubes; as a result, computing isn't as quick as it could be. What's more, semiconductors are susceptible to strong radiation, which can disrupt the atomic structure of the silicon such that the charges no longer move properly. That's a big problem for the military and NASA, which need their technology to work in radiation-harsh environments such as outer space.

"The computer you and I buy is what NASA buys, but they won't want it exactly the same way," says Meyya Meyyappan, an engineer at NASA Ames Research Center at Moffett Field in California. "It takes them a few years to radiation-proof it. Otherwise the computer you put in the space shuttle or the space station basically will get zapped and stop working."

The new device is a cross between today's transistors and the vacuum tubes of yesteryear. It's small and easily manufactured, but also fast and radiation-proof. Meyyappan, who co-developed the "nano vacuum tube," says it is created by etching a tiny cavity in phosphorous-doped silicon. The cavity is bordered by three electrodes: a source, a gate, and a drain. The source and drain are separated by just 150 nanometers, while the gate sits on top. Electrons are emitted from the source thanks to a voltage applied across it and the drain, while the gate controls the electron flow across the cavity. In their paper published online today in Applied Physics Letters, Meyyappan and colleagues estimate that their nano vacuum tube operates at frequencies up to 0.46 terahertz—some 10 times faster than the best silicon transistors.

The team's device isn't the first attempt at miniaturizing the vacuum tube. Contrary to previous work, however, the researchers do not need to create a "proper" vacuum: The separation of the source and drain is so small that the electrons stand very little chance of colliding with atoms in the air. This is a huge benefit, says Meyyappan, because it opens the door to mass production.

Electronics engineer Kristel Fobelets at Imperial College London agrees. "Vacuum technology within a semiconductor fabrication line would make fabrication costs very high," she says. Still, she cautions, the nano vacuum tube is more of a "proof of concept" than a working device, since its operational requirements do not yet match modern transistors. As one example, about 10 volts is needed to switch the device on, whereas modern transistors operate at about 1 volt; in this respect, the nano vacuum tube isn't compatible with modern circuits.

Even so, the potential is great, says Meyyappan. The new vacuum tube's inherent immunity to radiation could save the military and NASA a lot of time and money, while its faster operation makes it a rare candidate for so-called terahertz technology. Sitting between the microwave and infrared regions of the electromagnetic spectrum, the terahertz region can pick out the "fingerprints" of certain molecules. The technology could therefore be used at airports to safely scan for illicit drugs, for instance.

So are vacuum tubes poised to make a comeback? Meyyappan thinks so. "We are combining the best of the vacuum," he says, "and the best of what we have learned in the past 50 years about integrated-circuit manufacture."

http://guitarsquid.com/posts/vacuum-tubes-may-be-making-a-comeback-soon-but-you- probably-wont-recognize-them/10399/?frame=1


.
 

Gold Member
Username: Magfan

USA

Post Number: 2775
Registered: Oct-07
Jan, I've seen the nano vac tube article and the author made at least one technical mistake of serious proportions.

"Vacuum technology within a semiconductor fabrication line would make fabrication costs very high," she says

The implication was you don't use vac to make SS devices.
This of course, is 100% wrong. Vacuum (Hi-Vac) is used in etching machines, implant, CVD and Metal Deposition (sputter)

The idea of the 150nm pit etched in phos doped silicon....and it etches very quicly using RF of either Reactive ion or 'sputter' etch, is that of 'Mean Free Path'. The electron will hit the wall before collision with an air atom. This is the same phenom which sets limits on how good a vacuum can be attained and places a premium on shape / design of such chambers used for vacuum processes.

The PLUS? Well, you can use semiconductor processing techniques to make 'arrays' of such devices and use standard metalization and polysilicon(doped) interconnect techniques to make what ever you want. Maybe bunches of such devices in parallel could be a RF power device on a chip.

I'd like to see a cross sectional drawing of these devices. It is possible, I think, using the right start silicon and BACKSIDE wafer thinning and implant to make the backside of the device the DRAIN and thereby make cooling easier and raise powerhandling. This would also lower RdON which is good to lower resistive self heating.

Good stuff.
 

Platinum Member
Username: Glasswolf

Columbia, South Carolina America

Post Number: 14675
Registered: Dec-03
Nothing wrong with tubes, and they are still widely in use in higher end audio. I myself am in the process of building a valve RIAA phono pre-amp. I may eventually build a pair of monoblock Oddwatt tube amps as well. Plenty of people, particularly in the turntable world and amplifier world still using lots of vacuum tubes.

The majority of the market uses solid state and IC tech now simply because it's cheap and mass-produced, and doesn't need any sort of bias calibration as it ages.
 

Platinum Member
Username: Jan_b_vigne

Dallas, TX

Post Number: 17267
Registered: May-04
.

"'Vacuum technology within a semiconductor fabrication line would make fabrication costs very high,' she says"

'The implication was you don't use vac to make SS devices."


Don't know, leo, the article is rather sketchy since this doesn't seem to be anywhere close to production. I assumed the issue would be how vacuum tubes are evacuated of their gasses vs the vacuum process used in current chip manufacturing.


.
 

Platinum Member
Username: Jan_b_vigne

Dallas, TX

Post Number: 17268
Registered: May-04
.

At the recent Lone Star Audio Fest, I spent more time in the Oddwatt room than any other. They were displaying with a pair of discontinued Martin Logan box speakers and the sound was exceptional for the price being asked. By far not the best sound in the show but certainly one of the best for the budget category. Anyone with the skills to assemble a kit should, at least, give Oddwatt a look. Their tube selection was rather common, 12AX7's and EL34's in an industry somewhat obssessed with using oddball video tubes for audio circuits but the sound of their gear would be a good indication why 12AX7's and EL34 were ubiquitous during the days when all we had were tubes.

.
 

Gold Member
Username: Magfan

USA

Post Number: 2778
Registered: Oct-07
If you want to make a SS device with a vacuum sealed area, THAT could be a probem.
The article gives the dimension of the void in the silicon, at which size you don't NEED a vacuum because of mean free path considerations.
Pretty cool, actually.

Vacuum technology is, and has been pretty stable for a long time. The design of vacuum equipment, made to process silicon wafers is a known set of principles and parameters. Tubes, too.

The glitch? Well, you have to know what to do with the equipment and how to design a process of manufacture. Times? Tempertures? Vacuum level? Material? Cleaning parts? Equipment Maintenance?

Vacuum tube 'evacuation' is done with the same pumps as SS manufacture. The difference? Well, tubes are a fairly open shape and are pumpable. The addition of a getter helps get the tube to the final desired vac......what we called 6 or 7 range. (1e-6 Torr, for example) Since glass is malable, you can than 'pinch' it off hot and seal it. Don't most tubes have a little dimple?
 

Platinum Member
Username: Jan_b_vigne

Dallas, TX

Post Number: 17269
Registered: May-04
.

Yes, and that's generally how the evacuation process is completed. Heat the tube envelope then slam the door shut behing all those dumb gasses that took off into the open air. Hah! Bet they're sorry now!!!
 

Gold Member
Username: Superjazzyjames

Post Number: 2007
Registered: Oct-10
 

Gold Member
Username: Superjazzyjames

Post Number: 2013
Registered: Oct-10
Smiley faces! Non-commital smiley faces! Cheers, cheers baby!
 

Gold Member
Username: Magfan

USA

Post Number: 2887
Registered: Oct-07
I'd love to work in the R/D end of this process.
 

Gold Member
Username: Superjazzyjames

Post Number: 2020
Registered: Oct-10
I'd like to know how well they'd work as audio amps. We're probably talking big bucks!
 

Gold Member
Username: Magfan

USA

Post Number: 2891
Registered: Oct-07
semiconductor processing is mass batch (lots at once) and so not all that expensive for 'commodity' devices.
Now, depending on how complex the device is and a few other non-trivial issues, it may be expensive or not.
If demand is there, costs will come down. Some devices I used to make were run once a year....and only 1 lot even than...and so were a little more dear. If the device needs some kind of special processing or even special equipment costs will rise.
If it can be fabricated on standard semiconductor equipment, it will depend on number of layers and other fabrication complexity.
 

Gold Member
Username: Superjazzyjames

Post Number: 2022
Registered: Oct-10
Now you have me wondering if it's possible to get high end sound from these nanos more affordably and how long it might be before we see this happen, if ever.
 

New member
Username: Da_wei

New Zealand

Post Number: 2
Registered: Oct-12
10V may be big for digital, but its 'small bikkies' for audio. Could they be made to cope with, say 45V? Imagine, a 100W Direct Coupled 'valve' amp, maybe running in class A!
OK, I know they probably aren't exactly power valves - at least not yet...
 

Platinum Member
Username: Jan_b_vigne

Dallas, TX

Post Number: 17471
Registered: May-04
.

I'm not sure what you're wishing for
, but try this;
http://www.atma-sphere.com/Products/#MA-1

The Atma-sphere amps are designed following the basic rules of the Julius Futterman OTL amps. OTL stands for "ouptut transformer-less" which - obviously - means there are no output transformers attached to the power tubes' output. This provides a
"direct coupled" approach to tube power amplifiers more or less similar to what you would find in a high powered solid state amplifier using multiple and paralleled output devices per channel.

To achieve an output impedance from a vacuum tube suitable for direct connection to today's average (reactive load) consumer loudspeaker requires either a suitable intermediary output transformer to lower the impedance of the tube or using multiple tubes connected in parallel which eventually, given enough tubes, will lower the output impedance of the whole circuit. As you can see in the photo, a total of 14 tubes per channel in this amp to achieve 140 watts into 8 Ohms.

This parallel connection scheme still leaves the amp with a rather high (and very reactive) output impedance by today's standards and careful matching with suitably non-reactive and somewhat high sensitivy/high nominal impedance speakers would be manadatory if the overall frequency response of the system were to remain even close to stable. This matching is more or less the same required of most vacuum tube power amps using output transformers, so the benefits of OTL operation are somewhat negated by their more significant trade offs. The argument can easily be made that output transformers/autoformers are more desireable than direct coupling - even when using solid state components - since they can, when designed and executed well, serve as a more constant output impedance source through the entire frequency spectrum and also buffer the tubes from many of the problems inherent in and common to many of today's loudspeakers. OTL's have, therefore, gained the reputation of more a "Look what I can do" amplifier than a useful design alternative.

Running in class A operation means each tube conducts the full sinewave rather than only half as would be the case in a push pull amp. With no point in its operation where the tube is effectively "switched off", the tubes conduct an enormous amount of heat typical of any true class A circuit. Read any review of the amps and you'll see most any listener would find these amps to be quite undesirable during even a Canadian summer as they serve effectively as very good - though energy-wise very poor - room heaters. These are, then, amps you might care to listen to a few months out of the year.


45 Volts though? No, more like 450-600 VDC through the amp's main circuitry. Small tubes can be run at relatively low voltages as many were designed for and adapted to use in mobile communications systems for in field use during the wars before solid state existed. Of course, most of these tubes had very limited power output meant to run off a rechargeable battery of the time and typically are not well suited to home audio use today.


A designer could, however, run a group of, say, 12AX7 dual triode's or, better yet (in someways), EL84 pentodes, together in a parallel feed amplifier to end with a product similar to the Manley Stingray;
http://www.manley.com/mist.php The Stingray, however, is neither an OTL nor a class A amplifier.

But 45 Volts? Even a switching mode power supply outputting 45 Volts would be somewhat sizeable. What's the magic in 45 Volts?



.
 

Gold Member
Username: Magfan

USA

Post Number: 2913
Registered: Oct-07
diodes / SCRs / IGBTs power MOSFETs are routinely are rated up to 1200v.
I helped manufacture IGBTs rated to 1200v.

OTOH, Logic is low voltage. I think old TTL was 5v while modern computers? much lower.

That means that the insulator between Source and Drain is 'good' to that voltage while trying to maintain low voltage turn on. (Gate parameters...mainly oxide or insulator thickness) The resistance of the device while in the 'on' state will determine how much power it turns into heat. Various leakage measures..for example, source to drain when off will determine other parameters and suitability for audio applications.
The ON resistance of semiconductors is one measure that helps determine the Damping Factor of an amp.
 

Gold Member
Username: Dakulis

Spokane, Washington United States

Post Number: 1326
Registered: May-05
Guys, guys, guys, especially you Leo.

You had me at "tubes".

Then, you lost me at: "The idea of the 150nm pit etched in phos doped silicon....and it etches very quicly using RF of either Reactive ion or 'sputter' etch, is that of 'Mean Free Path'."

WOW, I'm such a noob when it comes to the physics and electrical engineering part of this hobby. I guess I'm just the "special" audio guy.
 

Gold Member
Username: Magfan

USA

Post Number: 2928
Registered: Oct-07
Dak
Mean free path is just how far an ion / particle / photon travels in a vacuum between impacts which change it's direction, absorb energy or induce noise. Longer is usually better. And since NO vacuum is perfect, it has to happen......eventually
That's why Vacuum tubes. All the phenom spoken about will work at atmosphere, just not as well. And why once a tube has been evacuated and sealed, a small getter is left behind so that when the tube is activated for the first time, it'll consume the last of the oxygen and improve the vacuum level. At very small distances, in the micron range, the chance of striking something between 'source' and 'goal' diminishes. If if you are in full atmosphere.
When I looked into vacuum tubes, I was surprised to learn just how good the vacuum level was. We called it '7 range' which is pretty darn low.
Don't feel bad. I worked semiconductor processing for about 3+ decades and worked R/D with some pretty smart cookies. I'm undegreed, as a disclaimer.

Can a hybrid tube/SS device be made? I'd like to see where the current state of the art lies. If you need a larger volume of vacuum, you'll have trouble 'sealing' the area, but that's just my first pass thought. That is, IF you need such a large space.
 

Platinum Member
Username: Jan_b_vigne

Dallas, TX

Post Number: 17488
Registered: May-04
.

"You had me at "tubes".

Then, you lost me ... "





We had you? I didn't know anyone would be available to be "had".


So, what are you available for, dak? That "tube magic" all the magazines write about?

I thought you had some tube magic with the Unico but decided to upgrade?



You should know I recommend everyone acquire as much technical knowledge and critical thinking skills as possible before they get too confused by all the competing voices. However, you don't need to know much about tubes to make them work in an audio system.

The major failing when using tubes is, IMO, not understanding how a 100 year old technology mates with today's components. This is especially true when trying to use a tubed amplifier with many of today's highly reactive loudspeakers which are not well suited to a transformer coupled amplifier. If you ignore the basic rules, then you're rather certain to be disappointed in what you thought was going to be "magical music" suddenly pouring forth into your room.

But that's what the forum is for, to provide basic guidance and try to keep people from doing dumb things.



I posted the article because I thought it was interesting, not because I was trying to "get" anyone. Those of us who use and prefer vacuum tubes are always interested when someone new promotes the technology. There are obvious advantages to using tubes and we like to see people talking about those advantages. We find it interesting that, after 106 years of use, research still continues into the development of tubes - and triodes at that! One the other hand, while solid state is the dominant technology of most electronics today, most folks think research into actually improving solid state ceased about thirty years ago. Since then everything solid state has been about making larger and larger integrated circuits which are smaller and smaller and which process purely digital data. This ties into another recent thread in the CD forum and many of us are not totally convinced the last half century of audio progress has actually resulted in much forward movement.



.
 

Gold Member
Username: Magfan

USA

Post Number: 2930
Registered: Oct-07
Jan, I know you are sincere, and may not even believe it, but trust me, research and improvement in SS Power devices is an ongoing thing with room for improvements yet. 'Most' folks, are, as usual, wrong.
One facet which is subject to improvement is one thing at which SS excells. The idea of Resistance of the Device when ON.....Rdon, for short. This is one of the parameters which is measured during final test when all devices on a silicon wafer are 'tested' Some devices are tested on a sample basis, while other devices are 100% tested. Some small wafers....which are 6" these days, will hold in the THOUSANDS of possible die of which you'd expect 90%+ to be good. Therefore, a sample is OK and you can package 'em all from passing wafers and weed 'em out at device test.
Unique chemistries are being investigated. Copper instead of Aluminum interconnects. New geometries of fabrication. Cost lowering thru improved process speed and yield.

Nope, people are misled and just like tubes are magic, so too, does MOSFET or BiPolar or JFet or whatever.....have the Mojo. But, as usual with any technology, it isn't the device, it's the implementation that counts.

And While you may or may not agree with progress having been made in the last half century, no doubt in my mind that progress can be made. Experts are traditionally wrong about predictions. Time was, experts thought you couldn't exceed the speed of sound. Just look at the record of Bell Labs. Basically a patent-a-day for several decades. Many with practical but unforseen applications. Even the transistor which was sold off since nobody could figure out quite what to do with it.

http://en.wikipedia.org/wiki/Bell_Labs
 

Platinum Member
Username: Jan_b_vigne

Dallas, TX

Post Number: 17489
Registered: May-04
.

Leo, I know you believe you are right, but I think we are once again at cross purposes and what you've posted merely confirms what exactly I have posted.

"One the other hand, while solid state is the dominant technology of most electronics today, most folks think research into actually improving solid state ceased about thirty years ago. Since then everything solid state has been about making larger and larger integrated circuits which are smaller and smaller and which process purely digital data."



Transistors are taught in engineering schools as the history of the field, not the future. By this I mean there is nothing earth shaking or future directing that is occurring in discrete transistor technology. That period existed for roughly the 25 years after discrete transistor technology emerged from the labs and into the day to day world of electronics manufacturing. By the mid-1970's both discrete transistor technology and analog media were being displaced by what I described in my post; a shift toward smaller and smaller yet more highly complex integrated citruits tasked with digital data processing.

I find it a bit astounding you seem to deny this while stating more or less the exact same position in your post. Where are your exact points of departure from what I have said? Or, from the reality of how electronics exist today?

Show some proof of what you claim, leo, not the technical "Rdon" or talk of wafer technology. Wafer technology is integrated circuit - if not "nano" - technology as described in the op. Show me where discrete transistors are still be driven by furthering research. Show me where digital data has not largely replaced analog technology in the vast sea of consumer choices. Please, show me because I certainly don't see it.

That is my point with regards to "solid state", leo; discrete transistors are ancient and there is no real world application in which they are the dominant contemporary technology other than the rarefied air of, say, high end audio. And what have we learned over the last thrity years in high end audio? To tell the customer your product is built with discrete transistors and not integrated circuits. Why? I'd say because integrated circuits simply dominate the electronics market and so few components are constructed with discrete transistors.

Certainly nothing in video is made from discrete transistor technology. Only the few hold outs in high end audio adhering to what "sounds good" are using anything other than integrated circuits passing digital data streams at faster and faster rates. That small maybe 2% are still tied to their analog recorders and playback equipment. But such recordings from the major and minor labels are distinct "throwbacks" to a different era - more concerned with what "sounds good" than to what sells. Or, more appropriately, to what makes the most profit because analog and non-solid state have been proven in the market to be money loosers.

Certainly no iPod, iPad or iPhone, Slingbox, CD/DVD/BluRay, fibre optic, information exchanging, medical or weapons technology would exist today if we had not left discrete transistor technology and analog reproduction circuits behind the curve of development and research.

That is what I see happening. That's what I would guess "most people" would recognize as the last thirty to forty years of solid state technology.

That the few hold outs to what "sounds good" still cling dearly to anaolog circuits and ancient technology proves, IMO, my contention very little has come of the measurement side of that last fifty years of supposed "progress". Meaurements have been devised to keep up with the technoolgy, I'm not in denial of that. However, new measurements meant for new products - or even just new theories as we see in the op - do not necessarily make for improved results. Just different results using different measurements for different technologies.

CD sound is "perfect forever" because the measurements say it is, but not the listening experience. From the moment of its arrival on the consumer market - well before CD's and back into the dark ages of "digitally mastered" LP's - digital sound reproduction has faced an uphill climb which, nearly forty years after the fact, it has not fully achieved. What are the two most familiar phrases in high end audio reviewing? That any component produces sound values which are both "analog" and "tubelike".

Read any review of any DAC and those are the two terms you're likely to see to alleviate any fears a prospective buyer might have over their sometimes quite literal distaste for anything which sounds "digital" or which exhibits "digititus". So over the last thirty years of CD technology, the aim of virtually every "refinement" of the integrated circuit based digital playback system has been to eradicate the ubiquitous and oft times unpleasant sound of digital music reproduction and to assure the buyer the product is assembled using discrete - not integrated - circuit technology throughout the analog output stages. In the high end consumer audio market, at least, the industry is always looking backward towards analog and vacuum tubes to find what "sounds good". This is pursued even while the industry is constantly pushing the boundaries of what digital data processing through ever smaller and ever more complex integrated circuits can accomplish.

Quite frankly, leo, I don't see how you can deny that fact.


As an aside, leo, I believe what dak was referring to when he posted, " Guys, guys, guys, especially you Leo", was your frequent dives into the very deepest weeds of things such as "RDon" and "gate parameters". I do encourage everyone to be as familiar as possible with the technology they are employing within ther systems. I do, however, feel most folks such as dak find your explanations a tad beyond what they are ever going to need to make an audio system operate well. You know this stuff and it's great that you do. But, I'm fairly certain no one is going to ask any audio salesperson to compare the "Rdon" value of two amplifiers.

.
 

Gold Member
Username: Magfan

USA

Post Number: 2931
Registered: Oct-07
As near as I can figure, Rdon is what eventually turns into damping factor. NFB also is a large contributor but a seperate chat.

We must keep digital processing and CPU / Logic applications seperate from Power applications.
While they operate in a similar manner, power devices to actually drive speakers or even motors, for that matter, are a different kettle of fish.

The First couple generations of Digital gear, (until....say.....'87 or '88) were just as flawed, but in different ways, as the first generations of SS amp / pre / tuners. My Philips Gen1 player was Golden compared to the Sony of the same time frame, yet falls far short of modern standards.
I suspect that tubes didn't just fall from the sky perfected, either. If I remember correctly, it was good-hearing people that pointed out to Sony just how BAD their original CD player was.......and it was withdrawn from market.

Any technology will have teething problems. You can't name a tech which started mature or didn't have plenty of room for improvements.

Don't sell the future of measurement short. Regardless of how it has performed up to now. Or how you feel it has performed up to now. I rather suspect we are not getting the whole story from you, but rather filtered thru persons of like mind. This is OK with me.....but one thing I know for certain is that the 'rest of the story' is what I'd be researching if so inclined. And I'm afraid I would go into minutia. People must have been driven, at least in audio, to try to measure just what they were hearing. And therefore to try to analyze and improve. It's human nature.

My knowledge of how semiconductors and power devices are processed / manufactured is indeed a 2 edged sword. Some of the stuff I read makes me laugh as much as any hifi myth you can come up with.
 

Platinum Member
Username: Jan_b_vigne

Dallas, TX

Post Number: 17490
Registered: May-04
.

"If I remember correctly, it was good-hearing people that pointed out to Sony just how BAD their original CD player was.......and it was withdrawn from market."



Having sold first generation CD players I would say that's not the real story. Sony and Philips were convinced their product was "perfect sound forever". It may have been Techniocs who used that advertising slogan, but digital audio had been around for a decade prior to the release of the CD format. As I said, digitally mastered LP's had already established what to expect from CD. But those companies with an investment in digital technology (Sony, Philips, Yamah and Technics to name the most prominent in consumer audio) brushed aside the criticisms of harshness, brittleness, lifeless musicality, etc ascribing them instead the the inferior the analog playback systems of the day as the digital juggernauts forged on with the development of a new digital format. That it was technically outdated by the time of its official introduction didn't seem to slow anyone down. Redbook standards had been firmly established by 1976 - the time of the Atari 400 video game - Pong in black and white. By 1982 and the introduction of the first generation of CD players Atari had virtually disappeared from the market. Super Mario Brothers in four colors had taken over the market. And there was the 1982 equivalent of Pong waiting to be sold to the audio consumer.

The first generation of CD payers actually did somewhat well as far as sales were concerned. With the limited library of discs to choose from - most of which were AAD recordings - classical music lovers embraced the format long before the general public did. No more surface noise from an LP and an entire symphony on one disc that didn't require cleaning or special care and, most importantly, did not require being flipped after each movement. The newness of the sound sold the product despite the average $1k in 1982 dollars pricing. We couldn't get enough of them fast enough to satisfy the demand.

However, CD was produced by the mass market companies and the mass market companies do not bow to consumer opinion, they feel their purpose is to shape consumer opinion. No mass market company embodied this mass market attitude more than Sony who had already shaped the market with new technological marvels for the previous twenty five years. Nothing was "withdrawn" from the first generation CD market. In fact, the players couldn't arrive in the stores fast enough to satisfy the desire for the new and well touted format. People who had no CD's bought machines in hopes of being able to secure two or three discs to play.

Model changes occurred rather rapidly and spread the pricing over lower ranges. One of the first Yamaha CD players I purchased in the 1980's sold for $400. Not checking the inflation schedules, my guess is that would be the equivalent to about $1200-1300 today. For a Yamaha!


No the mass markets did not withdraw any CD players from the market. We were selling virtually all the players we could get from the companies well into the early 1990's. As with all mass market gear, the changes and "improvements" were coming fast and furious and, yes, the market was responding to the generic complaints regarding digital sound but their "fixes" - such as they were - existed all over the board. 18 bit, 24 bit, 8 times oversampling, 16 times oversampling, anti-aliasing filters, one bit, brick wall filters, no filters, etc. Models changed quickly in response to the sales pitch of the competition, not necessarily to actually improve the sound.

But from the first day CD players hit the shelves until the early 1990's most dealers were in a goldrush mindset as CD became the hottest item with the greatest revival of profit the consumer audio industry - which had just recently gone through the first wattage/price wars in its history - had ever seen. There had been nothing like it for the retailers before and I can think of nothing since which has matched CD"s intial and long running profitability.



.
 

Platinum Member
Username: Jan_b_vigne

Dallas, TX

Post Number: 17491
Registered: May-04
.

"I suspect that tubes didn't just fall from the sky perfected, either."


I suppose that depends on your your definition of "perfected". Certainly, deForest's original invention took several years of development before the triode vacuum tube became an acceptable device for use in audio. However, in 1906, everything took several years of development before it became an acceptable device for audio use. There were no microphones at the time and vocalists sang - "shouted" - into a megaphone until the early 1920's when "electrical recordings" began to displace the purely mechanical ones. The use of vacuum tubes in rudimentary audio circuits did not occur until the latter half of the 1920's.

By that time "talkies" were hitting the theatres and telephones were becoming a far more common item in the everday household. - at least for the upper 10% of society they were. "Pork" was being spent by the government to extend electrical power and communications to the farthest reaches of the back hills. You can still travel to many areas of the nation - West texas for one - and see the lines which were run by the politicians to garner goodwill and votes. It was not uncommon for residents to buy electrica appliances long before and in anticipation of the lines actually reaching to their homes.

Bell Labs, Western Electric, RCA, Edison and Tesla were all in competition during the first quarter of the twentieth century to win government contracts for radio transmisssion and national/international telephone distribution along with military contracts. Audio was being driven by the advances in the communications industry and the profit from "home entertaiment" products was a hot sideline. Radio meant a new wave of entertainers were available for sale to anyone with a tubed or crystal tuner. The popularity of so many entertainers being heard across the nation meant there was a market for "records" made by those entertainers.

With the need for international telephone's longer lines and more powerful radio transmission to reach the outskirts of the state and beyond with advertising promotions came the need for higher powered vacuum tubes than any triode could manage. Tubes using more sections which would develop that power were quickly being developed and by the late 1930's beam power and pentode tubes along with an entire assortment of other styles and combinations of tubes had come about.

For the next twenty years research into vacuum tube technology mostly focussed on making what was available a better product. In Europe tube designers and manufacturers took the beam power tubes from the US and designed KT66's and KT88's as substitues for 6L6's and 6550's. The KT tubes were "kink-less" which became a significant improvement over the standard US tubes. The BBC became the headquarters for audio development in Britain as Germany began invading the Western nations of the continent. Multi-channel sound was being perfected, digital theory was being developed, "tape" or wire recorders were developed for natonal security purposes.

No one of these technologies hit the ground running, they all had development stages and improvement stages. "Manhattan projects" were developed around audio and communication circuits in order to be ahead of the Axis powers' own developments. "Radar" came from such projects. It wasn't until after the war's end that the higher grade of recorders which had been developed in Germany were made commercially available in the rest of the world.

From 1900 to well into the space age there was tremendous energy put into the audio and communications fields. Transistors passed through this development stage rather rapidly as the need quickly changed from audio curiousity to the need for smaller and lighter, yet faster and more powerful, integrated circuits. At that point discrete transistor technology was rapidly being left behind in the research labs.



What does it prove that tubes were not "perfect" when they first became acceptable to the consumer market? I never claimed they were.




.
 

Gold Member
Username: Magfan

USA

Post Number: 2932
Registered: Oct-07
To get back to your original post:
Vacuum tubes have not gone anywhere and therefore have no need to 'return'.

Of course you never claimed tubes were perfect when introduced. The question is what drives improvement. You have said that instrumentation doesn't help. A 'never' statement like that is almost always wrong. And using the past to predict the future is not accurate.

The original Sony player, as I remember, used a single DAC which time shared between channels. It was so audibly flawed that it got Zero traction. The Philips player, which cost me 250$ (not 1k$) was double DAC and 4x oversample at that....though only 14 bit! When I bought my player, I'm sure the 'next gen' stuff was already being shipped to dealers.
The Sony solution introduced some phase / time shift artifacts which people....even 'regular' people could audibly identify and did NOT enjoy. Strike 1 for CD.

Power management is one of the big challenges of our time and is driving improvements in technology. Semiconductor power devices will play a large part. Everything from Compact Flourescent Lights to smart refrigerator motors will depend on power devices. Many times, for sure, working in harmony with low level logic. Don't underestimate the improvements made and being made in power semiconductors. In some ways his has entered the 'incrimental' phase as opposed to the 'fundamental' phase. But wait for GaN devices or Si-Ge hetrojunction devices. They may even prove to be good for audio uses, not just power or logic applications.

Jan, in brief, please describe how you think products get improved. In this case, use your most familiar area, that of stereo. How has anything get better? Strictly by ear? Strictly by measure? (I even reject that approach) or by some combination? If some combination is a good and decent answer, what path leads to the future?


Nope, what's past is past. The future is not predictible, unless you are willing to talk about exponential growth in a finite resourse world. But that's a totally different chat.
 

Gold Member
Username: Magfan

USA

Post Number: 2934
Registered: Oct-07
Nelson Pass ( you knew we'd get to him sooner or later, right?) is maybe the best combination of Measure, Listen, knowledge, progressive and experimenter out there. Certainly on ANY reasonable short list.

Here is a little something he came up with to answer some of the SS objections....while emulating a tube characteristic. Look at the detail of measures Pass goes thru and Understands.........

Just for your consideration:

http://www.firstwatt.com/sitintro.html

And while Yamaha and Sony Bothe experimented with these type devices, it took Pass to take it to the next level....
 

Platinum Member
Username: Jan_b_vigne

Dallas, TX

Post Number: 17493
Registered: May-04
.

"You have said that instrumentation doesn't help."


I never said "never" in the context of this dicussion or on either thread, leo. Here's what I have said most recently in this thread, "IMO, my contention very little has come of the measurement side of that last fifty years of supposed "progress". Meaurements have been devised to keep up with the technoolgy, I'm not in denial of that. However, new measurements meant for new products - or even just new theories as we see in the op - do not necessarily make for improved results. Just different results using different measurements for different technologies."



"And using the past to predict the future is not accurate."


I don't believe I've done that either. However, the past is where discrete solid state technology now exists. The future of discrete solid state circuitry has been predicted to remain relatively isolated in niche marlets such as high end consumer audio. Otherwise, it's a more or less dead issue and new measurements are not being devised to improve discrete solid state circuitry itself. The markets, the industires, and the technology have all essentially stopped researching ways to improve discrete solid state circuits in favor of researching smaller and smaller, more powerful and more complex 'integrated" type circuits. These circuits process digital data rather than analong signals. A purely discrete transistor circuit could not easily provide either the speed nor the miniaturization required to build a first generation CD player. How many discrete components would it require, leo, to build a simple 16 bit DAC?


That's what I've said, leo. New tehcnologies have developed their own new measurements as developments have called for new tests. Obviously, measurements serve to create the technology and measurements also determine whether theory meets reality. Have these new measurements improved consumer audio very much? I would say not much at all - in fact, I have.


Vacuum tubes and analog (LP) playback are stilll making advances thanks to the power of solid state, digital circuitry. A new vacuum tube type was recently developed, the KT90. Turntable playback has improved dramatically over the last, say, fifteen years.

What has improved discrete solid state circuits in use in high end audio has been the passive parts and the components of the circuit long thought to be passive in their effect on music reproduction. Most of these improvements, however, are not "measurable" in the traditional sense and are disputed by the severe objectivists.


Please show me proof that discrete solid state technology is alive and well in the mainstream electronics market rather than just again repeating your opinion I have it wrong.


.



.
 

Platinum Member
Username: Jan_b_vigne

Dallas, TX

Post Number: 17494
Registered: May-04
.

"The original Sony player, as I remember, used a single DAC which time shared between channels. It was so audibly flawed that it got Zero traction. The Philips player, which cost me 250$ (not 1k$) was double DAC and 4x oversample at that....though only 14 bit! When I bought my player, I'm sure the 'next gen' stuff was already being shipped to dealers.
The Sony solution introduced some phase / time shift artifacts which people....even 'regular' people could audibly identify and did NOT enjoy. Strike 1 for CD."



Obviously, as I have stated, that's not the story I remember regarding the impact the first generation CD players had on the masses.

On October 1, 1982, Sony released the first ever CD player in Japan. Seven months later, it finally arrived in the U.S., complete with a vast library of 16 available albums. Its price? $1000.


Regarding the impact of the first generationS of CD players, please refer to my above post.

I wasn't selling Philips at the time of their first CD launch. If you say it cost you $250, leo, then that's what it cost. It would, however, have been very difficult to sell the numerous $700-1000 players against the Philips at that price. No?

You are correct the original Philips had dual DACs, it had to since they were only 14 bit processors. And it also had to oversample to get the useable sample rate up to CD's standard 16 bit. Though neither system used processors which were capable of anything more than a true 12 bit resolution.

Again, my recollection of what I was selling at the time was not that Sony had a balk in the CDP-101 but that most stores could not keep CD's or CD players in stock for years. You can research the impact of the CD on both the music and the audio industry if you'd like, leo. When people bought CD's they wanted a player. When people had a player, they wanted to buy CD's. It was a profitable period for both industries beginning with the 1982 introduction of the CD through to the early 1990's.



"The Sony solution introduced some phase / time shift artifacts which people....even 'regular' people could audibly identify and did NOT enjoy. Strike 1 for CD."


I have never claimed any first generation CD player sounded good. I have said peope bought them as fast as the manufactures could turn them out. This is the mass market where CD's first appeared and new models were being designed almost as fast as any other mass market item. Ther was a lag between the first CD player hitting Japan and then hitting the US. In that time, new models had been readied.

Thefore, I don't know what your point is with that statement. All CD players sounded rather bad until the Meridian showed up in 1986 or so. That didn't stop people from buying them and it certainly didn't stop the manufacturers from releasing new models.



.
 

Platinum Member
Username: Jan_b_vigne

Dallas, TX

Post Number: 17495
Registered: May-04
.

"Jan, in brief, please describe how you think products get improved. In this case, use your most familiar area, that of stereo. How has anything get better? Strictly by ear? Strictly by measure? (I even reject that approach) or by some combination? If some combination is a good and decent answer, what path leads to the future?"



I don't know what you're trying to prove, leo. That I'm wrong and you're right? Why?



How do products get improved? I suppose we'd have to agree on what "improved" actually means before we could address how improvement occurs. No?

Since there has been no definition provided, I have to suggest - again - that when a fifty year old tube amplifier can still be a class A recommended product at Stereophile, possibly "products" haven't really "improved" all that much despite the "alphabet soup" of measurements we can now perform on most products.

Are computers "faster" than they were ten years ago? In general, yes. What does that prove? That we didn't test the McIntosh tube amp for its RAM data processing speed fifty years ago? I'll grant you that. As I said, new technologies bring with them new measurements. Still that fifty year old McIntosh tube amp is as "good" as most of its modern and supposedly "improved" competitors when it comes to what the amp was designed to accomplish - reproducing interesting music in an interesting manner.

Here's what has "improved" in the building of that "new" fifty year old design. The circuit has remained virtually untouched other than to provide the upgradeability to use balanced lines with the amp. That's a technical value not an "improvement" per se. Balanced lines have offered identical gains for the last fifty or more years, they simply were not in common use in home audo gear fifty years ago.

While McIntosh always believed in tight tolerances and the use of the best available passive parts in their products, the reputation of McIntosh generally suffered with the high end audio reviewers through the 1980-90's. Their engineers objectively felt "parts were parts" for the most part and they did not use many of the "audiophile approved" passive parts found in the "average" Audio Research or CJ. After the sale of Mcintosh to another ownership things changed and the Mac products eventually included those same passive parts though there were still not objective facts to support their additonal cost.

This was a very difficult time at McIntosh as many of the original or second generation McIntosh engineers/designers were being asked to design circuits with parts they saw little to no value in adding or, possibly, removing.

At that point, after the inclusion of the "improved" parts, the general opinion was McIntosh had "improved" their products and they were once again competitive with the best available. Those parts are being used in the current production MC275. Whether the placebo effect of having specific parts in the design of the MC275 has increased its value or whether the amp actually does a more relalistic reproduction of music is not an issue I can address here.

Why had McIntosh refused for so many years to make the move to "high end" passive parts? Primarily because they could not measure any "improvements" in most of those parts vs the stock McIntosh parts used for decades. Fomer McIntosh engineer Roger Russell's webpage for wire guage is still pulled out without missing a lick when cables are being discussed. His opinion, like Les's, is that anything more expensive than 18AWG wire is a waste of resources when nothing more than 12' runs are being used.

However, it was a French audio writer who first suggested "listening to" cables. He did so due to his perception of differences and improvements when using certain cables but not others. It was primarily listeners who have made the subjective "improvements" in high end audio products since the late 1970's. It has been this way since the 1920's and the beginning of electrified music reproduction really. Someone recognizes an audible problem and sets out to solve it.

Are new measurements used to verify the "improvement"? At times, yes. Measurements such as "TIM" are thought up to explain what has been heard. The horse is now behind the cart and is being pulled along by the process of moving the cart forward in this instance.

Audible problem identified. Solutions are tested until a subjective evaluation suggests a solution has been effective. Once the problem has been resolved, there needs to be a repeatable method of duplicating the result, so a name for the problem and its test(s) are conjured up. Though, subjectively, it is not completely certain the changes made are resolving the the problem because the new tests say it is or whether some other coincidental issue has resolved some other problem which we have now identified as responding to our changes.

Just because we objectvely measure something does not mean we are actually measuring the correct value.

This is how many products came about, by setting out to design something completely unlike what the designer ended up having. How, then, can we say measurements actually improved a product the designer had no intention of designing?



None the less, the vast majority of passive parts being used in today's high end audio market are not chosen by specs or measurements. Many years ago Stereophile interviewed Conrad and Johnson and asked about their approach to bringing a new, "improved" product to the market - in that instance it was the ART pre amp. Their response was to say they first cover the bases with the essential design of the circuit. Or, as JA says of the MC275; good audio design is timeless.

Standard measurements comfirm all is right before they set about making subjective changes to passive parts which alter the voicing of the product until they are both satisfied with the "sound" of the component. For this stage of development, measurements were of little to no value since the parts did not inherently measure as not being identical in conventional tests. Their ears, their honed perception of live music and a familiarity with the "sound" of music were the tools they required to perform the final stages of voicing the component. Objective tests largely fell by the wayside in favor of a desired subjective response.

This is how most high end audio products have been designed for decades. Even the original MC275 tube amp was designed in much the same manner though at the time the wide assortment of passive parts we find today simply were not available. In the end, however, it had to sound as though it came from McIntosh Labs. The other two amplifiers in the line also had to "sound" not like distinct amplifiers but rather as if they could reproduce music as eloquently as did the MC275, just not quite as loudly as the MC275 driving the same speakers.

Once CJ had developed their flagship pre amp through such "tests", they set about designing a less expensive version using trickle down technology. Many of the changes and tweaks made to the flagship would be relevant to the new sibling, but at a lower cost. I think, should you read about the development of the ART II pre amp, you'll find much the same process occurred as the team tried to create a sonically identical product in as many ways as possible given the budget restraints, but did so with fewer parts and (often) less costly parts.

Therefore, it was not the measurements which created or "improved" the CJ pre amps, it was the ability of the designers to know and to design to a perceived subjective measurement which satisfied their ears. In the end, though each line has aduibly improved over the decades, a Mac is a Mac and a CJ is a CJ. It is not inherently how they measure - though they both measure quite well - but how they preform when reproducing music. It is the temporal and dynamic values of music which make an audio component poorly suited to the static measurements of most severe objectivists.




That is a common story in high end audio, leo. In another industry, other tests and other measurements might be used to ensure a
"correct" design. I am relatively certain though that in any product where a subjective evaluation of the product drives sales, at some point the basic design is "improved" through subjective testing of many, many variable parts which are not strictly tested by way of hard and fast measurements. Most wine an cigar retailers do not rely on specific measurements to produce a product. They might use measurements to test for specific values and have those values remain as their target. But most wine and cigar retailers understand how much the intangbles such as weather and soil influence their final product. They might rely on a specific measurement of, say, humidity to protect their investment, but they know within the first moment of a subjective evaluation whether the product hits their mark or not.



Does a chip manufacturer measure? Certainly. Has the measurement improved the product? Once again we would need to agree upon a common definition of what it means to "improve" a product. I do, however, concede new technologies bring with them new measurements. I've never not said as much.

Measurements might confirm that a specific part has been well chosen, but it is not the spec which has initially led to the selection of that part in products which are chosen largely by subjective evaluation.


Have strictly objective measurements therefore made an improvement in the subjective quality of a product? IMO you would be hard pressed to say that they have in most cases and do so with a completely straight face.



YMMV







.
 

Platinum Member
Username: Jan_b_vigne

Dallas, TX

Post Number: 17496
Registered: May-04
.

Sorry, you've once again baffled me with your reference to the Pass amplifier.

Were new measurements actually developed which "improved" the product? Or was the product simply an improvement in a subjective evaluation? Were any of the measurements taken by Pass in the creation of his amplifier "new" measurements? Or, were they merely new implementations of old measurements?

Actually, with cercain speakers, the current sourcing amp Pass had designed would sound no better - and by his own admission, might sound a bit less good - than a voltage sourced amplifier. So what is being tested here? Objective measurements? Yes. But is it is not the subjective evaluations which determine whether the objective tests have been of any value what so ever? Which "test" then has greater weight in the creation of the product or in its eventual improvement? If the objectve tests merely confirm a better audio design as experienced in subjective testing, have the objective measurements done anything to "improve" the product?

Would Pass introduce a product which measured perfectly but did not meet his subjective goals? You and I both know he wouldn't. Once again, leo, I would say that proves my point and doesn't do alot for your's.

Therefore, which measurements are really of value to the subjective impression of music being reproduced by the Pass amplifier? Which objective test has Pass employed which confirms the ability of the amplifier to reproduce "interesting" music?

Once he has performed "good audio design", is the rest not a matter of meeting only his subjective tests? Just as was done with the CJ pre amp?

And, while Pass's design is largely unique in today's market, he was hardly the first designer to think a current sourcing amplifier would perform to a higher standard when driving certain reactive loudspeaker loads. Peter Walker's famous Quad tube amplifier from the 1950's was, to my knowledge, the first successful implementation of such a concept. Therefore, is it not true that the measurements used by Pass to design his solid state current sourcing amplifier are not much more than new technology (transistors vs tubes) bringing new measurements?

Where, if the above question is answered in the affirmative, have measurements actually "improved" the product, leo?




.
 

Gold Member
Username: Magfan

USA

Post Number: 2935
Registered: Oct-07
I know that no progress is being made in power devices, but at least one company I know is content to throw away 12% or so of gross on R+D. And it is a billion $ company.

The point is pass both listens and measures. It is a feedback loop where one reinforces the other.

And AGAIN. making statements about the history of measurement says nothing about the future of measurement. You seem to think that since your opinion and that of a few of your strawmen say it hasn't helped that it will NEVER help. I doubt either part of the statement is true.
I suspect that looking at the history of measuring hifi with anything other than the telescope you use from on high would yield some more concrete data. Not my job.

You continue to characterize my approach as design it on paper and it must be right which is false. Certain rules of thumb need to be taken into account. Heat sink area, PS design for ripple / regulation. All are known or at least not unknowable. You have in other posts ref'd crossover design as being computer aided. Data had to come from somewhere, didn't it?

http://www.atma-sphere.com/Resources/Paradigms_in_Amplifier_Design.php
You may enjoy this, though I suspect you've already seen it.......

And as to Pass? he repeatedly talks about amp speaker matching.
Once he has performed "good audio design", is the rest not a matter of meeting only his subjective tests? Just as was done with the CJ pre amp?
Different words for something I've said before.....and heard before, for that matter. This process goes beyond mere voicing.
 

Platinum Member
Username: Jan_b_vigne

Dallas, TX

Post Number: 17498
Registered: May-04
.

Leo, you are going about this just like a d@mn Republican. You either don't pay attention to what I say or you don't care what I say, you say I said something I never said and then I'm supposed to argue what you said I said when I never said what you say I said.


Here are a few things I have never claimed to be true which you have claimed are my positions or where you have completely ignored what I said even in my last post (it's not like I'm asking you to have a long memory here, leo);

"And AGAIN. making statements about the history of measurement says nothing about the future of measurement. You seem to think that since your opinion and that of a few of your strawmen say it hasn't helped that it will NEVER help."

"I suspect that looking at the history of measuring hifi with anything other than the telescope you use from on high would yield some more concrete data."

"You continue to characterize my approach as design it on paper and it must be right which is false."

"You have in other posts ref'd crossover design as being computer aided. Data had to come from somewhere, didn't it?"






So what am I supposed to do with your distortions of my statements? You claim I am setting up strawmen when this, " ... at least one company I know is content to throw away 12% or so of gross on R+D. And it is a billion $ company.", is a classic case of a strawman which proves nothing in the context of the debate. Actually, it proves nothing at all other than you claim to know of such a company.

You saying something is occurring is not proof something is occurring. Show proof, leo, not empty words.

And when I have I said R&D is not occurring? I have flatly stated on more than one occasion new technologies bring with them new measurements. Which of the previous ten times that I have stated that in the last week have you missed?

But more than anything, you seem to have taken my basic sentiment and completely distorted what I said into something you cannot find any reconcilliation with.

What I had originally said in the previous thread was tubes are continuing to find research aimed at their development even after 106 years of existence and a supposed fifty years of obsolescence. Solid state (as seen in the first decade of its use) however was a relatively short lived technology which quickly moved away from analog signal processing to digital data management. In that process of switching from analog to digital, "solid state" has essentially done away with reasearch and development of discrete transistorized circuits and moved to R&D of integrated circuit technology. Discrete solid state circuits are only still found in a small niche market such as high end audio which has historically only pertained to approximately 2% of the audio market which in itself is only a tiny fraction of the entire "electronics" market. If you still disagree with that statement, leo, show me new products outside of a niche market which look like the inside of, say, an Adcom power amplifier.

Now, when Google, Nextel and Apple combined own more wealth than the cumulative national economies of all nations existing on the Earth, and you can find not a single spec of discrete solid state circuitry in any of their products, what do you have to argue about?

Saying you know of a company investing in R&D does not prove they are investing in discrete solid state circuits. Does it? If that is your proof, leo, then go aead and prove what it is you are setting up. Otherwise, you have established a pure and very simple strawman which is easily knocked down - just as the term "strawman" implies.

What is this mystery company R&D'ing? Show me, don't say soemthing and expect me to accept it as proof for your position which isn't entirely clear since you are arguing something which was never the issue.

I have never said measurements have not been involved in the development of progress in the electronics field. Never!!! If you think I have, show me where I said that. Once AGAIN, my position is encapsulated in the approach taken by McIntosh with their reissued MC275 tube amplifier. Once AGAIN! if a fifty year old tube amplifier design can be considered completely competitive with the very best that has come after fifty years of R&D, how can we say "measurements" have resulted in progress to which we can all agree? I readily agree that measurements in the years prior to the original introduction of the MC275 in 1961 were important to our understanding of electronic theory put into real world application. Where I have I said otherwise? Nowhere.

My point has been, since I first said discrete solid state was a short lived technology, I contend that measurements into those fields where discrete solid state still exists have probably not accomplished much which accounts to "progress". I have said that progress - as far as it is commonly understood - has come largely from the subjective side of the R&D. Is that a statemnent you have completely missed? Or, just conveniently ignored?


" Once he has performed "good audio design", is the rest not a matter of meeting only his subjective tests? Just as was done with the CJ pre amp?"

"Different words for something I've said before.....and heard before, for that matter. This process goes beyond mere voicing."


You contradict yourself with two sentences, leo. First, the idea of "voicing" is something you've "said before". Then, "This process goes beyond mere voicing." Those two statements would not appear to be able to stand side by side as you have placed them.


What "process" are you discussing? And what does it have to do with your position? I'll be honest, leo, I don't have a clue as to what you are talking about when you make those types of statements. I know somewhere in your head you know what you mean, but for the rest of the world not inhabiting leo's brain, they make no sense. "Voicing" is not accomplished in high end audio by taking measurements first. It is accomplished by subjective evaluations which inform the designer whether or not their design as it sits on paper and in the reality of a working model has the ability to reproduce music in the manner to which they perceive live, unamplified (mostly) music to exist. When parts are swapped, they have been selected not so much on their measurements but on their sonic values. A designer doesn't take out "this" value cap and simply insert another value of cap. No, once the rules of "good audio design" have been met, similar values are swapped between various types of caps, not dissimilar values of caps. Many times the sonic values of these passive parts are still hidden from conventional measurements - change one cable running from the input jacks to the circuit board for another cable for example - and the designer is making choces based not on what a scope tells them but on what they subjectively know to be worthy of experimentation. Change a cap from one type to another type, then listen. If the result was more pleasing to the music reproduction by way of subjective testing, make notes of the changes and then, and only then, measure not what has brought about the change but measure only to ensure the electrical values of "good audio design" have not been altered in any significant way. A talented designer will soon learn which caps do what to affect which values of music reproduction. They wil learn to use those caps when "this" effect is desired in the subjective testing of their circuits. But good audio design is a soup pot of ingredients and finding the most pleasant tasting final product is a function of "voicing", which is a series of small tastings after each addition of another ingredient. After they insert "that" cap into any circuit and achieve the desired subjective result, they will test to objectively make certain no aspect of good audio design has been changed. And it probably won't since the caps they are swapping are virtually all objectively identical.

So where are the measurements taking the design, leo? They are coming after the subjective evaluations have confirmed the resullt. They are not driving the objective or the subjective tests. That, as the Atmas-Phere paper points out, has largely been done on paper before any working models have been built up by the R&D department. Good audio design is not a complete mystery. Interesting music reproduction is far more a black art of garnering "this" subjective result from "that" objectively measured design.


That is what I have said several times now, leo. If you would care to actually argue those statements, please procede. If you prefer to toss more spaghetti against the wall to see if it sticks, I'll have to drop out of this because spaghetti as an argument doesn't appeal to me.


As to your linked paper, what? What is the take away I'm supposed to gain from reading it? You have to kind of tell me what you think the paper says and not just toss it against the wall. What do you see in that paper, leo, which is relevant to what either of us have said?

Here's what I see, Any audiophile will agree that the most valuable thing they have with respect to their audio system is their own hearing. In fact human hearing defines the reality of audio. As these words are written, the high-end audio industry has been experiencing a shrinking market for over ten years. It is no surprise- in order for the market to expand, the industry has to touch, move and inspire the marketplace with the possibility of real music. 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.



That is what I have been saying regarding the "progress" achieved in high end audio in the last fifty years. It has largely not been objective measurements which have accomplished that progress. Objectively, large amounts of NFB actually work to the designer's advantage. Subjectively, they are the exact opposite of what "good audio design" should produce. Argue that, leo, not something I have never said.




As to your statements, I see this quote as being relevant to the discussion, "The problem here is human nature. We tend to stay within the limits thus set by the existing paradigms and to resist changes that threaten one's viewpoint of the world. When someone else creates challenges to the paradigms, it is normal also to try to protect one's world view by preventing the new idea from gaining ground. "



If my opinion is banging up against your world view, leo, I can see why you are putting up such a fuss over this. I cannot, however, see why you cannot argue exactly what I have stated on numerous occasions. But I can see why you continue to set up strawmen such as you have.


Going forward, leo, please debate what I have said, not what you think you have read. And, if you wish to add a reference to an article such as the Atmas-Phere owner's, please, tell me what I'm supposed to take away from it. Making me guess at what you see as being relevant to our discussion is unlikely to actually advance this debate.


OK?


Thanks.




.
 

Gold Member
Username: Magfan

USA

Post Number: 2936
Registered: Oct-07
http://www.gurufocus.com/StockLink.php?type=sec&symbol=IRF&date=2012-08-22&repor t=10-K

Please see pages 30 / 31 for some fiancial details, including the fact that IR, a primarily power device company, spent 12.9% of OVER a billion dollars gross revenue on R+D. This number has exceeded 10% for at least the last 3 years.
I guess they figure the age of discrete is over.

And what any of this has to do with anyone's political peccadillos is up for grabs. I think the total cost of this election cycle (presidential only) is nearly 2 Billion $. ouch.
 

Platinum Member
Username: Jan_b_vigne

Dallas, TX

Post Number: 17499
Registered: May-04
.

That's wonderful, leo.

But analog, discrete solid state would appear to be a niche market even for INTERNATIONAL RECTIFIER CORPORATION.

From p. 30; During fiscal year 2012, in order to provide us additional flexibility to adjust our internal manufacturing footprint to changes in demand, we added additional capacity to our existing contract wafer fabrication and assembly capacity related to certain proprietary and higher value-added products and programs. In the past we have targeted external contractors for up to 30 percent of our wafer fabrication needs, and 60 percent of our packaging needs. Going forward, we plan to further expand our use of external contractors for up to 50 percent of our wafer fabrication needs, and 70 percent of our packaging needs. We will continue to monitor the demand environment and we may seek to further adjust our operational footprint to reflect changes in demand.

IR products ...
http://www.alliedelec.com/international-rectifier/?utm_source=bing&utm_medium=cp c&utm_campaign=bing_Supplier%20Keywords%20-%20New



"International Rectifier Corp. is a manufacturer of power semiconductors (Integrated Circuits, MOSFET, and IGBTs) located in El Segundo, California, USA. It has a market capitalization of $1.6 billion USD and is listed on the S&P Midcap 400.

Currently employing over 4900 employees, it is a global leader in power management technology[citation needed], with products found in a very diverse range of devices, such as calculators and airplanes. With aggressive funding in research and development, their proprietary products are used by many computer manufacturers worldwide.

[edit] History
1954: commercialization of germanium rectifiers
1959: creation of the first silicon-based rectifier
1974: first power and Darlington transistors which used glass passivation
1979: first hexagonal MOSFET
1983: first intelligent power ICs
2000: FlipFETTM wafer packaging
2002: DirectFETR, a MOSFET packaging technology developed to address thermal limitations found in advanced computing, consumer and communications applications
http://en.wikipedia.org/wiki/International_Rectifier




So, how much of that "12.9% of OVER a billion dollars gross revenue" is spent on R&D of discrete solid state?




.
 

Gold Member
Username: Magfan

USA

Post Number: 2937
Registered: Oct-07
Jan Quote:
'But analog, discrete solid state would appear to be a niche market even for INTERNATIONAL RECTIFIER CORPORATION. '

Niche for IR?
Are you smoking something?
When i worked for IR, here in SoCal, the fab(s) I worked in totaled over 2 acred (>90,000 sq ft.) and had a designed capacity to start probably 25,000 6" wafers per WEEK. The reason I don't have exact numbers is I left during the expansion. My 1/2 of the fab could easily start 15,000 wafers per week and they were more than doubling capacity. My guess of capacity is if anything, low. Some wafers would yield thousands of good die.
That was 100% discreet including HEXFETs, IGBTs, FREDs but NO ICs.
Where I worked was considered the 'cash cow' of the company. At one point the company sold all the older tech stuff to Vishay who used OUR fab as a foundry. That is the area which was converted over to 6" and the products we made in the other fab....where I was a plankholder, having been with the startup crew.

One additional note: The page #30 quote you cite? That means in part, that IR is moving production overseas. Fabs which build product for anyone who can write a check are called 'foundries'. They are less expensive. Vishay for example, is heavily into foundry fabs.
Many older tech and commodity devices are sourced thru foundries while the original companies retain more advanced and higher profit devices under their own roof. IR does lots of assembly down in Tijuana, which was moved from the facility in which I worked.

I'm not about to figure out the breakdown. That IR, which is primarily a Power Management focused company spends such a large sum on R+D, I'll take as reasonable evidence that they are interested in the future and Progressive Improvement of the product line.

They ARE investing in GaN devices which are a fairly new technology. In the past, SS was built on or 'of' Silicon or Germanium. Some specialty Gallium Arsenide and other materials were used, and I once saw experiments done with what was called a Silicon-Germanium hetrojunction. You'll have to look the LAST up in a back issue of Scientific American.

Cheers.
 

Platinum Member
Username: Jan_b_vigne

Dallas, TX

Post Number: 17501
Registered: May-04
.

OK, leo, if you're convinced you're right, that's all that matters.
 

Silver Member
Username: Ornello

Post Number: 124
Registered: Dec-12
Generally speaking, I prefer the sound of good solid-state amplifiers to any tube amplifier that I have heard.

The transformers necessary with tubes create distortions that I find unpleasant, and impair bass damping. Also, power is limited (except at stratospheric prices).
 

Gold Member
Username: Magfan

USA

Post Number: 2990
Registered: Oct-07
a properly designed and executed tube amp will have no such problem.

a 'good' output transformer for such an amp will be good to 80Khz or above and have terrific square wave response.

Bass damping would appear to be an interaction of speaker and amp...and to blame poor damping of a speaker on low damping factor would be to ignore such interaction as causal.

Check out the Atmasphere OTL designs......
 

Silver Member
Username: Ornello

Post Number: 127
Registered: Dec-12
Yes, but transformers have weaknesses no matter how 'good' (i.e., expensive) they are. You are not getting the signal directly from the tube circuit.

Tube amplifiers are of limited application (primarily with very sensitive speakers). I never heard one I liked. The midrange is merely OK, the high end is pinched, and the low end muddy. And this is true even of the biggest and best AR stuff that can be used to heat a house.
 

Platinum Member
Username: Jan_b_vigne

Dallas, TX

Post Number: 17694
Registered: May-04
.

Leo has his facts right. The other guy, the troll, doesn't have a clue. But, then, the troll has proven since the day he showed up here that he's not much on real world information.

Ornello posted on Friday, January 25, 2013 - 01:23 GMT

I cannot imagine a phono input on any major company's product without a phono section.

Ornello posted on Wednesday, January 30, 2013 - 22:19 GMT

An amp for a cell phone? I don't they even make such an article.




Proceed at your own risk, leo. The troll has been digging into threads that haven't been active in months looking to stir up some attention for himself. There's no fact you can present here that the troll will accept. That would actually end the attention he wants.

Remember; https://www.ecoustics.com/electronics/forum/home-audio/706244.html

Stupid is as stupid does. And you're jumping right into "stupid" with this troll.



.
 

Silver Member
Username: Ornello

Post Number: 129
Registered: Dec-12
So, what difference does it make, how old a thread is?

I just joined, and some things are of interest to me regardless of their age.
 

Platinum Member
Username: Jan_b_vigne

Dallas, TX

Post Number: 17695
Registered: May-04
.

What matters is you are a supreme troll uninterested in an actual conversation and very uninformed when it comes to audio. Nothing you have posted regarding tube amps is correct and you won't listen to facts. What you think of tube amps is the typical troll BS about tube amps. Therefore, there's no reason to give you more air time to be a troll.


http://search.yahoo.com/search?p=internet+troll&ei=utf-8&fr=veri-ie8





.
 

Silver Member
Username: Ornello

Post Number: 132
Registered: Dec-12
I find you quite a curious case. May go down in medical history.

I do understand the limitations of tube amps, and every one I have ever heard (including the big 300 or 600 wpc Audio Research room heaters) is affected by these limitations.

They have audible distortion, they are slow, and the bass is sloppy or weak or both (transformers fall off at low frequencies). I can hear these things.

http://www.ebay.com/itm/Audio-Research-Reference-600-MK-II-/261159028819?pt=US_Home_Audio_Amplifiers_Preamps&hash=item3cce4a8853

If they are so good, why is the owner selling them?

Hmmmmmm?

Look here, bubba:

http://books.google.com/books?id=sh94bLWOTY4C&pg=PA424&lpg=PA424&dq=transformer+frequency+response+amplifier&source=bl&ots=e9NlyJvZcI&sig=1q1fLI-Ir9VPKk-Q4lfXihnJii4&hl=en&sa=X&ei=RbAKUaK0E-r-2QWal4DIDw&ved=0CDUQ6AEwAQ

and here:

http://www.aikenamps.com/OutputTransformers.html

The problem is not entirely the tubes themselves. A large part of the problem with tube amps is the transformers.
 

Gold Member
Username: Hawkbilly

Nova Scotia Canada

Post Number: 1548
Registered: Jul-07
"I just joined, and some things are of interest to me regardless of their age."

But then.....

"They have audible distortion, they are slow, and the bass is sloppy or weak or both "



Oh yeah, you're REAL interested in a tube amp thread.

"Who cares about manners? I do, and we all should."

 

Silver Member
Username: Ornello

Post Number: 144
Registered: Dec-12
and a pleasant day to you, Sir!
 

Gold Member
Username: Magfan

USA

Post Number: 2992
Registered: Oct-07
Orn, a PROPERLY integrated tube system will have good bass. My panels, though a low sensitivity load, respond well to tubes since they don't have any wacky phase problems.

An avenue of investigation for you would be tubes and phased loads....And do tube or do they not like capacitive or inductive loads? Will some speakers of Q>=1.0 sound good with tubes or SS....at least in the bass?

To make a blanket statement like.....

"They have audible distortion, they are slow, and the bass is sloppy or weak or both "

is almost automatically wrong. The logican in you should already know that.....

And personally? I've heard some great bass out of tube systems. With the proper speaker of reasonable inpedance 'curve' and low reactivity.

As a bonus, I'm throwing in an article by Nelson Pass, who I'm sure you've taught much...but he must have forgotten somewhere along the way.

http://www.passdiy.com/pdf/cs-amps-speakers.pdf
 

Silver Member
Username: Ornello

Post Number: 145
Registered: Dec-12
I have never ever heard a tube amplifier (including the big AR stuff) that I liked as well as good transistor stuff, such as Mark Levinson.

Tube amplifiers almost have to use transformers, and those devices are the weakest links.
 

Silver Member
Username: Ornello

Post Number: 147
Registered: Dec-12
"Orn, a PROPERLY integrated tube system will have good bass."

Trouble is, there are no such PROPERLY integrated tube systems.

I have always wondered why certain people gushed over tubes, so I spent some time listening to some fairly pricey tube amplifiers hooked up to some worthy speakers. Not impressed.
« Previous Thread Next Thread »



Main Forums

Today's Posts

Forum Help

Follow Us