EC8010 said:
That's where I defend global feedback practises from o/p tranny to input stage. It does correct for circuit imbalances especially from phase splitter circuits and corrects response from poor quality output trannies. Although I agree that too much NFB is poor for sound, I've always get good comments when 10-15dB is used in PP UL amps. As EC8010 points out The 10KHz square wave test is pretty infalliable, and anyone with a tube amp doeinmg such a test won't expect to see quite the same in and out square wave performance. That also applies to conventional preamp stages that use ECC83's and the like!
Remember Fourier analysis, example: to reproduce a 15KHz square wave (which is rare in music) the third harmonic must have no attenuation til 45KHz implying the amp must have a bandwidth of that figure at a max 3dB down. If an amp cannot reproduce this then a square wave becomes a sloppy sine wave with considerably high 3rd HD. Tough Yes. So if a particular tube amp claims -3dB down at 40KHz at full power and quote HD figures at that level (ignore 1watt level) it is a very well designed amp with superb trannies.
Nearly all p-p tube amp designs of yesteryear had 3dB response down at 20-40 KHz which satisfied all guitar and hi users at the time. The fact that on paper tube amps have a worse spectral spec than SS amps doesn't imply SS is better.....as many comments show the contrary. All the people I've had around my place noticed an immediate difference in tube sound compared to crummy home theatre sound.If you are using a tube preamp, (now I would mentioned it) it could be creating higher HD and lousy performance than the main amp. So we are back where we are started.
rdf mentioned 0.1% HD over midband on an SE amp. Not mentioning tons % intermodulation HD. ..I would consider too high for norm listening levels especially classics brass and music with alot of deep drum expansion. The upper MR will prob sound like a sick parrot. I fully agree with the 1960 suggestion that a dual 20W tube amp is sufficient for most home hi-fi needs with 93dB sens LS system Yes /no ??
richj
Hi Joel,
I think we're saying the same thing, you from the more coherent end. Essentially engineering aspects beneath the raw numbers count highly. You spoke of loop stability and me of harmonic structure. Neither earth shattering or new yet too often forgetten in the reduction of quality to a single metric, whether it be one THD number, IMD, TID, or zero feedback.
Hi richwalters
The IMD figures are also very, very good for a pentode SE. About 0.5% full power (all of 3 watts) if I recall. I didn't get into much circuit detail because the amps don't use the same output transformers and as I discovered recently IMD is as much affected by power supply and primary inductance as the circuit. The Mullard uses a first-gen Hammond 1628SE - saturation is not an issue! - and the home brew a James. The Hammond has problems up top which can be tamed, but never eliminated, with judicious choice of feedback cap while the James is happily flat past 50 kHz with no feedback at all. It passes a very nice 10 kHz square wave into a resistor and shows the traditional cycle or two of ultrasonic ringing into something more challenging like a fullrange driver. I'm sure the Hammond's design flaws have some impact on what I hear as well but the distortion spectra is such a stand out I can't help but think it's the main contributor.
A PDF's worth a thousand words though. Obviously both amps have some power supply demons to tackle. To return to the original comments, I have a hard time believing from these graphs that the second amp's higher 2nd order distortion is what drives my preference for it.
I think we're saying the same thing, you from the more coherent end. Essentially engineering aspects beneath the raw numbers count highly. You spoke of loop stability and me of harmonic structure. Neither earth shattering or new yet too often forgetten in the reduction of quality to a single metric, whether it be one THD number, IMD, TID, or zero feedback.
Hi richwalters
The IMD figures are also very, very good for a pentode SE. About 0.5% full power (all of 3 watts) if I recall. I didn't get into much circuit detail because the amps don't use the same output transformers and as I discovered recently IMD is as much affected by power supply and primary inductance as the circuit. The Mullard uses a first-gen Hammond 1628SE - saturation is not an issue! - and the home brew a James. The Hammond has problems up top which can be tamed, but never eliminated, with judicious choice of feedback cap while the James is happily flat past 50 kHz with no feedback at all. It passes a very nice 10 kHz square wave into a resistor and shows the traditional cycle or two of ultrasonic ringing into something more challenging like a fullrange driver. I'm sure the Hammond's design flaws have some impact on what I hear as well but the distortion spectra is such a stand out I can't help but think it's the main contributor.
A PDF's worth a thousand words though. Obviously both amps have some power supply demons to tackle. To return to the original comments, I have a hard time believing from these graphs that the second amp's higher 2nd order distortion is what drives my preference for it.
gingertube said:
The correct attire for a walk through an audio forum;
An externally hosted image should be here but it was not working when we last tested it.
Okay I'm beginning to duck behind my bench. In all the years of tube amp revelations it didn't take me long to find that in a conventional 3 or 4 stage p-p amp, how each tube stage has an influence on the loading of the preceeding tube. The situation isn't quite clear cut and far more complex than meets the eye. The Miller effect paints quite a different picture when it comes subtle detail of complex HD between stages, cancelling and addition goes on, albeit at quite low levels. Take the example of the concertina phase splitter followed by a balanced Common Cath push pull voltage driver i.e Williamson type. In all but the worst misbalanced cases that particular configuration does an excellent job in ironing out the slight misbalances created by the phase splitter. Some might say I'm defending the concertina, (yes I am), and mention that using this configuration one is really creating a 4 stage amp. The noise contribution from a split load med mu triode stage can be ignored. My arguement is when it comes to phaseshift, I see it as a 3 stage because the anode of the preceeding tube is usually strapped to the grid of the concertina which doesn't create phaseshift.
In the common cathode type driver looking at the common cathode with a Oss''cope the misbalance can be easily seen, the secret lies but splitting the cathode to ground resistor and then providing a dab of local feedback by strapping the grid resistors. to the split cathode resistor. see dia.
The poor PSRR of the anode (concertina) can be ironed out by the use of global feedback,but I prefer proper stabilised B+.
For those tube newbies wondering the importance of global and local NFB; simply,one can compare NFB to the effects of cycle/cycle correction in switchmode psu topology. i.e correction on a cycle per cycle basis. In a tube amp all goes to pot to high thd when bandwidth slewing starts. How much global NFB ? How much S/N ratio can one tolerate ? so many issues relating to sound quality.
Results; for a start I nearly always use pentodes confiured as triodes. ...even in preamps....the better bandwidth and low capacitance arguments of ex video tubes is compelling evidence to reduce Miller influence through other stages....these make excellent building blocks. The individual stage performances have improved, but we still have a little problem with the high frequency end of the output transformer and how much HF compensation to throw in.The square wave business which EC mentioned is a great help, but becareful with dummy loads and the real LS with a wretched crossover networks. The scope problem can also do it’s silly things with earthing, and where lead placement is...I can assure everyone that the algorithm of HD, NFB and phaseshifts is more than looking at a water melon pulverised by a shot gun.
richj.
In the common cathode type driver looking at the common cathode with a Oss''cope the misbalance can be easily seen, the secret lies but splitting the cathode to ground resistor and then providing a dab of local feedback by strapping the grid resistors. to the split cathode resistor. see dia.
The poor PSRR of the anode (concertina) can be ironed out by the use of global feedback,but I prefer proper stabilised B+.
For those tube newbies wondering the importance of global and local NFB; simply,one can compare NFB to the effects of cycle/cycle correction in switchmode psu topology. i.e correction on a cycle per cycle basis. In a tube amp all goes to pot to high thd when bandwidth slewing starts. How much global NFB ? How much S/N ratio can one tolerate ? so many issues relating to sound quality.
Results; for a start I nearly always use pentodes confiured as triodes. ...even in preamps....the better bandwidth and low capacitance arguments of ex video tubes is compelling evidence to reduce Miller influence through other stages....these make excellent building blocks. The individual stage performances have improved, but we still have a little problem with the high frequency end of the output transformer and how much HF compensation to throw in.The square wave business which EC mentioned is a great help, but becareful with dummy loads and the real LS with a wretched crossover networks. The scope problem can also do it’s silly things with earthing, and where lead placement is...I can assure everyone that the algorithm of HD, NFB and phaseshifts is more than looking at a water melon pulverised by a shot gun.
richj.
Attachments
amperex said:
I agree with the idea of that... that the vast majority of people painfully underestimate what their auditory system is capable of, but I don't think our ears can beat the best electronics anymore.
The problem isn't the electronics, it's our interpretation of what it's telling us.
eeka chu said:
I can't see how there is such a thing as best electronics.........how do you define it ? I recently broused in an electronics store to listen to some posh LS systems and despite so much hype about equipment prowess all I could hear when a particular unit when it was switched on was buzz from the transformer. More noise came from the unit than the s/n. Best hi-fi electronics NO...basically someone in manufacturing had no experience how to put things together.....destined for the muck heap. The most reliable electronics heads are in mil or medical systems and anyone who works in either will tell you things will fail.
richj
Who would argue that the ear is unnecessary? This is simple logic when talking about audio equipment. But the problem with your particular brand of attitude is that you then discount all other forms of measurement and information collection. Being an "ears only" person is as limiting to design as being a "scope only" person.
I guess it's inevitable that some would learn absolutely nothing from this thread. We're back to religion again.
Joel
I'm sorry to be so confrontational... But that statement contains no useful info without a context. It is pointless IMO to discuss any topology without a context. SE amplifiers do not sound good with all speakers. SE amplifiers do not all sound the same. The same amplifier (especially SET) amplifiers do not sound the same on different speakers. For that matter PP amplifiers do not all sound the same, or even SS amplifiers...
I tried 8 watt SE amp on CSS FR125 that does not work..(hey another pointless remark...because it depends on the size of the room. 8 watt SE on an FR125 sounds good in a small computer room..but SUCKS in my living room)
What is it all about? Was your question... and really imo it is all so very very simple.
Because it is only obsolete in terms of efficiency, size, cost etc. No matter what the guys in the advertising departments said...(see attached ad)
Often true.
True, SE's sound bad at full power. Thing is ..if you use the proper speakers you'll never need full power.
That depends on the loudspeakers they have. And looking at most folks that have SE..I would have to say..not they are not. And while we are on the subject of the so called reason for that euphonic tube sound...(distortion)...personally I have never ever heard a warm / romantic sound from a 300B SE amp...
My answer..no we are all positively...absolutely not lovers of grunge. And though it may be difficult to make a PP valve amplifier with a "nice" distortion spectrum, it is possible.
So what is it all about?...It's about making choices on how we like our music reproduced...and then making the right choices on the right mix of equipment. One can have perfect sound forever whether it be through SS, tubes (PP or SE) CD or VINYL, Single Drivers, Home cinema set. Dipole, bipole...whatever. The combination is important, the room is important, the sound level one likes to listen at is important and the person that is listening is important. Last but not least the urge to improve and change even after audio nirvana has been reached must be "cured" through group and phsychotherapy. Or by simply listening to one's wife for a change.
Attachments
Good stuff Bas....What's it all about ?? Apart from loud music, I drive and fire steamtrains.. those big ones.......believe me the carriages are guaranteed full and the turn-on from everyone is fantastic.....just as those lured to vintage hi-fi. It's the smell of, old noise, dust and heat, just as hot oil and coal.
I gave up SE 35 years ago, despite raucous disapproval from others....When I listen to brass band, SE simply can't deliver the gut punch you wanna feel.
---------------------------------------------------------------
It's not suprising that we are dealing with the least efficient things that mankind has ever made, but the pleasure in hearing nostalgia has never been it's greatest.
richj
I gave up SE 35 years ago, despite raucous disapproval from others....When I listen to brass band, SE simply can't deliver the gut punch you wanna feel.
---------------------------------------------------------------
It's not suprising that we are dealing with the least efficient things that mankind has ever made, but the pleasure in hearing nostalgia has never been it's greatest.
richj
That's an interesting point richwalters, musical preference and listening habits have a large influence on perspective. My current listen is 2 watts into concentric Tannoys and I've confirmed with a dual trace scope monitoring input stage grid vs. cathode voltage that clipping wouldn't fly with the neighbours for long. Most of my current music is solo guitar, piano and the like so maybe, for example, my particular set of trade offs don't weigh bottom end performance and high volumes as heavily. As some wise man around here keeps saying: "horses for courses."
Somewhat off-topic but ironic you should mention vintage hi-fi. While grabbing lunch I scooped the January issue of The Absoulute Sound, the one with the rave review of an amp so clean it "made it hard for me to believe I was listening to electronic music." The MC275 Series IV re-issue.
Somewhat off-topic but ironic you should mention vintage hi-fi. While grabbing lunch I scooped the January issue of The Absoulute Sound, the one with the rave review of an amp so clean it "made it hard for me to believe I was listening to electronic music." The MC275 Series IV re-issue.
Like in all tube vs. solid state discussions I have ever been witness to, it sounds like there are two camps here: those who want amps and speakers to add to the music and become an additional "instrument" by selectively adding distortion to the music played and those who want amps and speakers to be as neutral as possible and accurately render what is on the recording--relying on the recording only to generate sound.
I guess that begs two questions: what is your intent and what sounds good to you? If your intent is to accurately reproduce music (or if you need a ton of power!), don’t use tubes. If this is not your intent and you want to add an additional effect that you think makes your music sound better, use the tubes if you like them.
I personally like for my system to render recordings as accurately as possible, leaving the addition of effects, distortion or “ambiance” to the producers or concert hall. However, if selectively adding in distortion makes the music “sound” better to somebody, I see no reason to argue over whether it is good or bad, just as long as we agree that it is not accurate. The same can be argued for “bass boost”, “tone” or “loudness” selectors—it is not usually accurate to the recorded material, but many people find it pleasant. I’m not sure why “tube effect” should be thought of as any different.
Lots of recording studios add reverb and other effects to tracks to make them "sound" better. Concert halls have their own characteristic "sounds" that are tuned based on size, shape and decor. And yes, electric guitars can add “grunge” effects, all in the name of “sounding better.” You can measure all of these effects quite easily, and some people find them pleasing and some people don’t! Most of the electric guitar, bass, and keyboard amps that I have really liked have been tube based. (Not a musician myself, but I always notice gear when I listen to shows ‘cause I’m weird like that!)
I’m not adding anything scientific to this discussion, but in the course of reading through this thread, I couldn’t help but notice that no one had made this basic observation: If you like the distortion, use the tubes and if you don’t, don’t!
Now that this issue is settled I am sure no one will ever bring this topic up ever, ever again!
I guess that begs two questions: what is your intent and what sounds good to you? If your intent is to accurately reproduce music (or if you need a ton of power!), don’t use tubes. If this is not your intent and you want to add an additional effect that you think makes your music sound better, use the tubes if you like them.
I personally like for my system to render recordings as accurately as possible, leaving the addition of effects, distortion or “ambiance” to the producers or concert hall. However, if selectively adding in distortion makes the music “sound” better to somebody, I see no reason to argue over whether it is good or bad, just as long as we agree that it is not accurate. The same can be argued for “bass boost”, “tone” or “loudness” selectors—it is not usually accurate to the recorded material, but many people find it pleasant. I’m not sure why “tube effect” should be thought of as any different.
Lots of recording studios add reverb and other effects to tracks to make them "sound" better. Concert halls have their own characteristic "sounds" that are tuned based on size, shape and decor. And yes, electric guitars can add “grunge” effects, all in the name of “sounding better.” You can measure all of these effects quite easily, and some people find them pleasing and some people don’t! Most of the electric guitar, bass, and keyboard amps that I have really liked have been tube based. (Not a musician myself, but I always notice gear when I listen to shows ‘cause I’m weird like that!)
I’m not adding anything scientific to this discussion, but in the course of reading through this thread, I couldn’t help but notice that no one had made this basic observation: If you like the distortion, use the tubes and if you don’t, don’t!
Now that this issue is settled I am sure no one will ever bring this topic up ever, ever again!
That brings me to the next point. In the Netherlands there is a saying: "Meten is Weten" it rhymes in dutch...making it even more fun to use
...and it means.."To Measure is to Know" The statement is generally true.Except in audio, and the reason for this escapes me. Or at least boggles my mind. As far as I know the one figure that everyone uses for distortion is THD. Maybe it's the marketing departments. Because I don't often see engineers sparring with specs (in fact hardly ever).
And I think it is a shame that an organization like the AES can't come up with something better!
AES stands for Audio Engineering Society and I quote from their website : http://www.aes.org/info/
Since it can be scientifically proven that THD has no correlation with how an amp sounds. I reject the AES's reality and substitute my own.
Or actually it is Russel O Hamm's reality... which says that higher order distortion is more objectionable than lower order distortion. In other words the entire spectrum of distortion should be viewed. And when you do this you'd have to accept this statement as the truth. If you like the distortion, use the transistors and if you don’t, don’t! 
Hi,
That's just one big assumption.
The question remains however to what extent the measured distortion relates to how we perceive sound as being pleasant or unpleasant.
I'd even forward that amps that measure perfectly well to any standard do not necessarily sound good.
Distortion is a deviation from the original signal so ideally an amplifier should only amplify this signal and present it to a speaker so this speaker can translate it into sound.
Unless things have evolved at warp speed lately, which I doubt, I still have to hear the first system that would render a piano recording so faithfully that it would make me believe it's the real instrument being played.
Regardless of the amps' specs or the speaker used I've never been fooled. Have you?
So, are we measuring the wrong thing, are our measurements incomplete or are we too darn clever to be fooled yet desperately trying to?
Food for thought perhaps,
That's just one big assumption.
The question remains however to what extent the measured distortion relates to how we perceive sound as being pleasant or unpleasant.
I'd even forward that amps that measure perfectly well to any standard do not necessarily sound good.
Distortion is a deviation from the original signal so ideally an amplifier should only amplify this signal and present it to a speaker so this speaker can translate it into sound.
Unless things have evolved at warp speed lately, which I doubt, I still have to hear the first system that would render a piano recording so faithfully that it would make me believe it's the real instrument being played.
Regardless of the amps' specs or the speaker used I've never been fooled. Have you?
So, are we measuring the wrong thing, are our measurements incomplete or are we too darn clever to be fooled yet desperately trying to?
Food for thought perhaps,
Bas Horneman said:
I think it's more accurate to say it's a meaningless, or as normally applied meaninglessly broad, concept. Re: the harmonic spectrum of the two amps I posted earlier, by standard practices of the last few decades the Mullard should perform 'over three times better', or at the very least be less at risk of displaying a sonic character. It doesn't sound that way to me. A single numerical representation of THD alone is a bit like using skidpad G-force as the defining characteristic of a car's handling. It only captures part of the picture.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.