valves better than solid state?

[x9moto]

i thought i would ask this question as its usually a hot topic, and i have read that transisters are suposed to be superior yet there is a destinct difference in the sound to me, i prefer the valve, but could it be that the tube amp i built is simply better than my tranny amp and that a better tranny amp would be superior?

 

[503Timber]

The "superior" product is your ears.....
Each design has it's place, strengths and weaknesses that are quite different. Other things come into play as well....like your speakers, they may like SS better than valves.

As far as going up the chain in one design to pass the quality of your other amps quality is a never ending cycle

 

[simon7000]

It is easier to build a good amplifier with valves rather than solid state. See http://www.diyaudio.com/forums/anal...ch-preamplifier-part-ii-1310.html#post2622356

 

[dsavitsk]

Tubes are generally more linear as amplifying devices, but in reality, the difference most people hear between tubes and sand has more to do with different topologies being implemented than with huge differences in the parts. Build a tube topology with solid state parts, and it will sound surprisingly like a tube amp. And, build a solid state topology with tubes, and it will sound like a sand amp.

JC saying one thing is better than another hardly seems like authoritative evidence ...

 

[Globulator]

i thought i would ask this question as its usually a hot topic, and i have read that transisters are suposed to be superior yet there is a destinct difference in the sound to me, i prefer the valve, but could it be that the tube amp i built is simply better than my tranny amp and that a better tranny amp would be superior?

Transistors are far superior in terms of mechanical ruggedness, but as far as linearity for audio work they are pretty poor. A transistor covers FETs, MOSFETs and BJTs, plus the odd IGBT. Most people think of them as BJTs I suspect.

Additionally the problem with transistors is that they can work with low voltages, they are small and cheap. All very bad things - causing amplifier designers to use too many of them and at low voltages.

The voltage isn't such a big deal perhaps - but headroom and linearity suffer when you traverse more of the transfer curve. The key with a tube amp is using very few rather linear devices in a smaller part of their range. A second effect of this is that the amount of GNFB is less, which in my view also improves the sound.

 

[kstagger]

I used to be firmly in the tube camp - but to my ears (and recent speaker preferences), I've been enjoying my first all-SS system. Threshold gear had different strengths and weaknesses over my DIY tube gear, but for most of the music I like, it really does the job with my Magnepans. Heck, I'm using (the horror!) op-amps in my current MC phono preamp - Audio Sector's Phono Stage - and it sounds very fast and detailed.

The best part - especially from a DIY part - is the electrical toughness of tubes. It's a great way to learn (and make mistakes!) without destroying your creation. It's also fairly hard to make a truly wretched sounding tube circuit - I've (accidentally) gone far off any proper parameters but still ended up with something that "sounded good" but would have measured terribly.

 

[DF96]

Which is the better car: Bentley or Aston Martin? No, I'm not starting a car debate and you can substitute any pair of quite different cars from your own country. It is essentially a meaningless question, as good and bad amps can be made with either technology.

Beware of people who like "tube sound" - a good valve amp should not have much sound at all, you just hear the music. Beware of people who think that no/lots of feedback is necessarily a good thing. Beware even more of people who think that a 'little' feedback is a good thing. Beware of people who think that digital/analogue/valve/SS is the only sound worth listening to. There, I've managed to annoy almost everyone apart from a few sensible people like me and those who agree with me! Seriously, real questions don't have such simple answers.

 

[x9moto]

im not really biased towards either its just that currently the tube amp i own has a nicer sound to me than any of my other amps, seems more detailed and the high's seem high rather than muffled and unclear.

i notice bits of the music in the midrange that seem to be either enhanced on the tube or it is more linear and the tranny amp has a flat spot but most of my tranny amps seem to sound like that, ofcourse it could be that my amp is exaggerating that part but it seems to sound right, clearer.

may be worth mentioning that i only have the one set of speakers so all sets were tried on the same speakers, which were skitech 8ohm, 12inch woofer and horn loaded tweeter.

 

[Miles Prower]

There has never been a technology without a downside, and there never will be.

i thought i would ask this question as its usually a hot topic, and i have read that transisters are suposed to be superior yet there is a destinct difference in the sound to me, i prefer the valve, but could it be that the tube amp i built is simply better than my tranny amp and that a better tranny amp would be superior?

Yep, it's opening a big can of worms here. Transistors are superior in a good many respects: they make really good switches, much better than VTs in that regard. Since transistors are high gain devices, it's less critical to choose a type for a design unless you need some unusual property if the type in question can process the signal frequencies. They're not so swell when it comes to amplification, as the linearity is horrible. A lot of SS acolytes make a BFD over the need for OPTs with hollow state, but linear device + somewhat nonlinear OPT= more linear than very nonlinear transistor.

It is simply easier to get good sound from a hollow state design, since you're starting off with a far more linear device with a far more graceful clipping behavior on overdrive. That doesn't mean it can't be done with solid state (you just have to work at it more) and there is no reason why solid state based implementations need sound as hideous as they do. It is unfortunate, but they're not teaching the fundamentals of sonic performance in electronic engineer school. Instead, the marketing dept gets to work to convince you that that "MOSFET sound" is uber-kewel, as opposed to what it is: nasty harmonic distortion. It is also unfortunate that a good many CDs are mastered to cover up for the sonic defects of the SS amps they'll likely be played through. These sound "off" when heard through a hollow state design.

It is also unfortunate that solid state is moving towards digital and switching applications. Good lateral MOSFETs are already a niche item. IGBTs are useless for audio, except as something different and quirky for the sake of being different and quirky.

I've done it myself: designed a really good sounding solid state amp that comes very close to the sonic quality of a hollow state amp. It doesn't quite get there, but it's close. Of course, I disregarded a lot of nonsense and folk wisdom you see concerning solid state design.

 

[M Gregg]

Could it be that hybrid is better than either!

LOL.. only joking

This is impossible to answer....Build one of each and swap with your mood!

You haven't asked if this was PP, SE, or OTL either or does chip sound better than discrete components.

One other thing is (DIY aside) : how much would a tube amp cost to buy and how does it compare with a SS amp of similar value!

Regards
M. Gregg

 

[Salas]

Miles's post I find very learned and useful. I have done both and then I mix some, and is always down to what fits better to a plan, or its about what excursion you wanna head to today, its a hobby, its nice to experiment. We get to know what works for us after we build some.

 

[CBS240]

Transistors are far superior in terms of mechanical ruggedness, but as far as linearity for audio work they are pretty poor. A transistor covers FETs, MOSFETs and BJTs, plus the odd IGBT. Most people think of them as BJTs I suspect.

Additionally the problem with transistors is that they can work with low voltages, they are small and cheap. All very bad things - causing amplifier designers to use too many of them and at low voltages.

The voltage isn't such a big deal perhaps - but headroom and linearity suffer when you traverse more of the transfer curve. The key with a tube amp is using very few rather linear devices in a smaller part of their range. A second effect of this is that the amount of GNFB is less, which in my view also improves the sound.

Tubes have a much larger 'linear' operating region compared to any transistor. Also the distortion spectrum of valves is typically even harmonic compared to the odd harmonic distortion of most transistors, perhaps excluding Lateral mosfets. Lat fets typically have a 'Square Law' like Gm vs Id curve, this makes the Lat fet as close to 'tube like' as you will find in solid state pwr. J-fets are nice too.
Because the transistor distortion spectrum contains unpleasant odd harmonics, good solid state amplifiers have very low overall distortion. Many speakers today are less sensitive requiring certain amounts of damping with large currents and do not do well to be driven by a transformer unless the iron is massive.

Transistors are cheap, and because of this and the fact that they are poor linear amplifiers, many transistors can be and are used. IMO, it does not matter how many devices or stages there are, it is all in how they are used. Using multiple devices to mitigate the shortcomings of a single transistor is common. For example, how do you solve the issues of a VAS transistor that has inadequate Vceo? What issues are there? Of course you can use a larger Vceo device but the lower Vceo device may have superior current gain and Ft but also a much lower Early Voltage. Simple solution is to cascode with a higher Vceo device and drive it common base from the other. This takes more transistors and I can assure you in SS that reducing the number of components just for that sake will cut into possible quality like a knife. The use of error correction schemes for the output transistors is certainly valid and works phenomenally well but it may require several extra components. Since valves have a large linear gain region and are not plagued by the significantly bad sounding distortions of sand, the number of active components can be few.



BTW I am contemplating a balanced solid state amp that uses a direct coupled duel triode, something like an E88cc, as the VAS. After all this is just a hobby and I think it would be a neat way to make a hybrid.

It is unfortunate, but they're not teaching the fundamentals of sonic performance in electronic engineer school. Instead, the marketing dept gets to work to convince you that that "MOSFET sound" is uber-kewel, as opposed to what it is: nasty harmonic distortion. It is also unfortunate that a good many CDs are mastered to cover up for the sonic defects of the SS amps they'll likely be played through.

Fundamentals of sonic performance is really fundamentals of linear analog electronics design. This is a subject that is neglected much more now that digital electronics has taken over nearly everything. I think the old school instructors who are very skilled in analog design are retiring and the newer instructors just don't have the experiences of their prior peers. There is a way to take out that "mosfet sound" and create very acurate linear operation from them. But it takes more sand and more parts.

 

[tomtt]

JC saying one thing is better than another hardly seems like authoritative evidence ...

i'd say it does.

note his ' track record '


i prefer tubes, for most situations ...

 

[tinitus]

my best experience with a tube power amp have been that it was able to recreate a harmonic sound and substance, on the inadequate speakers of the past

there are also some cases where a tube amp designer have done his utmost to make it perfect
and the resulting sound more like a poorly designed SS amp

with todays better speakers you can achieve the same with SS amps
and there seem to be a much better understanding of the speaker - amp related interaction

I have always like a good tube pre
but I remember a very embarrasing 'test'
I had to choose an IC based preamp over a simple tube pre
and I say had to, because it was clear that some high frequency details simply dissappeared on the tube pre
well, unfortunately I forgot that the rest sounded much nicer on the tube pre

but I also remember another simple tube riia/line pre
probably the best I ever owned

but I guess no point in denying it, very often some of the top and bottom will be 'missing' on many tube amps
that is, if there is anything to be 'missing'

 

[x9moto]

another interesting thing i have noticed is the difference in sound by changing the configuration.

my amp is a pp class A with 4 valves and removing the parallel ones seems to result in superb subsonic bass response but lacks power ever so slightly, the sound is also cleaner than when running on all 4.

removing the tone amp valve also has a positive effect on the sound at the expense of some bass.

 

[Salas]

Also the distortion spectrum of valves is typically even harmonic compared to the odd harmonic distortion of most transistors, perhaps excluding Lateral mosfets. Lat fets typically have a 'Square Law' like Gm vs Id curve, this makes the Lat fet as close to 'tube like' as you will find in solid state pwr. J-fets are nice too.

Power Jfets like the exotic recent Semisouth and old Nec and Sony specials are also worth mentioning for nearing 45deg triode like transfer curves in the SS power devices all time wall of fame.

 

[DF96]

Also the distortion spectrum of valves is typically even harmonic compared to the odd harmonic distortion of most transistors

People often say this, but it is not really true. Both the ideal valve (3/2 law) and the ideal BJT (exponential law) have odd and even terms in an infinite series, with second-order dominating for most typical signal values.

The useful thing about valves is that a small-signal valve roughly corresponds to a medium-power transistor in terms of signal handling. On the other hand, for really low noise you need to get away from a hot cathode. Horses for courses.

 

[Globulator]

BTW I am contemplating a balanced solid state amp that uses a direct coupled duel triode, something like an E88cc, as the VAS. After all this is just a hobby and I think it would be a neat way to make a hybrid.

The closest I get to a hybrid for DIY is solid state in the PSU
I used to do a lot of transistor DIY years ago, but these days I prefer the ease of using tubes in designs!

 

[Globulator]

for really low noise you need to get away from a hot cathode. Horses for courses.

Surely the D3a gives us the low noise tube?!

 

[M Gregg]

The closest I get to a hybrid for DIY is solid state in the PSU

No CCS?

Regards
M. Gregg