Stereophile's Jim Austin disagrees w Atkinson; says tubes have something that can't be measured

[solderdude]

It's not so much the active devices that are used.
It is more about HOW they are used and longevity.
One can design amps using tubes that sound and perform the same as a good, equally powerful, SS amp but... that's not what tube amp buyers are looking for.
They want something 'different' and that is easy to do with tubes AND they generally look very, very different and nostalgic with mysterious looking and glowing tubes.
Other may desire swinging needles or impressive looking boxes or lots of buttons/knobs or displays to look at.
It is often more about the experience rather than technical performance for the majority of audio-philes/audio-phools/audio-afficionados/music lovers/gear-heads/braggers & show-offs.

Both JA's cater for both camps. Stereophile may need that to continue the path of making a living (or add to their income) and be able to continue to be sent unaffordable gear to play with either for free or on loan.
When one is a reader (I am not) one can choose to 'believe' what they like to believe. The ironic part is MOST people really want the magic and you then need a writer/reviewer that gives them exactly that.
No one HAS to read AND believe anything they say/write but should be free to choose so.... Magic vs science... best to cater for all.

 

[theREALdotnet]

Go ahead and prove me wrong on any of it.

You don't think there is a reason why tubes were obsoleted in the 70s? You don't think basic transistors do anything a tube can do - much better? High quality power transistors are beautiful in behavior. Tubes have never been, nor will they ever be. Enjoy their distortion all you want.

You’re clearly going off on a wrong tangent here, and I’m not going to stop you. For the record (no, not the vinyl one), I don’t believe audio equipment these days has any use for tubes, for the reasons I stated. But claiming that reliable, linear tube devices cannot be designed and built, or that tubes are simply rubbish in general is just nonsense and shows complete ignorance of electronics history.

 

[Galliardist]

I haven’t had much success playing CDs with mine, though

Is this finally the explanation for the vinyl revival, and can we finally see off THAT thread?

 

[computer-audiophile]

View attachment 285698
© Troels Gravesen

It depends what you mean by “very good”.

I think I have to explain. I know the measurements and experiments of Troels Gravensen from his website, but I don't agree with everything he does, his approach and what he concludes. I don't care either. It is not my point of view.

At first, it is not easy to find perfectly preserved specimens of this old speaker today. But purely aurally the Philips 9710M Alnico is one of the best vintage full range speakers. In my opinion, they develop a special sonic charm e.g. when it comes to voices and small classic ensembles. You have to hear it live to make a judgment about it. To stay in the topic of the thread, this 'charm' is something that can't be measured. By the way, these speakers fit perfectly with tube amps, as they were designed for them.

I have spent decades in this field of full range vintage drivers, built many special cabinets for them and had a lot of listening pleasure. I was not alone with this, have received enthusiastic comments from other audiophiles at scene meetings and my design has also been featured in the German audiophile magazine 'Hörerlebnis'.

I consider these speakers to be in a class of their own. They do not compete with modern speakers like e.g. with my Neumann KH120a, with which I have been listening in the near field for the last 10 years.

Tube amplifiers are also in a class of their own, and it's okay for me if you don't want to deal with them.

 

[computer-audiophile]

Addendum to #433 Photo for illustration. Especially the vintage SABA Greencone loudspeakers have been in the focus of my attention. I show a picture of the cabinets I have built for these.

 

[Sal1950]

I show a picture of the cabinets I have built for these.

Simply beautiful woodwork, I envy your craftsmanship.
One word of suggestion, I think you forgot to put in 2 or 3 more
drivers in that large cabinet.

 

[atmasphere]

Go ahead and prove me wrong on any of it.

You don't think there is a reason why tubes were obsoleted in the 70s? You don't think basic transistors do anything a tube can do - much better? High quality power transistors are beautiful in behavior. Tubes have never been, nor will they ever be. Enjoy their distortion all you want.

They were declared obsolete in the 1960s, FWIW. When germanium transistors were king.

Also FWIW, triodes are far more linear than transistors- they can put transistors to shame in that regard. If you build the tube amp properly and design your feedback loop(s) properly, they can perform extremely well. We've gotten zero feedback tube amplifiers to have IMD under 0.05% at full power. The circuit can do considerably better than that with feedback (BTW our tube amps are OTLs so one of the phase margin issues commonly associated with tubes is eliminated). Full power bandwidth from 1Hz to 200KHz. Fully differential from input to output (our MA-1 was the first amplifier offered to the home stereo market with a balanced input supporting AES48). Output section risetime of 600V/usec. The output section has bandwidth such that its still linear at 30MHz.

I'm not advocating tubes here. I'm just saying (apparently, again) that if you are going to denigrate tubes (or anything else for that matter), its helpful to do with with facts rather than myth.

 

[Cbdb2]

The same can be had from transistors at a fraction of the cost and the transistors never have to be replaced. Theres a reason tubes almost disappeared.

 

[computer-audiophile]

The same can be had from transistors at a fraction of the cost and the transistors never have to be replaced. Theres a reason tubes almost disappeared.

Kind of the same - but different. Because there is something with tubes...

 

[Sal1950]

Kind of the same - but different. Because there is something with tubes...

Yep, they love to blow up.
Wish I had a dollar for every tube I've replaced since around 1958

 

[atmasphere]

The same can be had from transistors at a fraction of the cost and the transistors never have to be replaced. Theres a reason tubes almost disappeared.

This really isn't true. First, tubes are still with us. What's going to kill them is class D guitar amplifiers invading the guitar market- nothing to do with hifi. Yes, sometimes they are cheaper. But the output devices we use in our class D amp are more expensive than the power tubes we use....

Second, if you think transistors 'never have to be replaced' you're living in a fantasy world- gnomes jumping over mushrooms, that sort of thing.

Transistors and chips die due to corrosion (amongst other things). So they can have a pretty long life, but they do on occasion have to be replaced. Certain transistors are notorious for failure- there's a list of them over on audiokarma.
Those that spend any time working on vintage effects pedals, synthesizers and audio equipment know what I'm talking about. The scary bit about semiconductors is the list of obsolete devices is pretty daunting- the semiconductor industry is always making strides so older devices go n/l/a. That means it can get tricky to repair older solid state equipment. For the time being at least, tubes for older and newer equipment are pretty easy to find, except for the newer KT150, which seems to have caused Audio Research some problems lately since their stuff is designed for that tube.

 

[ahofer]

This really isn't true. First, tubes are still with us. What's going to kill them is class D guitar amplifiers invading the guitar market- nothing to do with hifi. Yes, sometimes they are cheaper. But the output devices we use in our class D amp are more expensive than the power tubes we use....

Second, if you think transistors 'never have to be replaced' you're living in a fantasy world- gnomes jumping over mushrooms, that sort of thing.

Transistors and chips die due to corrosion (amongst other things). So they can have a pretty long life, but they do on occasion have to be replaced. Certain transistors are notorious for failure- there's a list of them over on audiokarma.
Those that spend any time working on vintage effects pedals, synthesizers and audio equipment know what I'm talking about. The scary bit about semiconductors is the list of obsolete devices is pretty daunting- the semiconductor industry is always making strides so older devices go n/l/a. That means it can get tricky to repair older solid state equipment. For the time being at least, tubes for older and newer equipment are pretty easy to find, except for the newer KT150, which seems to have caused Audio Research some problems lately since their stuff is designed for that tube.

Returning to the thread topic, I recall you had a theory that the tube difference had something to do with transient distortion on dynamic peaks. This, at least, is measurable. Did you ever get together with @pkane to see if it could be simulated and tested?

 

[atmasphere]

Returning to the thread topic, I recall you had a theory that the tube difference had something to do with transient distortion on dynamic peaks. This, at least, is measurable. Did you ever get together with @pkane to see if it could be simulated and tested?

No, that was more about distortion rising with frequency. He apparently had written the code for that but never implemented it. The thing is, most amplifiers made that employ feedback have rising distortion with frequency so its not anything obscure, and THD tends to hide the actual distortion by sweeping it under the carpet (depending on what frequency the amplifier is tested).

 

[pablolie]

This really isn't true. First, tubes are still with us. What's going to kill them is class D guitar amplifiers invading the guitar market- nothing to do with hifi. Yes, sometimes they are cheaper. But the output devices we use in our class D amp are more expensive than the power tubes we use....

Second, if you think transistors 'never have to be replaced' you're living in a fantasy world- gnomes jumping over mushrooms, that sort of thing.

Transistors and chips die due to corrosion (amongst other things). So they can have a pretty long life, but they do on occasion have to be replaced. Certain transistors are notorious for failure- there's a list of them over on audiokarma.
Those that spend any time working on vintage effects pedals, synthesizers and audio equipment know what I'm talking about. The scary bit about semiconductors is the list of obsolete devices is pretty daunting- the semiconductor industry is always making strides so older devices go n/l/a. That means it can get tricky to repair older solid state equipment. For the time being at least, tubes for older and newer equipment are pretty easy to find, except for the newer KT150, which seems to have caused Audio Research some problems lately since their stuff is designed for that tube.

You start to sound like you have an agenda. :-D "Tubes rule, semiconductors suck"... has no one ever ever said since Shockley and his team did some work...

Incidentally you seem to confuse circuit design with the merits of discrete components' specs. Putzey's white papers may provide some help there. Try to deliver on those specs with your "superior" tube tech, from linearity to power to distortion.

 

[theREALdotnet]

You start to sound like you have an agenda. :-D "Tubes rule, semiconductors suck"... has no one ever ever said

Including in this thread. So, why do you keep tilting at this windmill?

 

[SIY]

Also FWIW, triodes are far more linear than transistors- they can put transistors to shame in that regard.

Depends on the topology. Devices alone don't have linearity, circuits do. In common cathode/source/emitter, a triode will be more linear BUT have lower gain so that once feedback is applied, it will lose the linearity game. In common plate/drain/collector, the gains will be similar, but the transistor will be much more linear.

 

[pablolie]

Including in this thread. So, why do you keep tilting at this windmill?

because i have reading comprehension? :-D

 

[pablolie]

Depends on the topology. Devices alone don't have linearity, circuits do. In common cathode/source/emitter, a triode will be more linear BUT have lower gain so that once feedback is applied, it will lose the linearity game. In common plate/drain/collector, the gains will be similar, but the transistor will be much more linear.

with transistor circuits you can implement something called feedback more effectively.

 

[SIY]

with transistor circuits you can implement something called feedback more effectively.

Well, typically there's higher open loop gain. But still, it's not difficult to engineer tube circuits to give vastly lower levels of distortion than humans can detect by ear.

 

[amirm]

You start to sound like you have an agenda. :-D "Tubes rule, semiconductors suck"... has no one ever ever said since Shockley and his team did some work...

Incidentally you seem to confuse circuit design with the merits of discrete components' specs. Putzey's white papers may provide some help there. Try to deliver on those specs with your "superior" tube tech, from linearity to power to distortion.

Ralph is a very experienced designer (tube and now transistor). Let's disagree more professionally.