[B said:
The main reason why in real life the tube amps sound much more better than solid state amps (at least in most cases), is primarily therefore because most solid state amplifiers runs not far enough in Class-A mode (in contrast to tubes). By both kinds of amps there are 20mA until 50mA quiescent current through the output stage. In case of solid state output voltage supply goes from 50V - 120V but in case of tube output stage from 250V to 600V (except OTL - this is through the current control output again not directly comparable).
Secondly, the so-called "true complementary" solid state amplifiers also not comparable with the tube amp in class-A operation (even not in pure class-A), because there are no P-CH tubes. Already through this fact tube amps for audio must be better in general cause the better matching of the two halves.
This means clearly that only "Single Ended" and CSPP (PPP, Circlotron) solid state power stages in Class-A mode (by "Single Ended" unavoidable) are to compare with tube power stages, for avoiding to compare apples to pears or oranges.
In light of these conditions, solid state amplifiers clearly outperform each tube amp in every respect. Otherwise there are heavy design error(s) in the solid state amplifier.
Unfortunately there are not a wide range of such commercial solid state amps on the marked; commercial and diy (kit) examples I have listed there:
http://www.diyaudio.com/forums/soli...ended-integrated-power-amplifier-devices.html
and there
http://www.diyaudio.com/forums/soli...e-ended-related-solid-state-output-stage.html
But nobody could show me until this day a commercial tube amplifier, which clear outperform the biggest single ended solid state amp series from Nelson Pass, the "Aleph" - e. g. the "Aleph 1.2"
To outperform a normally solid state Amplifier in Class-AB mode with low transition aera from "A" to "B" (independend of the price class) is easy to show with most average tube amp, independend whether there are an ultralinear-, Single ended-, or PPP topology.
Sadly but true is that the largest percentage of solid state audio amplifiers uses small amounts of quiescent currents and so called "True complementary" output devices in the output power stages. I love this by theoretical explantations for beginners and on the schematics, but I hate this in real life .
This is for me even the only reason why so many music lovers (rightly) say's, solid state amplifier did sound unmusical (even in traditionally measured THD values that are extrem low and therefore hardly detectable). They have never heard unfortunately the other solid state amps like Aleph.
Last edited:
The best thing in that link is "Attention: This product is no longer available for sale".
A good news by time in time... 😉
Piercarlo
?
Unless I'm really misunderstanding you, which is certainly possible, you seem to be saying that one reason that tubes are better is because they use the same polarity of output devices in both halves of the push-pull configuration. But there are a lot of solid state amps that have done this as well, like the Phase Linear 300, or the SUMO The Nine.
If you find PNP/P-Channel devices undesirable, you can certainly build a solid state (any class) amplifier without them.
??? I thought his post was pretty factual. You (or I) may not agree, but that's not the point is it.
jd
So let's compare about output devices. I would expect the fastest transistor to be faster than the fastest tube if for no other reason than physical size and inductance. That's fine but we're talking about output devices in audio amplifiers. What happens if we compare devices by normalizing them by power output?
I'm no tube expert but I believe the KT88 is used a lot in audio output stages and I think it is a 40W device. How fast is the KT88 compared with n-channel 40W BJT, MOSFET, JFET (is there a 40W JFET?), IGBT?
I believe even "fast" audio power output bjt are device physics limited
metal gate Mosfet "speed" however is primarily package parasitic limited in popular TO-220/247 leaded packaging
IXYS RF: HF/VHF Linear MOSFETs
metal gate Mosfet "speed" however is primarily package parasitic limited in popular TO-220/247 leaded packaging
IXYS RF: HF/VHF Linear MOSFETs
You don't misunderstanding this but I think you misunderstand the whole.
What means "a lot of" for you ??
SUMO "The NINE" could be one of few exceptions, because it is an CSPP topology (see attachement), but I don't know about the idle current values. If it runs typically arround 50mA, nevertheless you cannot compare with tube amp.
The Phase Linear 300 use low idle current values and therefore, it is not comparable anyway to any tube amp.
To make long explanations short - the conclusion of my statement from previous is follow:
By compare SS amps with tube amps, it is important to know, to allow only "Single Ended" and CSPP (PPP, Circlotron) solid state power stages in pure Class-A mode.
This means first, that solid state Class AB modes are not to allow for compare (independend of topology), because I have never heard about tube amp devices for home hifi in Class AB (except any few OTL's)
Secondly true complementary (also quasi complementary) SS in pure classA are also not to allow for the aim of comparsion to tube amp, because there are no tubes with reverse polarity.
By breaking this rule the audible differences causes first through basically different operating modes and circuit topologies, but many musiclovers still believe to the superior properties of the tubes.
Bonsai, all what I say are facts, not opinions - so I think.
But if you want to convince me otherwise - please, be happy;
I look forward to this.
What means ":d :d :d" - please let me know
I'm no expert for the typical abbreviations in the "Forums" language - sorry
Member
Joined 2009
Paid Member
For very high frequencies at high power, tubes still reign supreme. Examples are microwave ovens (now there's high frequency for you) and commercial radio station transmitters. The issue is power. At low powers, SS is ubeatable, there are devices that emit signals in the THz region.
---
WHAT have you guys done to me - I'm now stuck trying to understand how to make a low distortion non-global-nfb amplifier with SS parts, with a 'nice' distortion profile (at least in my simulations). It's really really really very darn difficult 😡
Member
Joined 2009
Paid Member
They do have some extra stuff going on with magnetic fields, but there's a heater, anodes and cathodes with a vacuum and lots of electrons. Yes, I agree, they are quite different though. But it's an example where power and high frequency doesn't lend it'self so well to SS.
It would seem to me that once you have to go through an output transformer, the speed of the device (in this case the tube) becomes largely irrelevant.
Power MOSFETs are on average about ten times faster than RET BJTs in terms of equivalant ft, but they have lower transconductance.
A high-quality SS amp with a good soft clip circuit and a couple tenths of an ohm in series with the output can sound a lot like a tube amp.
Cheers,
Bob
I think about this that is a bit extremized though. In my view working classes, idling currents and even switching times, transition frequencies, beta linearity and so on (included the main dispute between vacuum or solid state devices) are, if catched isolately each from other, enough marginal items: that is building the real framework is, imho, not single issues but their crossed interaction.
Or, in other words: looking for single "exhaustive" or "definitive" answers that are in the same time even simple answers is simply wasting the time. In my opinion don't exist any of such "simple answer". I've heard good sound from any kind of anpliifers and, ever from any kind of amplifiers, I've heard equally bad sound. Luckly most of the former than the latter but, for me at least, is not this the crucial detail. That really matter in finding an answer is: "exist some kind of 'magical combinaction' of factors which can assure good sound if attended and bad sound if unattended? And, if exist, can we make it easily understandable and reproducible?"
Hi
Piercarlo.
Play along, why not? How do the speeds compare of tube, FET and BJT all of the same max power output?
And let me raise this hypothetical question. If only N-channel devices existed in all types and you have to use an output transformer, which device type would you use for best sound quality?
Agreed, Bob
And may be an amp capable to handle heavy loads at good power levels without clipping can sound very well, like Krells and MLs.
This is an important point, IMHO, how, when it happens, the amp clips.
Regards,