|
Audio Asylum Thread Printer Get a view of an entire thread on one page |
For Sale Ads |
71.96.141.150
I've heard the one watt in a tube amplifier is roughly equivalent to two watts in a SS amp. Mathematically, a watt is a watt? Can somebody shed some light?
Follow Ups:
I know I'm right but don't want to explain the above so as to play with your mind, if not your diet.Tubes clip differently. Different speakers make different demands on amps in terms of the load they present - even two equally spec'd solid state amps might behave differently into, say, a reactive load.
Many people are saying that the new things like the Teac 700 and the clari T behave as though they have more watts than measured (or at least advertised.)
A watt is not a watt.
- This signature is two channel only -
Here is an article from Stereophile where they tested tube amps to figure out why they sound different and louder. (see the link)In the end, it turns out that due to amplifier-loudspeaker interaction, the tube amps DO put out more power on transient peaks.
They measure a 11 watt 300B amp put out 39Vp into a 8 ohm speaker. Thats the same as a 80watt amp would put out. Read the article and see.
Its as if tube amps, and in particular single-ended amps are optimized for high output into transients.
It was apples versus oranges far more than it was tubes versus transistors.As a convenient example, tell me about the output transformers used on the transistor amp. :)
se
doesn't matter what anyone says or why, a 30-watt tube amp will fill a club up with the same sound level as a 60-watt solid state amp. One of the most popular amps among professionals right now is a 17-watter that totally kills. In home stereo, it's not quite as pronounced, but still a large difference (especially since my solid state amp is always unplugged!)
What brand is the 17 watt Guitar amp?Thanks,JD
Goodsell Super 17. I heard they're made by a guy who used to repair Hammond B3's, and he makes the amps with old Hammond opt's. Price has doubled on them v. quickly due to demand. Killer.
.
Do they have a tube version of that thing yet? I bet it'd sound better.
The whole idea that a tube watt is more powerful than a SS watt is absolute crap. A watt is a watt no matter how you slice it. The reason the whole idea that Tube watts are more powerful comes from the fact that when a tube amplifier clips its not harsh on the ears or tweeters like a SS amp. I actually think that a good SS amp will act more powerful than a tube amp since most SS amps have more current capability and usually can drive difficult loads better than tubes. This is a very general statement and there are exceptions of course, but as for a tube watt being more than a SS watt its BS, it may sound slightly more pwerful because it wont clip like a SS amp when overdriven it will clip very gently. Am I right here anyone agree or disagree?
There actually is a reason why tube amps generally sound more powerful than most equivalently rated SS amps. Most tube amps have a push-pull output stage. With this configuration, imperfections in the power supply (hum, noise, etc.) are rejected by the output transformer. So at high output levels, the "sag" that occurs from loading the power supply is ignored by the audio circuit and the amp sounds more powerful.This does not apply to single-ended tube amps (which are usually only a few watts anyway). However, the same thing happens with SS amps that use a balanced-bridge output stage. These amps tend to sound more powerful than their ratings would suggest.
Most power amps are biased in Class AB. This means they allow the active device that carries the current for the opposite polarity to cut off during large signal excursions.A push-pull tube amp with conventional transformer output coupling rejects power supply hum and noise as long as both tubes are conducting. The configuration of the transformer causes power supply hum and noise that modulates one tube to cancel the (same) hum and noise that modulates the other tube. The transformer achieves a balancing act between the two tubes.
However, when a large signal excursion causes one tube to be cut off, it cannot affect the distortion introduced by power supply sag.
Biasing such a stage into Class A would prevent either tube from ever cutting off, but this would require enormous standby power and heat generation.
...you are overstating the limitations of a tube amp running in class AB. Let's look at this in more detail.The transformer operates by magnetic flux, which in turn is determined by current. When one of the tubes is cut off, the current in that half-winding cannot change. Therefore no common-mode rejection of power supply imperfections is possible.
So far, we both agree. Where the divergence comes in is where cutoff typically takes place. You say that running in class A "would require enormous standby power and heat generation". This is true for high-power solid-state amps, but most vacuum-tube amps already run very close to class A.
Let's look at a Dynaco Stereo 70 (about as typical as you can get). The output tubes are run at 50 mA each. The primary impedance is 4300 ohms, for a turns ratio of 23:1 for the 8 ohm tap. This means that the amp can deliver 2.3 amps peak current before one of the tubes cuts off. In turn, this means that a Stereo 70 will deliver over 21 of its 35 watts in class A mode.
I don't know about solid-state amps in general, but this seems to imply particularly flimsy design if they are as bad as this, so that the tube amp sounds more powerful due to power supply rejection. Perhaps it is most true of mid-fi amps and receivers, and explains why there are so many tweak postings about adding capacitance to the power supply filters.
...so if the supply voltage sagged 10% at full power, then it would sag 5% at half power.
Gee, a first rate discussion characterized by (pretty much) defined terms, data and science, and honest disagreements that did not descend to "your-old-man-wears-panty-hose" emotionalism. Thanks all.
The solid-state equivalent to a tube push-pull stage is a complementary output, with NPN between the positive rail and the load, and PNP between the load and the negative rail.Until recently, there were no good matches of NPN and PNP power transistors for audio service. The PNP uses holes in the base as minority carriers, and these have much lower mobility than the electrons in the NPN bases.
ON Semiconductor makes a nice complementary pair: MJL3281A and MJL1302A. These are used in the InnerSound ESL-300 and similar amps.
However, nothing in my experience beats Ralph Karsten's Circlotronic OTL tube output stage.
Al wrote,"The solid-state equivalent to a tube push-pull stage is a complementary output." This is incorrect.
:)
Since positive and negative signals suck from the same 'teat', push-pull vacuum tubes might be different from comp-symmetry amp designs that don't. Keep an open mind, Al. You might learn something,
Say John, since we now have some good quality, high voltage power devices available in both bipolar and MOSFET flavors, y'ever thought of implementing some of the more classic tube topologies in silicon?I think it would be interesting. At the very least it would give us a much more apples to apples comparison of tubes versus transistors than what's usually offered up (i.e. comparing single-ended triode amps to complimentary-symmetry solid state amps).
se
Since positive and negative signals suck from the same 'teat', push-pull vacuum tubes might be different from comp-symmetry amp designs that don't.
Sure. Not to mention that the output stage in a solid state CS amp is typically acting as a follower, direct coupled to the loudspeaker rather than as an amplifier driving the speakers through an output transformer as is typical in PP tube amps.
se
The other thing you want to look at is odd-ordered harmonic generation. Transistor amps, in general, tend to generate more odd harmonics in the region of the 9th harmonic and beyond- these harmonics are used by the human ear as loudness cues. Thus a transistor amplifier can develop harshness in the highs, prompting one to turn the volume down. Tube amplifiers do this to a far lesser extent and so the volume control resides at higher levels of output without pain.This often give the tube amplifier and impression of more power, as between this and the generally softer clipping character (read, again: less odd-ordered harmonics) allow one to push the tube amplifier into clipping without immediately bad results.
IOW, watts are still watts and so distortion is distortion. Tube amplifiers make more distortion in general, but actually less of the distortion we don't like. Our ears just do not like odd-ordered harmonics. Tube amplifiers follow the rules that our ears have evolved to a bit better then transistor amplifiers.
IOW, watts are still watts and so distortion is distortion. Tube amplifiers make more distortion in general, but actually less of the distortion we don't like.
Not sure that's the case.
Our ears just do not like odd-ordered harmonics.
If that's the case, how can we stand to listen to music at all given that many musical instruments produce prodigous amounts of odd-ordered harmonics?
Tube amplifiers follow the rules that our ears have evolved to a bit better then transistor amplifiers.
Still not sure that's the case.
Some examples (if you don't look at all of them, be sure to go down to the last one).
Here's the spectra of the Musical Reference RM200. Its spectra is predominantly odd-ordered (indicated by the red arrows), including some stuff above the ninth as well as well above the ninth.
Here's the spectra of the Hovland Sapphire. Yikes! 'Nuff said. :)
The BAT VK-150SE.
The Cary 280 in triode mode.
The Wavelength Gemini 45.
The Lamm ML1. Another example of odd-ordered dominant spectra, though it tapers off rather quickly without any really high order stuff.
The Sonic Frontiers Power 3.
The Nagra VPA. As with the Hovland, yikes! :)
The VAC PA 80/80.
The Lamm ML2.1.
The ARC VS-110.
The Air Tight ATM-211.
The Manley Neo Classic 250 in triode mode.
The Quicksilver something or other.
For the last image, the Final Labs Music 6.
This isn't a tube amp. It's a battery powered solid state amp that uses inexpensive monolithic power opamps. It's spectra is actually even-order dominant throughout most of the spectrum (owing probably to the asymmetrical quasi-complimentary output stages used by monolithic devices), with odd-order overtaking at the 11th and 13th harmonics.
Bottom line, both in terms of odd-order and high-order distortion, it's no worse and even better than most all of the tube amps above.
So maybe it's not just "transistor amps." Maybe it's all those big-ass discrete, fully complimentary transistor amps. :)
se
P.S. I believe my use of distortion spectra graphs from Stereophile falls under the fair use doctrine as I've used them here. I'll leave JA and/or The Bored to make the final call.
se
Do you have any spectra of how these things do when reacting with something other then a sine wave? Trust me it would be very different! That's where we usually get into trouble with the graphs.GE did well-documented research back in the 60s that demonstrated the human response to even and odd ordered harmonics and it is that to which I refer.
Of course, musical instruments have odd ordered harmonics. Most sounds in nature do- what GE found was that responding to that for loudness cues is much more efficient then responding to the entire sound.
You might try a simple test- play a 400Hz sine wave at 0db, then distort the hell out of it and play it back at -20db and see which sounds louder- This experiment is a real eye opener. Even though the meter is telling you that the undistorted signal has more energy, your ears want to tell you something else as they use different rules.
What seems to be needed to really see what is going on is a test that allows one to measure the harmonics generated with an input waveform that is in constant and complex change. So far no such a test has been devised so the results we get do not always reflect what we hear. The industry has spent way too much time measuring things of no importance ...
Do you have any spectra of how these things do when reacting with something other then a sine wave?
Nope.
Trust me it would be very different!
Well sure. Simply using something other than a single sinewave would guarantee that. :)
That's where we usually get into trouble with the graphs.
That of course depends on what you're expecting of the graphs.
GE did well-documented research back in the 60s that demonstrated the human response to even and odd ordered harmonics and it is that to which I refer.
GE? As in General Electric? Can't say I'm familiar with it. Do you have a more specific citation so I can try and dig it up?
You might try a simple test- play a 400Hz sine wave at 0db, then distort the hell out of it and play it back at -20db and see which sounds louder- This experiment is a real eye opener.
Nothing particularly eye opening there. I've been playing electric guitar for over 25 years. :)
Even though the meter is telling you that the undistorted signal has more energy, your ears want to tell you something else as they use different rules.
What "meter" do you mean exactly? A simple FFT will tell you that the undistorted waveform has less energy.
What seems to be needed to really see what is going on is a test that allows one to measure the harmonics generated with an input waveform that is in constant and complex change. So far no such a test has been devised...
Since when? What about the null test that David Hafler popularized some decades ago? Subtract the input and output from each other (accounting for gain) and feed the residual to an FFT.
Nice thing about this approach is you can use any signal, including actual music.
The industry has spent way too much time measuring things of no importance ...
Well, you're in the industry. What things of importance have you spent time measuring? ;)
se
GE in this case is indeed General Electric. What they found was that people will tolerate up to nearly 40% of even ordered harmonic distortion, but odd-ordered, and particularly the harmonics past the 7th or 9th, are not tolerated even in 'insignificant' quantities.With respect to the little test: The meter would be a VU meter, displaying the signal driving an 8 ohm speaker. In this case, a distorted waveform at -20db (considerably less energy (!) then the undistorted waveform), can easily sound louder. although it clearly is not as far as the instruments are concerned.
However, we *percive* it as louder, due to the function of our ears.
Hafler's test is a good step in the right direction, but suffers the same problem that a lot of physics tests do- that the results change depending on the observer, in this case the null circuit has problems of its own!
As far as industry- I've been touting this stuff for years, not that too many have listened. At any rate I have designed our gear to generate as few odd ordered harmonics as possible. It is actually possible to identify things in a circuit that do that.... so, as 'industry', that's been what I have tried to contribute, FWIW.
GE in this case is indeed General Electric. What they found was that people will tolerate up to nearly 40% of even ordered harmonic distortion, but odd-ordered, and particularly the harmonics past the 7th or 9th, are not tolerated even in 'insignificant' quantities.
Ok. So you can't give me anything more specific about its publication so I can try and locate a copy?
With respect to the little test: The meter would be a VU meter, displaying the signal driving an 8 ohm speaker. In this case, a distorted waveform at -20db (considerably less energy (!) then the undistorted waveform), can easily sound louder. although it clearly is not as far as the instruments are concerned.
Um, a VU meter? I thought you might have been referring to some real test and measurement gear. Not something as crude as a VU meter.
What's your point?
Hafler's test is a good step in the right direction, but suffers the same problem that a lot of physics tests do- that the results change depending on the observer, in this case the null circuit has problems of its own!
I really hope you're not alluding to the Heisenberg Uncertainty Principle here.
As far as industry- I've been touting this stuff for years, not that too many have listened.
Then stop touting and start doing. Why do you expect someone else to do it for you?
At any rate I have designed our gear to generate as few odd ordered harmonics as possible. It is actually possible to identify things in a circuit that do that.... so, as 'industry', that's been what I have tried to contribute, FWIW.
That's cool.
se
> I really hope you're not alluding to the Heisenberg Uncertainty Principle here.You mean where its not certain if there's another can of cat food for Schrodinger's cat? Wasn't there another can before you opened the cupboard? And does it matter, since the cat may not be around for it anyway...
Nah. I just don't like the test as saying much- too many variables.
You mean where its not certain if there's another can of cat food for Schrodinger's cat? Wasn't there another can before you opened the cupboard? And does it matter, since the cat may not be around for it anyway...
Say what!? Dude, Shroeder didn't have no damn cat. Shroeder had a piano. Charlie Brown had a dog. Now he was a really cool cat, but still a damn dog. So where you gettin' this Schroeder's cat stuff, man? Don't you ever read the comics? ;)
Nah. I just don't like the test as saying much- too many variables.
Fair 'nuff.
se
I've often commented on the fact that distortion adds "apparent volume" to a system i.e. it sounds louder than it really is. The less distorted a system is, the louder that you can play it without listening fatigue. This is something that i've learned from experience, not actual tests that i've performed or read about elsewhere. It's also one of the reasons why i believe that increased levels of dynamic headroom in an amplifier and dynamic range in a speaker system result in a much more "listenable" i.e. "lower distortion" / "natural sounding" system. Nothing is stressed, so distortions ( of all types ) are reduced.It's good to see that there are others that have not only noticed this, but also conducted the proper tests and studies that support my own personal experiences. Thanks for sharing that bit of info with us Ralph. Sean
>
Which one sounds best?
Which one makes the best sounding music?
Which one sounds best?
Which one makes the best sounding music?
To whom?
se
THAT helps.Of course you could have lied a little....
Good job.
To whom?Now that's the real test. Good one, Steve-o.
Looks like a "who's who in distortion generators", not a "who's who in high linearity electronics". Can you imagine what these things look like trying to drive a reactive load under dynamic signal conditions instead of just a dummy load and a steady state signal??? Like i've said in the past, most of high end audio is over-priced & under-designed junk. Thanks for proving my point. Sean
>* That laugh was patented by Jack G and was reproduced without his authorized consent. Having said that, i'm at least acknowledging and accrediting the originator of said laugh : )
It's also worth pointing out that the higher order harmonic distortion is far more objectionable than lower order, for example 3rd order is much less offensive than 7th order, but it's still not a good thing.Even order harmonic distortion is euphonic, although like odd order, the higher the order the worse it is. Second harmonic distortion is used by some designers to make a component sound warmer; too much even order distortion and you get syrupy, even muddy, sound. Better not to have any.
nt
Watts = Volts x Amps x cos(phase angle)Speakers are seldom purely resistive loads, certainly not at all frequencies.
And those strange phase angles are what cause loading problems and instability in the amp, resulting in time delay ( smearing ), increased harmonic content and deviations in amplitude linearity. This is why i said that when we can have an amp that will drive ALL loads equally, we'll have solved a large part of the problem. Sean
>
I'm definitly not an expert on the subject but here is where the thinking comes from.First you are correct a watt is a watt, be it tube, SS or digital. So on paper at least a 40wpc SS amp will put out the same amount of power as a 40wpc tube amp.
Where the difference is suppose to be, is when the amps are driven to higher levels. When clipping occurs music will tend to sound harsh, brittle and edgy. Increasing all the more as we get deeper into clipping. When you look at a picture of a soundwave clipping the top is flat or clipped.
When you look at the picture of a tube amp driven to clipping levels the osciliscope picture of the soundwave looks different. It dosen't have the same clipped shape. Clipping also tends to be more gradual. So when clipping occurs with a tube amp it tends to not be as audible and have a more gradual onset. Also the clipping pattern unlike a SS amp with the flattened top does not tend to damage speakers as easily.
All things being equal; the 40wpc SS amp with its clipping patteren will tend to sound harsher and damage speakers earlier then a tube amp.
Whats really happening is the 40wpc tube amp can be driven to higher level before audible distortion and speaker damage sets in. This in effect allows more of the 40wpc to be used before distortion and speaker damage sets in.
This of course is just a general rule. But this is where the thinking one tube watt equals two SS watts.
BW Maxx
The differences come into play LONG before clipping sets in i.e. transient response and circuit stability related. An amp doesn't have to "clip" in order for distortion to be produced or the signal smeared. If it isn't "fast enough", the transients lack intensity and the peaks aren't of the proper amplitude. These are both distortions, but are different in the fact that they deal with both the volume of the signal and the duration of the signal.It is a commonly understood fact that MOST ( not "ALL" ) tubed circuits are TYPICALLY a LOT slower than most well designed SS gear. In effect, the slow speed of the tubed gear "saves itself" from damaging speakers that might otherwise be exposed to a higher amplitude and longer duration signal that would be produced by the "more responsive" SS gear. It is also that "slow and round" quality in tube circuitry that helps maintain a more consistent average power, which would otherwise be more dynamic by nature IF the output was linearly tracking the amplitude of the input. Our ears hear that difference in average power vs dynamic power, even though the variances may be neither long in duration or large in amplitude. This is why amps are measured in RMS wattage ( a conservative "average" figure ) rather than in peak power. We are FAR more sensitive to sustained energy levels than we are to dynamic variations.
Obviously, there are some exceptions to the aforementioned generalizations, especially in the tube camp. As an example, Atmasphere tubed amps are quite fast in terms of transient response making them a notable exception, but they lack stability ( due to their somewhat higher output impedance ). As such, they simply aren't a good match for lower impedance speaker loads. If one can mate them with an appropriate loudspeaker load i.e. one that stays above 8 ohms or more, they can achieve most of the benefits associated with the "speed" of SS without incurring many of the drawbacks associated with the "slow & round" tube sound. Obviously, this takes a lot of care and thought to build such a system, but then again, isn't that why we are here to begin with??? : ) Sean
>
You are on the right track, and stability is indeed a concern, but I don't agree that solid-state and tube circuits differ greatly in "speed" if they are made in similar fashions (ie, large chassis, discrete transistors in circuits similar in size to tube layouts).Very small transistors (integrated circuits) can be made to switch very quickly, but these are not suitable for audio service.
The Routh-Hurwitz Stability Criterion applies to any feedback amplifier, regardless of whether it is tube or solid-state. The time delay through a stage depends on the driving resistance and input capacitance, and these are to be determined by the designer (assuming he knows what he is doing). Any amp that incurs more than 180 degrees of total phase shift at the unity-gain upper frequency is unstable. Amps that come close are marginally stable and will suffer from the audible problems you described. There are plenty of marginally-stable solid-state amps.
Atma-Sphere amps do not suffer from stability problems. The high output impedance makes them unsuitable for low-impedance loads, but not because of stability concerns. Paul Speltz' ZERO autotransformers are an excellent method of getting the best sound with Atma-Sphere amps and low-impedance speakers.
One can pick and choose the criteria as to what is or isn't a stable design and where you draw the line, but when the reflected EMF of a loudspeaker modulates the output stage of an amp, that's directly due to a lack of contral via instability. This occurs quite frequently, and not just with high impedance output tubed designs. Sean
>
If the first watt doesn't sound good, then all the rest are nothing but numbers. I would rather have a tube watt than any solid state watt. There is a distortion not being measured that only the ears can hear. Listener fatigue is not measurable either, but it exists.
There is a difference between first listening, and long term listening. Sometimes what sounds good for 5 minutes will eat away at you in the long run.
Having said that, a watt is a watt. I have a Dyanco Stereo 70 II that I wouldn't trade for the world, but I can't push it into low efficiency Magnapan 10.1 speakers and expect it to produce the same volume that a SS amp can make without audible clipping. I still prefer to listen to the tube amp at a lower volume for my long term musical enjoyment though.
the first watts - plural.My speakers have 86 dB sensitivity and I listen at a bit over 2 metres distance in the nearfield where doubling distance reduces level by closer to 6 dB. The pair of speakers would give 89 dB at 1 metre for 1 watt input (doubling due to 2 speakers) but doubling the distance to 2 metres losed 6 dB for that 1 watt input leaving me with a level of 83 dB or so for that first watt. I don't listen at loud levels, but I do prefer listening levels closer to 85 dB or so, up to 90 dB or so on some music and that takes more than 1 watt. Note that the above calculation ignores room effects and my room has a fair amount of absorption so I don't get as much reinforcement from it as some do. That may end up requiring a bit more power. Its the first 5 to 10 watts or so that are important for me.
The first watt line is really only true for high efficiency speakers. It's definitely not true for low efficiency speakers unless you like listening at really low levels.
This is exactly what i said above i.e. that appr 8 - 10 watts of Class A power is typically sufficient for average efficiency speakers with "reasonable" listening levels and 8 ohm loads at typical listening distances. Alter any of those variables i.e. lower impedance speakers, lower efficiency speakers, louder listening levels, greater listening distances, etc... and you need a higher level of Class A power. As such, it is good to see those with common sense also share common listening skills and experiences : ) Sean
>
Everyone has thier opinions. MY thoughtson the whole tube vs SS is this. Theres good SS and Tube amps and bad tube and SS amp. If the amp is very well designed and built it will sound good regardless if its tube or SS. I use both tubes and SS in my system. I prefer to have tubes in the pre-amp section and SS for the power amp, I find this givbes you the best of both worlds, the natural, warm and liquid sound of tubes with the speed, bass slam and power/speaker driving ability of SS. I think both tubes and SS have stereotypes attached to them, most people think tubes are slow, laid back, warm and polite sounding, while SS is fast, bright, grainy and not very musical, I think this is BS! I think that describes the sound of badly designed amps, because ive heard amps that were just the opposite of those stereotypes from both tube and SS. To quick examples
1. manley labs-there tube stuff is very fast and powerful and not overly smooth or laid back at all
2. Plinius-very smooth and warm sounding not bright or harsh at allSo I think these stereotypes come from people hearing a badly designed amp and or using the amp in a system that its not meant to be used in. SS and Tubes can both sound excellent
My ears lied to me also. Nothing worse than all that tube noise ya'll pretend not to hear. If none of you tunephiles dont hear it your not true to yourself.The Dynamics & flat out control & response and bass just doesn't happen with tubes. SS as usual kicks butt everytime.
You tube guys just keep pumping each other up. Ya'll need the internal support,(all that money spent)Iam sure they all believe tubes are the best. Until You Switch. And can enjoy your music & not the system. See its the Music, Not the System.
Your words, two different opinions.
I have had many solid state amps in my lifetime. Some were very expensive, others were not. Brands included NAD, Carver, B&K, and Sherbourne. Not one of them brought me closer to the music than my dad's Scott 299C that he built in the 1960's with 17-25 watts per channel. I kept trying to replicate that sound with thousands of dollars wasted on SS gear. I finally found out that the only way to get the tube sound is to buy a tube amplifier.
If you want to tell me I'm hearing noise, then you must know something I don't. Good for you. I'll enjoy the music as much as you do, with my Dynaco Stereo 70 II!! See, It's the music, not the system. That's what it's all about.
I used a SS phono preamp and a SS integrated. I switched to a tube phono preamp (Art Audio Vinyl One) and have enjoyed the difference. The tube phono unit is dead quiet. I realize that not all are, but this one is. Like anything else in this hobby, I think its very wrong to place absolutes on a certain topology or architecture. It comes down to implementation by the equipment designers.I still like my SS integrated (Plinius 9200) but may eventually move to a tube unit after my positive experience with the phono preamp.
For what it's worth a good triode tube is the most linear output device available(defined as if 1 volt in gives 10 out then 2 volts in should give 20 out). Without feedback, trodes, pentodes, which includes solid state devices don't come near. And that is the reason a proper tube amp sounds more powerful until it clips of couse(and if clipping is at ten watts it doesn't matter if it's tubes or solid state).
Without feedback, trodes, pentodes, which includes solid state devices don't come near.
Perhaps due to the inherent negative feedback in the triodes themselves.
Kind of apples and oranges to compare a tube with inherent negative feedback to tubes without it and without applied negative feedback isn't it?
se
It's my understanding that it's the circuit, not the active device, that implements any feedback i.e., the unbypassed cathode resistor allowing local feedback. Perhaps you could explain how a device with 3 physically and electrically separate elements posess it inherently, and why a simple npn transistor is immune?
It's my understanding that it's the circuit, not the active device, that implements any feedback i.e., the unbypassed cathode resistor allowing local feedback. Perhaps you could explain how a device with 3 physically and electrically separate elements posess it inherently, and why a simple npn transistor is immune?
The three elements in a triode while physically separate are not entirely electrically separate. That's why screens are added to tubes such as pentodes, to further electrically separate them so as to reduce the inherent internal negative feedback which allows for more gain in a single stage so that global negative feedback can be more effective.
Here's a piece on the internal negative feedback of triodes by H. Sotckman, published originally in the April, 1953 issue of Wireless Engineer:
se
To take this paper from Bill Perkins' website?For those interested in vacuum tube technology, Bill Perkins website has many more technical papers available for free perusal. Bill runs Perkins' Electro-Acoustic Research Labs (PEARL), most famous for his tube coolers.
The link below is to his literature archive page. Bill has spent years gathering old seminal articles on vacuum tube technology. He is now in the process of reformatting them into digitally generated PDF files (instead of simply scanning them). I doubt he is happy that someone is appropriating his work without credit...
To take this paper from Bill Perkins' website?
No. The document's in the public domain. Why would I need permission?
Besides, I didn't download it from Bill Perkins' website. I downloaded it over a year ago from Ken Gilbert's website.
I posted a link to it in March, 2004 on Prop Heads:
http://www.audioasylum.com/forums/prophead/messages/8195.html
It was (and still is) in a collection of technical articles and it said nothing more than "Ever wonder why triodes are so damned linear? They can be thought of as pentodes with negative feedback applied. For more info, click here ."
I doubt he is happy that someone is appropriating his work without credit...
I wouldn't know. Though I can say I'm not too happy with your jumping to conclusions and making false accusations against me.
se
... I didn't copy that CD. I copied it from someone else who copied it.
... I didn't copy that CD. I copied it from someone else who copied it.
Bad analogy.
Bill Perkins does not hold any copyright to that article whatsoever. In fact, he's technically in violation of copyright law for copying it in toto and disseminating it to the public.
So if you want to get all high and mighty about intellectual proptery, Bill Bill Perkins should be at the head of the line.
Now, about those false accusations you made against me...
se
If you've done much browsing on the internet, you would probably know that linking to someone else's web site is pretty much fair game. On the other hand, copying something from someone else's web site and putting it on your own is generally frowned upon.(In fact, I seem to recall a flap in this very forum about this very issue only a week or two ago....)
I would expect that when a normal person found out they had inadvertantly misappropriated someone else's work to rectify the situation. You know, replace the misappropriated material with a link, or something like that.
On the other hand, I would expect you to argue about this, just like you argue about everything else under the sun. In this, you have exceeded my expectations.
If you've done much browsing on the internet, you would probably know that linking to someone else's web site is pretty much fair game. On the other hand, copying something from someone else's web site and putting it on your own is generally frowned upon.
Yes, when someone has some claim of copyright to the material.
(In fact, I seem to recall a flap in this very forum about this very issue only a week or two ago....)
If we're thinking about the same flap here, that was a case wherein someone copied copyrighted material (in this case from Stereophile's e-newsletter) and made it available to the public.
And ironically, that's also what Bill Perkins has done.
Also, the file isn't on my website. It's on the server that hosts my website. There's no link on my website to that file. Only a direct link to the file in my post.
I would expect that when a normal person found out they had inadvertantly misappropriated someone else's work to rectify the situation.
I don't see that it was "misappropriated." Bill Perkins has no copyright whatsoever to that article.
You falsely accused me of taking the file from Bill's website and posting it here without giving him credit for it.
As I explained, I didn't take the file from his website so I had no idea he was responsible for having created it and therefore could not have given him credit for it at the time I posted it.
He's now been given credit for it. So what's the problem?
se
You say you didn't know that the paper was prepared by Bill Perkins. OK, fair enough. You found the paper on Ken Gilbert's website, and although there are many places where he credits Bill Perkins (including a page dedicated to Bill's white papers), Ken apparently forgot to credit Bill for this one.So the presumably you thought that Ken Gilbert prepared the paper. Then when someone asks about tubes, you download the paper from Ken Gilbert's site to your computer. Then you FTP the paper to your own website. Then you publicly refer people there.
Maybe there is some innocent explanation, but it's not obvious to me. It seems to me that either you wanted people to think your website is a great resource for technical articles or that you meticulously transcribed the original article. Why didn't you just post a link to Ken's website (since you thought he was the one who prepared the paper)?
By the way, how long do you think it would take you transcribe the original article (with complex equations and greek symbols) and reproduce all the circuit diagrams (as Bill obviously did)? Seems to me that Bill put *many* hours of work into this project...
You say you didn't know that the paper was prepared by Bill Perkins.
That's what I said.
OK, fair enough. You found the paper on Ken Gilbert's website, and although there are many places where he credits Bill Perkins (including a page dedicated to Bill's white papers), Ken apparently forgot to credit Bill for this one.
Apparently.
So the presumably you thought that Ken Gilbert prepared the paper.
No. I didn't have any idea who prepared the paper.
Then when someone asks about tubes, you download the paper from Ken Gilbert's site to your computer.
Not quite.
I downloaded the paper from Ken Gilbert's site (via a Google search) over a year ago when I was doing some research of my own regarding the internal feedback of trodes.
It's been on my hard drive ever since.
Then you FTP the paper to your own website. Then you publicly refer people there.
I didn't refer anyone to my website. The link in my post links directly to the pdf file.
And yes, I did FTP it to my web server the other day because I couldn't remember where I'd originally got the file.
After you made your big stink about it and accused me of "misappropriating" it from Bill Perkins' website and intentionally neglected to give him credit, I did a search to find where I'd made mention of it on AA before and that turned up a post from me on Prop Heads.
Until then, I didn't remember where exactly I got it originally.
Maybe there is some innocent explanation, but it's not obvious to me.
Gee, no surprise there.
It seems to me that either you wanted people to think your website is a great resource for technical articles or that you meticulously transcribed the original article
Huh?
How would it have made anyone think that my website is a great resource for techincal articles when anyone who actually went to my website wouldn't find any link to that or any other technical article there at all?
And why would I give a shit about whether or not anyone thought that I had meticulously transcribed the article myself? What, I think some headhunder looking for an ace transcriber might stumble upon it and offer me a lucrative job?
Put down the crack pipe, Charles.
I uploaded the file to my web server and linked to it in my post for the simple and mundane reason that I didn't remember where I'd got it from originally, that I've had it on my hard drive for over a year and already knew where it was and it didn't take me more than about 30 seconds to do it.
Why didn't you just post a link to Ken's website (since you thought he was the one who prepared the paper)?
Again, I had no idea who had actually prepared the paper.
By the way, how long do you think it would take you transcribe the original article (with complex equations and greek symbols) and reproduce all the circuit diagrams (as Bill obviously did)?
Maybe a couple hours. Three tops.
It's only three pages. There are no "complex equations" just some simple algebraic expressions. Greek symbols? That's just a mouse click to select the font. And the circuit diagrams are exceedingly simple.
Seems to me that Bill put *many* hours of work into this project...
Great. It's appreciated. And he's been given credit. What, you want me to offer mow his lawn or wash his car for him or something? What?
Now, what of your false accusations against me? You seem to be doing a pretty good job of dancing around that.
se
It appears that you intended no slight of Bill Perkins or PEARL. I appreciate you filling in the details.
It appears that you intended no slight of Bill Perkins or PEARL. I appreciate you filling in the details.
Sure.
Now I'd appreciate an apology for the false accusations.
se
Let's see now...I asked "Did you get permission to take this paper from Bill Perkins' website?" It turns out that you didn't take it from Perkins' website, but from Ken Gilbert's website instead. And you want an apology? For what?
Everyone here knows that you are the one who should be apologizing for your continuing stream of rude, profane posts full of personal insults.
I asked "Did you get permission to take this paper from Bill Perkins' website?" It turns out that you didn't take it from Perkins' website, but from Ken Gilbert's website instead.
Yes. You falsely accused me of having taken it from Perkins' website.
"I doubt he is happy that someone is appropriating his work without credit... "
se
... are you saying that I made a "false accusation" when I said "I doubt he [Bill Perkins] is happy that someone is appropriating his work without credit"?Doesn't seem like a "false accusation" to me. Seems like I stated my opinion. Do you think he *is* happy about it?
How about we call him up and ask him?
... are you saying that I made a "false accusation" when I said "I doubt he [Bill Perkins] is happy that someone is appropriating his work without credit"?
No. I added that just to underscore the false accusation itself. That being your having falsely accused me of taking the file from Bill Perkins' website .
Did you get permission to take this paper from Bill Perkins' website?
Whether I asked permission or not, the accusation is that I took it from Bill Perkins' website.
I doubt he is happy that someone is appropriating his work without credit...
Appropriating it from where? From the first quote, from Bill Perkins' website.
So, whether I asked permission or not, or whether Bill Perkins is happy or not, you accused me of taking it from his website . Which I did not. Making the accusation a false one.
Is that too difficult for you to grasp? Or am I expecing too much?
se
Bill Perkins carefully prepared a reprint of a 50 year old paper by Stockman. It ended up on your website. I asked you if you had permission to do so. You explained (to my satisfaction, at least) that this was done innocently.Now to me (and also to most people, I would imagine), that would be the end of the story. But for some reason you seem to think you deserve an apology from me. I don't and you're not getting one. Is that too difficult for you to grasp?
Never ceases to amaze me how people will argue till they are blue in face, rather than apologise. Mr. Charles Hansen you owe the man an apology for accusing him wrongly of taking the paper from Mr. Perkins, which clearly he did not. The paper is in public domain, and Mr. Gilbert may have acquired the paper without the assistance of Mr. Perkins. I have included a link to another site where the article in question is published.
Music making the painting, recording it the photograph
Thanks for the input. If it were anybody else besides Steve Eddy, I might have a different response. However if you've followed his postings for long, you would know that he doesn't deserve an apology from me for asking the question that I asked him. See the link below for more information on this.Regarding the article itself, the link you provided was to a different version. Without question, the version that Gilbert posted on his website (and Eddy subsequently uploaded to his server) was the version that Perkins prepared. You can either compare it with the version at the PEARL website:
http://www.pearl-hifi.com/06_Lit_Archive/07_Misc_Downloads/1150_Inherent_Triode_Fdback.pdf
or simply look at the footnote below the references, where Bill Perkins' initials appear.
it's an interesting read. I would favor the term degeneration for the electrical action, instead of feedback. thanks.
I would favor the term degeneration for the electrical action, instead of feedback.
A rose by any other name...
But if it makes you feel better, I won't argue. :)
se
Well, your understanding is wrong, the control grid is implementing a form of negative feedback, applying negative bias to the control grid linearizes the current delivery of the device, since a simple npn transistor does not have a control grid, negative feedback has to be implemented in the circuit.
Music making the painting, recording it the photograph
transistors also have bias, and you might say the base is the equivalent to the control grid. Please explain how a control grid in itself defines feedback without a circuit. It still seems to me that your explanation is circuit defined, not inherent. If you are refering to leakage capacitance between grids, let's keep the discussion in respect to audio frequencies. And how does applying a separate neg. control voltage to the control grid induce current or voltage feedback between grids?if you look at v. vs c. tube curves for a 300b triode, you will see straight lines graphed on a good portion of each for various control voltages indicating perfect linearity and no feedback.
Sorry to butt in here, but SE will make sure that you are thoroughly confused, if I don't.
Technically triodes have a form of feedback, because the output current current is changed by the output voltage. With pentodes, ideal bipolar transistors or fets, this is not true, and they are free of this 'feedback' mechanism.
Actually, triodes are probably the most linear device available. This is because the change in output conductance that occurs with changes with the output voltage at the plate of the tube, tends to cancel the rising transconductance (input voltage to output current) to linearise (make equal) the spacing BETWEEN the lines as seen on a curve tracer, (see previous article cited on tubes).
This effect can also happen with fets and bipolar transistors, and is more tricky to implement. For example it is possible to cancel the 2'nd harmonic in a single fet or bipolar transistor without any push-pull action by careful selection of device, load resistance and operating point. For further explanation, contact Bascom King, who showed me this over 30 years ago.
In any case, triodes are generally the most linear devices, and the first watt is the most important one, in any amplifer. Why? Because that is where we normally operate when listening to music. Higher powers are usually reserved for special effects, like cannons, explosions or very loud rock.
If you are in doubt, just put a simple AC voltmeter across one of your loudspeakers and not its average output. Then convert what you find to watts. You might be surprised.
However, IF you want significantly higher levels, to create explosions, or to impress your friends, you will find that it takes lots of watts, and that is one reason why we usually make much larger amps, the minimum usually being 100W.
Hi John, and thanks for taking the time for a proper explanation. My Radiotron is packed away from a recent move and could not reference. I knew there had to be a specific mechanism involved, more than just a parroted heresay...
TAZ!! You couldn't have stated that more perfectly.So many times we will make a change in our system and your first intincts are telling you that is was an improvement being that it sounds different.Many times it is an instant improvement perception but once things settle back in and you start the analytical process,you realize that the change wasnt good.BTW I would also take a tube watt anytime over a SS watt and im sure 99% of the people here would.Now your st70II could push the maggies if you have good beefy power supply but you may have to biamp.I use a modded dyna st70 on a pair of acoustat model 2s from time to time in full range but it does break up on super heavy passages like the 1812 which most low power amps do.
Speakers are driven by voltage. As such, the amp with the higher voltage potential will typically sound both cleaner and more powerful, so long as it has the necessary current required to sustain that voltage under any operating conditions. Given that tubes tend to run at a higher rail voltage, they exhibit traits that follow that supposition, but only up to a point. Most tubes aren't capable of a high current output, so that's where they run into trouble. On the other hand, SS amps typically run at a lower rail voltage, but are capable of passing quite a bit more current. This is why many tubed designs offer "finesse" whereas many SS designs offer "brute force".With that info at hand, it would make sense that tubes would work best with a higher impedance speaker i.e. a load that didn't require much current. This would allow one to achieve all of the benefits of the design with very little of the drawbacks. At the same time, a well designed SS amp is at home with a wider variety of speakers, but the sound may not be as good at any given time as what one could achieve with a well designed tube amp. Then again, tubes tend to introduce a higher noise floor into the equation with increased levels of distortion, whereas SS amps are typically much quieter and can be more linear in operation. Each design approach has pro's and con's.
Having said that, the ultimate would be to have an amplifier that ran very high rail voltages with gobs of current, a low output impedance with a very high S/N ratio, very fast transient response with a very wide power bandwidth, circuitry that was naturally linear in operation so that there would be minimal need for negative feedback, was biased into Class A operation above the power levels required for the average listening session and had more than adequate amounts of heatsinking with proper ventilation. In most cases, 10 watts of Class A power would be adequate for a nominal 8 ohm load, but with a lower impedance or low sensitivity speaker, higher levels may be required. After that point, the amp could run in Class B mode for what should be several hundred watts per channel. Power should "double down" as impedance is halved, not only at the rated output, but at the point of clipping also. This means both a hefty power supply and a high current capacity output stage.
Extremely high efficiency designs ( 100+ dB's @ 1 watt @ 1 meter ) still require enough voltage and current to overcome thermal loading losses and reflected EMF, so the available power should be kept up as high as is possible while still being able to retain the aforementioned design goals.
With those variables in mind, an SS amp has more potential to achieve those results than a tubed design. At least, the tube designs using tubes that are currently being marketed for "audiophile" purposes. Why a manufacturer hasn't produced such a product is beyond me, although i do see some heading in that direction. Until we get to that point, not all amps or "watts" are created equally. That's because not all amps, tubed or ss, can deliver the same amount of power with the same amount of quality into all loads equally, nor do they recover from impedance changes in the load or dynamic changes in the signal in the same fashion. As mentioned above, an amp that is stable into any given load is what we are looking for, and even then, that stability has to be achieved with a high level of linearity and proper design traits. Too many engineers attempt to seek those traits, but utilize circuit design trickery ( gobs of feedback ) to cover up the lack of quality in the circuitry that they've devised to attempt such a task.
As such, pick your poison i.e. tubed or SS. Just make sure that the other components in your system and listening environment are the antidote for that poison, as a good system is one that balances the pro's and con's in equal manner. Otherwise, specific areas are highlighted and others are treated in a subtractive manner. While this type of system can be quite enjoyable to listen to, it is strictly a matter of personal preference and might not be to universal liking. Sean
>
Didn't I see a post of yours about Sunfires and some modifications? If so, how are you running them (current output, voltage output, combination)?
The modified Sunfire's are running in my HT system, which is one of five systems in my house. Both amps are fed in balanced form from the Pre-Pro.For the front mains, i'm running a modified Sunfire Signature into heavily modified AR 9's. In case you're not familiar with them, the 9's run a .75" dome tweeter, a 1.5" dome upper midrange, an 8" mid-woofer and dual 12" low Q woofers in a 4+ cu ft sealed cabinet. These are bi-wired using Kimber 4TC to the high frequency section with Kimber 8TC to the woofers. Both the top and bottom sections of the speaker are fed from the "voltage source" as labeled from the Sunfire factory. If you know anything about these amps, then you know that this is really the "current source", not the other way around. This amp was "dyno'd" at 1480 wpc @ 4 ohms. It is fed from a 30 amp breaker with 8 gauge wire.
For the center, rears and subs, i'm running a modified Sunfire Cinema Grand Signature. The rears are highly modified AR 90's, which are basically slightly smaller versions of the 9's. They use a slighlty shorter cabinet housing the same .75" dome tweeter, 1.5" dome upper mid, 8" mid-woofer and dual 10" drivers rather than the 9's dual 12's.
The center channel is a custom built piece using OEM AR parts that match the parts used in the high frequency section of the 9's and 90's i.e. dual 8" mid-woofers, a 1.5" dome upper midrange and a .75" dome tweeter. The domes are actually part of a rare "plate assembly" that was used on the AR 98's. The subs sometime vary in this system, but i'm currently using two large low Q sealed boxes with dual 12" drivers per box. All of these speakers are a nominal 4 ohm load, at which this 5 channel amp is rated to deliver 810 wpc RMS. This amp, like the other amp, is fed from a 30 amp breaker with 8 gauge wire.
All of the speakers were built / modified in similar fashion. That is, all of the baskets on the drivers were heavily damped. This drastically increased the mass of the baskets, reducing ringing and improving transparency. On top of that, all of the switched attenuators were bypassed and hardwired in the "wide open" position, as this gave me the shortest signal path and the most linear frequency response. The cabinets all received an internal layer of vinyl damping sheets, additional bracing and a small amount of fiberglass stuffing along with the factory supplied poly damping material. Internal wiring consists of Kimber 4TC and 8TC to the appropriate sections.
The crossover network was rebuilt using high grade poly caps and the two boards condensed down into one. Not only did this make wiring it up much easier, it also shortened the signal path and the amount of connections needed to get the job done. Special attention was paid to how the crossover parts were oriented so as to minimize crosstalk between the various parts of the ciruit due to EM coupling via the fields generated near the inductors. Both sets of binding posts per cabinet were replaced with pieces manufactured by Axon. Due to how the crossover networks had to be laid out, the binding posts are now on the bottom of the cabinet rather than in the back.
Most of my other systems utilize actively crossed speakers with multiple amps and / or full range drivers with mono-blocks driving them. The main system uses twelve channels of mono-block amplification that is actively tri-amped to power a line array of electrostic tweeters, electrostatic mids and multiple low Q sealed and stuffed woofers per channel. The aggregate power of this system would be rated at 3800 wpc. In the future, i may add four more channels of amplification and another set of e-stat panels and actively quad amp. Goertz speaker cabling is used through-out this system.
My home office system uses modified 800 wpc mono-blocks to drive a pair of omni-directional full range drivers. No active or passive crossovers here, just one funky looking Walsh driver per box directly connected to the amp via Goertz speaker cabling. On some occassions, i'll substitute a high powered stereo amp into the system just for fun. In doing so, i have to change the interconnects and speaker cabling being used, but the speaker cabling remains Goertz MI of a longer length.
My bedroom system uses two modified 100 wpc stereo amps run in an actively crossed mode. The speakers are small pyramid shaped time-aligned satellites with a low Q sealed and stuffed down-firing sub per channel. Speaker cabling consists of modified XLO for the satellites and home-brewed low inductance 8 gauge for the subs.
The one exception to my "active is best" rule is my tubed system, which uses highly modified horns and is driven by one stereo tubed amp. Goertz speaker cable is used here also.
Whenever possible, i try to use Goertz MI flat series of speaker cabling as i think it is the most electrically and sonically neutral brand on the market. It is hard to beat a wide-bandwidth solid conductor with minimal skin effect wrapped in a high grade dielectric that offers a very low nominal impedance. Other designs try to emulate this simplistic design, but end up both complicating the situation and lowering the performance of the system.
I used the Kimber 4TC and 8TC in my HT system and to re-wire the associated speakers internally as the sligthly leaner tonal balance ( higher nominal impedance ) helped to even out and open up the slightly soft sounding AR drivers. It did this while still keeping inductance and high frequency loss to a minimum.
The XLO cabling was geometrically altered in a way that makes it more comparable to the Kimber design than the XLO design. The XLO design is higher in inductance than the Kimber, which tends to lower the quality and quantity of treble information available. This cable was selected for the bedroom system as it tends to balance out the somewhat "smooth and warm" sound of the electronics being used there. On top of that, i already had it and it was handy : ) Sean
>
A watt's a watt. 2.83V at 8 ohms is a watt. Good ol' Google sez 1 horsepower = 745.699872 watts.What a a speaker (load) does with a watt generated by an amplifier is going to depend on a lot of things.
Tubes tend to compress before they clip, AND when they do clip, it tends to produce more pleasant-sounding harmonic distortion than does a ss amp. This is (IMNSHO) the genesis of the "tube watt" truism.
A measure of power. Shoulda mentioned that before. See link below.
all the best,
mrh
:-)
This post is made possible by the generous support of people like you and our sponsors: