DIY Audio Projects Forum

I don't believe gofar99 was judging amps based on the measured THD rather than by listening. He was merely making some relative comparisons of distortion levels IMO.

There is more than one type of distortion as you point out. Some are almost inaudible at relatively high levels, some sound horrible at low levels.

OVERALL HD means very little UNLESS IT’S EXTREMELY LOW. If you can hear distortion in a SS amp with a TOTAL HD of less than 0.05% but you can’t hear a triode distort at 5% then your listening experience is 180 degrees the polar opposite of mine. I won't use a triode alone due to the excessive distortion unless I WANT TO HEAR DISTORTION. My 300B amp distorts. It's difficult to hear even at high levels for two reasons: The distortion is low order and the distortion harmonics are even. It’s almost an octave generator, with the 2nd harmonic dominating the others. And an octave is consonant and pleasant to listen to.

Compare that with intermodulation distortion where the sum and difference products are completely unrelated musically to the program material. IMD sounds awful. Low order even harmonic distortion can sound good or even be inaudible at levels where the same amount of IMD would sonically destroy the music. You might even think it's part of the recording unless you did an A/B test with a SS amp with very low distortion or even an oddwatt.

If I use a triode in an amp it’s only going to be the power stage. Triode preamps distort too much. If I use a triode in a preamp or driver it will be an SRPP configuration to minimize the distortion. No common cathode single triode stage will ever find its way into an amp of mine unless it’s a SE output tube which is SUPPOSED to distort IMHO.

Also IMHO QUALITY SS amps that are well designed can reproduce music with an accuracy and a total distortion level so low that there is essentially no audible coloration or distortion. However, this doesn’t necessarily sound better than 2nd order triode distortion thrown into the mix, or even a push-pull amp with much higher distortion than the SS. Just because you prefer the pleasant distortion produced by triodes doesn’t mean the SS amp is inferior. The SS amp is actually more accurately reproducing the input signal. I don’t prefer it over the alteration of the sound by a tube amp. Once you get used to good tube amps, quality SS sounds “lifeless”.

You just don’t like a totally clean, “sterile” sound, and that IS personal preference. That’s fine, but I simply think you’re mistaken you say you (or anyone) can hear the distortion at 0.05% in a SS amp but can’t hear a triode distorting at 5%. Sorry, that’s just not my experience at all. I CAN hear the triode distort, although at low distortion levels it sounds like part of the music (and I like it). Especially on strings like violins and cellos – it sounds almost like more players have been added. Switching to a clean SS amp makes the same music sound less “life like” even though technically it’s more accurate.

Put on some classical music and crank up a 300B SE amp and you’ll appreciate what I’m talking about. I’ve seen the term “euphonic distortion” coined and that’s precisely what my 300B delivers. I didn’t build it for low distortion – the TOTAL distortion is high. But the TYPE of distortion is very pleasant compared to any other that I’m aware of, and sometimes I like a little distortion.

I’d also add I play the guitar and use a heavily distorted sound produced mostly by overdriven 12AX7 triodes. The distortion level is probably too high to measure, but those overdriven triodes sound SOOO good. I’ve tried MANY SS guitar amps that were supposed to sound like tube amps but NONE of them ever have. There’s at least one DSP digital modeler that I know of that will reproduce tube distortion very accurately (check out the AXE-FX Ultra if you play the guitar – it will blow your mind) but NOT ONE SS analog circuit that will emulate tubes convincingly – not one.

I agree with a lot of what you say, I just disagree that SS amps sound bad because of the way they distort. I think they sound good, but they sound “sterile” precisely because they DON’T distort, at least not enough to be audible to the vast majority of people. And let’s be clear, I’m speaking of QUALITY SS amps; professional studio monitoring amps – not home theatre systems or cheap receiver/amplifiers.

The only thing I’d add about SS amps is that they can not under ANY circumstances be allowed to clip. If the distortion at 100 watts is 0.01% the distortion at 110 watts could very well be 20% with predominantly high order (high frequency) odd harmonics that might quite possibly destroy your tweeters. At the very least it will sound horrible. Even transient peaks MUST be below the clip point of the amp. SS amps need to be oversized to work properly and the user must take care not to exceed the thermal rating of the speakers and burn them up. Most people (and even most instruction manuals!!!) tell you the "red clip indicator should only briefly light up occasionally". WHAT??????????????!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! If that's how you run a SS amp then expect it to sound like crap.

Tube amps with low NFB on the other hand don’t clip. They get more and more non-linear as they exceed their power rating and they EVENTUALLY “flatline”, but the onset of distortion is much more gradual. That’s one reason you can’t hear a triode distort as easily as a SS amp. It’s a gradual process and it doesn’t produce a bunch of “trash” coming out of your tweeters. It also makes it more difficult to exceed the thermal (and momentary excursion) limitations of your speakers if the amp is sized properly.

Just my 2 cents.

_________________
Stephen

HI.
Please substantiate this claim scientifically. Are you telling us pentodes generates less distortion than triodes assuming both tubes are operating normally?

I've already repeatedly stated:- the transfer curve of a triode is much much more linear than any P/N junction of a sand device or a pentode (which both exhibit simiilar knee-shaped not-so-linear transfer curves). This is strictly scientific.

That's why I use power pentodes only triode-strapped to please my picky ears.

I've auditioned many 300Bs power amps, including very hi-end brandname 300B PP, & 300B PSE (Audio Note Kegon, all triodes in-out, tagged Stg 85,000 = USD130,000), my impression was always very positive. No sign of distortion at any room listening SPLs with different hi-end brandname loudspeakers.

My comment to your claim is the 300B amp you found distorted could be defective.

Sorry, I've never had such aural impression with any 300B power amps I ever auditioned.

Sorry, I totally disagree.

FYI, my DIY phonostages are all triodes & sound nice to my very picky ears & to whoever ever auditioned them - transparent, fast, airy, precise imaging & huge lifelike soundstaging.

FYI, the most costly brandname phono-preamps & power amps, like Audio Note, are all-triode built.

As I already stated, I never dispute SS amps can be immaculately design/built. BUT, but their distortion spectrums do not fit in our ears non-linear auditory masking envelop like a properly disign/built triode amp does. So our brain perceives better sound from a triode amp than any SS amp irrespective how excellent the SS amp can be built.

Nope. I like clean but not "sterile" sound like many SS amps. My tube amps are all very clean but musical involving sound which few SS amp can come close if any at all. I bet every music lovers like clean and muscial sound, which is personal preference indeed, of many many many.

Most most SS power amp get class B PP which involves cross-over distortion. Many many qualtiy tube amps are class A SE / PP or class AB PP which get much more ear forgving sonic
effect.

c-J

PS: I appreciate you need SS gears for yr recording job. But here we are dealing with home audios.

_________________
"If the facts don't fit the theory, change the facts" quoted Albert Einstein.

Most tube amps I've seen, even very expensive ones, have very high distortion levels, often greater than 1%. You may not percieve the distortion as distortion, but if you claim to hear no distortion when in fact a test instrument clearly shows it's there then you've just admitted that you are unable to hear distortion.

As far as the 300B, ANY 300B amp will distort AUDIBLY if it's driven too hard. My 300B is designed such that the power tubes will audibly distort before the driver, so almost all distortion, audible as distortion or not, is produced by the 300B tubes.

As far as a triode, I said I'd never use a self biasing triode alone by itself in an amp unless it was an SE amp using an output power triode. In an SRPP or with an active load absolutely, but not alone by itself.

I would suggest you find a 300B amp that you believe sounds good and turn up the drive level if you want to hear it distort. Additionally I would suggest that you find a 300B amp that you believe sounds "clean" with no distortion, then put a distortion analyzer on the amp. You'll discover that the amp is indeed distorting - you just can't hear it. But don't feel bad, neither can most other people. But you're kiding yourself if you think SE power triode amps aren't producing distortion. Test equipment shows otherwise. My ears can percieve it now that I know what to listen for, and A/B comparisons make it clear that it definitely sounds "different", especially when driven to the edge of what would be perceived as audible distortion by the ear.

As far as SS amps, THEY SHOULD NEVER BE DRIVEN TO DISTORTION. If you can hear 0.01% distortion when the amp is operated properly then you sir have much better ears than mine. The "distortion spectrum" of quality SS amps is IRRELEVANT unless you're overdrivng the amp into clipping, in which case you're not using the amp properly.

I totally agree with you that triode distortion is in general "inaudible" as distortion per se, but the distortion is there and can be measured. If you want an amp with very low distortion figures (like the oddwatt) you can't use a simple single triode stage as a preamp or driver. An SRPP stage is fine, and I'm sure strapped pentodes work very well with very low distortion if used with an active load - I never said I had a problem with tubes with active loads to reduce the distortion. But using a single pentode strapped as a triode and self biased with a cathode resisitor will probably introduce in excess of 1% distortion into the signal. I didn't say you could hear it, but it can be measured. I also didn't say it sounds bad. I think triodes sound great, but if I want to build a tube amp with very low distortion figures I can't use a simple single self biased triode in the signal path.

If I had to guess I'd say my 300B is producing distortion levels up to 5% before it becomes audible as distortion. Low order even harmonic distortion doesn't sound like what most people would consider distortion. It sounds "clean" even though the distortion levels are very high realative to SS amps and PP tube amps that are designed for very low distortion like the oddwatt. A typical SS amp driven to 5% THD would sound HORRIBLE, but that would be serious misuse of the amplifer.

Get some test equipment so you can measure what you're actually listening to. You'll be amazed. Triodes produce distortion that's difficult to perceive unless they're driven very hard, and the distortion they produce is pleasnat to the ear in small quantities. I think we agree on that point. Maybe not.

_________________
Stephen

Hi.

What "test equipment" available to us today can tell what we're "actually listening to" ?????
None, bud. Not in the foreable futrue, IMO as we don't even know what data we are going to measure to tell us what we actually listen.

THD? Forget about it.

The only thing measurable todate that can give us some idea of how we are listening is the distortion spectrum analysis displaying all orders of distortion. Yet it is so remotely related to what or how we percieve in our brain that it is still a guess rather than some quantitative analysis.

As I said better looking of the harmonic distortion display does not mean our ears will listen the same way as our ears don't work the way how an amp is measured.

Yes, it is a gauge to how well design/built is an audio amp from engineering viewpoint. But sonically,
these measurements are helpless.

Our ears are still the best gauging instrument.

c-J

_________________
"If the facts don't fit the theory, change the facts" quoted Albert Einstein.

Trust your ears. If they fool you you may never know.

_________________
Projects: "Noir” - Psvane tube preamp | ”fiih-300” - 300W Holton based amp | "Sortilege" - 180W Tripath Class D Mundorf upgrade amp | “Mantra” - tube MM phono preamp |

Wow, this could be a new thread. My comments were sort of in line with much of this tread. We tend to place too much reliance on numbers and not enough (IMHO) on our ears. Second that with newer parts and designs we have the ability to do better than the pioneers in the field were able to do in days long ago. Cool discussion though.

On a more in topic note. I now have about 200 hours on the KT120s and they are still in balance and still sound fine. My initial thoughts still hold - they sound more powerful than the KT88s and seem to have more inner detail than most tubes I have tried in the amps. For folks who love a warmer sound, I still recommend the JJ Blue Glass KT88s and EL34s.

Good listening
Bruce

_________________
Some of my DIY Tube Amplifier Projects:

• PoddWatt: 5751 / EL84 Push-Pull DIY Tube Amplifier Project
• OddBlocks: KT88 Push-Pull DIY Tube Amplifiers (Mono Blocks)
• ForeWatt: ECC802S (12AU7 / ECC82) DIY Tube Preamplifier Project
• GrooveWatt: Tube (Valve) RIAA Phono Preamplifier Project

Bruce: I agree all the way with the comments about measurements and music reproduction. Is it just us old farts that agree here or are the young picking this up? I think the young are. There will always be the ruler and guage faternity from which nothing can err. Not saying its wrong or right, its just another way of justifying what we do and why we do it.

Agree also with your comments with relation to KT120s etc. I still have the Merlot Black top which will run EL34 (Black Sables) JJ KT88s etc and for that reason I may never part with it. As good and all as the KT120s are there is still room in my room, head and hart for the extended highs and brigher mids of the 88s and the all over delicate presentation of an EL34 spin-off.

This is what make this hobby so good. Viva la difference!

_________________
Projects: "Noir” - Psvane tube preamp | ”fiih-300” - 300W Holton based amp | "Sortilege" - 180W Tripath Class D Mundorf upgrade amp | “Mantra” - tube MM phono preamp |

My main point was that if you want an amplifier that will have a measurable THD of less than 1% you can't use a simple common cathode self biased triode in the signal path. I didn't say triodes don't sound good or that they would produce a sound that wasn't "clean" to the ear. My definition of distortion is ANY deviation from the original signal. Some people's definition of distortion is something that is objectionable to listen to. That in my opinion is an incorrect and incomplete definition. There are so many different ways a signal can be distorted that the word "distortion" is almost meaningless. More specific terms must be used to have a coherent discussion on the subject.

My second main point was that the total %THD is NOT RELATED to how good the amp sounds, but if a test instrument measures 5% THD, the sound IS being colored. It's possible that it will sound FANTASTIC even at those levels if it's the "right" kind of distortion, and it can sound perfectly clean to the ear at several percent.

I also believe that a lot of useful information can be obtained with test instruments. They CERTAINLY don't tell all, and your ears should DEFINITLEY be your final guide. Agreed.

But find me just ONE amp that shows high levels (several percent) of IMD and high order odd harmonic distortion (when measured with a test instument) that sounds good and I'll change my mind about that.

Why cheap-jack can't get a 300B amp to distort audibly is beyond my comprehension. DRIVE THE AMP WITH A STRONGER SIGNAL. I give up on that one. cheap-jack is right - 300B amps simply won't audibly distort!!!!! Mine amp MUST BE DEFECTIVE WHEN I TRY TO GET 30 WATTS OUT OF EACH CHANNEL AND I CAN HEAR DISTORTION. Maybe I can bring it in to a shop to have it fixed since beyond 8 or 10 watts it starts to audibly distort. It must be broken.

And AGAIN my comments have been taken out of context and their meaning distorted (no pun intended).

_________________
Stephen

Hi,

I have the impression that those ripples are spurious in nature and probably generated by the measurement itself.

I'm not familiar with the software/hardware you use for producing the data but I guess there must be a periodic frequency sweep involved. Such sweep could explain the features you see in the transfer function.

Have you tried to inject pure tones at the "bump" frequencies?
How does the phase look like?

Hope it helps.
Cheers!

--Rodrigo.

Could be. Not a bad idea to try a few pure tones. I'll give it a try. However, my other tube amp (300B SE) exhibits no such ripples, which leads me to believe it's the amp. The software uses a quick tone sweep from 20 to 20K with a duration of a couple of seconds to measure the response.

_________________
Stephen