Whatever Joe, I'm afraid your response was all too predictable. Besides what you and I are discussing(?) is off topic in this thread. I and others have shown that your concern over current distortion at the amplifier output is misplaced. If and when you provide some evidence I will happily look at it with an open mind, that's all.
Alas, Pavel did some good work there, but some of his conclusions are wrong. I have tried to OPEN a conversation with him, but he shut it down, did not want what was offered, constructive criticism given in the right spirit.
1. The conclusion that current drive can reduce distortions by up to 20dB is correct!
2. The conclusion that current driver can reduce the distortion of the driver by up to 20dB is incorrect.
But it is easy to see why that can cause confusion, the above two statements may seem contradictory until you explain why they are not. "Understanding current is not easy" (quoting Menno Vanderveen) and by extention current drive is not easy to understand.
Here is the explanation, but one that is open to scrutiny and I welcome it.
Under current drive the measured distortion of the driver can be seen and measured to be less than voltage drive. Pavel confirms that. But it is not the driver that is distorting more under voltage drive, it is the amplifier that is.
I know, this may not be easy to understand, in fact it was hard even for me to come to that conclusion, but it is the only one that stacks up.
So you can see, the discussion is not being shut down, even if there have been attempts on blowtorch and elsewhere.
Scott, please see the fuller picture, that an idea or an understanding often needs its moment in time. The interest in understanding better the role of current, and not just current drive, is now higher than it has ever been. Some are getting on the train and want to enjoy the ride, others are stand-offish and ridiculing the travelers getting on the train.
Which of the two groups are you?
I was hoping Pavel would have been a fellow traveler.
Cheers, Joe
Last edited:
It is precisely ON TOPIC!
We are discussing the sound of amplifiers, right?
Once again, YOU and others are trying to shut down the conversation. Why? Search your soul for the answer to that question. There I cannot help you.
Please, ScottJoplin and Joe Rassmussen, stop already!
The original postings were about “is there a sound specific to tubes” (paraphrased). There have been a lot of answers; some have gone down the track of “properly designed, there shouldn't be”. Others have variously cited the output storage dynamics of a valve-amplifier's transformer, of the subtle feed-back that the OT gives to the driving tubes themselves, and all that. Others still have maintained with dignified insouciance, that amplification of valves intrinsically is different than with op-amps or transistors, because of their markedly different IV transconductance curves.
Yet, it all comes back — again — to the idea that amplifiers' designs exist on a continuüm of design-ideals-and-schooled-criteria. And the continuüm ain't a function of 2 polar ideals either! More like 5 or 6. A many-dimensional continuüm.
EVEN SO, insofar as I can tell, at the topmost level, is the continuüm of “errorless amplification” on one side and “minimalist signal flow” on the other. Because valves for all they ARE are not particularly linear “at the edges” (yea, even the 6SN7, a legendarily linear valve!), over-amplification-and-error-correction-thru-negative-feedback is the generally recognized taming design maneuver.
But just even saying that is an over-simplification, because NFB can be broken down into global NFB and local NFB. It can (less popularly) also be embodied in “limited stage gain” and also “phase-split signal amplification” to induce the concept of unavoidable-but-musically-acceptable distortion symmetrically around the zero of quiescence.
In the end though, “pure vacuum tube amplification” designers come to 'hang their hat' on some small design-criteria area of the continuüm.
I, as a matter of personality, generally prefer more amplification stages, shooting for triode-distortion that is symmetric by an early split in phases, and I'm not particularly religious about having a really good triode or pentode be the front-end first stage.
I'm just as happy with quality JFETs, with well designed modest-gain bipolar long-tailed-pair configurations and any of the other designs that don't employ global negative feedback, but which do deploy local negative feedback as a matter of course. Un-capacitor-bypassed cathode bias resistors, for example. Same for source-side bias resistors for JFET front-ends. Etc., etc. I don't go “crazy” either with crafting rock-solid constant-current or constant-voltage sources. A little rubbery — but well defined — operation seems to hit the sweet spot.
Anyway, again gentlemen, without the rancor please.
⋅-⋅-⋅ Just saying, ⋅-⋅-⋅
⋅-=≡ GoatGuy ✓ ≡=-⋅
The original postings were about “is there a sound specific to tubes” (paraphrased). There have been a lot of answers; some have gone down the track of “properly designed, there shouldn't be”. Others have variously cited the output storage dynamics of a valve-amplifier's transformer, of the subtle feed-back that the OT gives to the driving tubes themselves, and all that. Others still have maintained with dignified insouciance, that amplification of valves intrinsically is different than with op-amps or transistors, because of their markedly different IV transconductance curves.
Yet, it all comes back — again — to the idea that amplifiers' designs exist on a continuüm of design-ideals-and-schooled-criteria. And the continuüm ain't a function of 2 polar ideals either! More like 5 or 6. A many-dimensional continuüm.
EVEN SO, insofar as I can tell, at the topmost level, is the continuüm of “errorless amplification” on one side and “minimalist signal flow” on the other. Because valves for all they ARE are not particularly linear “at the edges” (yea, even the 6SN7, a legendarily linear valve!), over-amplification-and-error-correction-thru-negative-feedback is the generally recognized taming design maneuver.
But just even saying that is an over-simplification, because NFB can be broken down into global NFB and local NFB. It can (less popularly) also be embodied in “limited stage gain” and also “phase-split signal amplification” to induce the concept of unavoidable-but-musically-acceptable distortion symmetrically around the zero of quiescence.
In the end though, “pure vacuum tube amplification” designers come to 'hang their hat' on some small design-criteria area of the continuüm.
I, as a matter of personality, generally prefer more amplification stages, shooting for triode-distortion that is symmetric by an early split in phases, and I'm not particularly religious about having a really good triode or pentode be the front-end first stage.
I'm just as happy with quality JFETs, with well designed modest-gain bipolar long-tailed-pair configurations and any of the other designs that don't employ global negative feedback, but which do deploy local negative feedback as a matter of course. Un-capacitor-bypassed cathode bias resistors, for example. Same for source-side bias resistors for JFET front-ends. Etc., etc. I don't go “crazy” either with crafting rock-solid constant-current or constant-voltage sources. A little rubbery — but well defined — operation seems to hit the sweet spot.
Anyway, again gentlemen, without the rancor please.
⋅-⋅-⋅ Just saying, ⋅-⋅-⋅
⋅-=≡ GoatGuy ✓ ≡=-⋅
Based on answers here + my own experiences I've determined the probability that tubes themselves do not have a sound. Even if we are talking about the I/V curves, I can make a solid state amp with identical curves to a triode, you would not be able to tell the difference by looking at it.
I was gunna make a plate driven super triode amp with variable current sharing (and therefore variable output impedance) for a friend but I don't see the point at all if the triode does not do anything unique to the sound.
I never liked that method of current amp design, I think it's lazy and will provide sub-par output impedance and unwanted DC offsets among other things.
I was gunna make a plate driven super triode amp with variable current sharing (and therefore variable output impedance) for a friend but I don't see the point at all if the triode does not do anything unique to the sound.
I never liked that method of current amp design, I think it's lazy and will provide sub-par output impedance and unwanted DC offsets among other things.
Last edited:
triodes have nfb built in, that is why they can be made without the need for gnfb..
i built a triode connected KT88 pp amp, 8 output tubes per channel, based on the carver silver 7 design, while i provided for gnfb, after biasing the tubes and listening to the amp on speakers, i found out that there really is no need to connect the gnfb....that was about 7 years ago and the amp still lives...
YES exactly the reason why you should never ever take stuff at face value...bring in the doubts and show the conclusions to be complete unprofessional rubbish.
Popular science is full of it, - even more so our friends from Eastern Europe once they get going....which is why serious organisations like IRCAM,CNRS, Bologna universities exist.
Why?
Agenda based science will prove whatever you like,and be wrong nearly all the time.
I am really tired of it, it's same as AGW, - not happening but blown out of all proportions by the mass media, based on massive unscientific con jobs,
VALIDATION is not a swear word!
If you are going to do science use Bruel and Kjaer or DPA, use proper sources, do proper IMD measurements....
What does our czech friend do.....use a pathetic cheap noisy Behringer.
In short everything EXCEPT valid science.
Hey well done!
AND
AND
AND
24 pages of it:-
mmmmm nice!
Statistics do not apply.
This is not statistics, ie a very educated guess from a much smaller sample, it is the real world, and objective measures show that they do exist. There may not be many examples, but some show that they do exist, contrary to your statement. And there is nothing stopping there from being more, and maybe there will with greater level of understanding how you get around the shortcomings of current drive and near current drive can be overcome.
And the huge popularity of amps like the ACA means there are many who need to pay attention.
As an analogy, if i was in the early 1900s, the same statistical argument could be applied to cars… and biy would that argument have been shown wrong.
Then why are you here? Unless you are interested in playing with or have such a high impedance amplifier, this thread is off-topic for you.
dave
I have the book and read it twice. He does not directly say that. He correctly states that finding a suitable woofer for current drive is almost impossible unless you get a batch made. It can be done. It has been done and there is no reason it can’t be done.
For now we have to use tricks like Joe’s (which i am not particualrily enamoured of) or a low Qms driver has to be put into a box that heavily damps the driver further. This too can be done.
Given i am a big fan of biamped WAWs using a voltage amp on the bottom and whatever i want on the top (impedance is flat) is an easy way out.
dave
Some context would help, I'd be interested in the stated reasons it isn't 'best'?
You are only changing the amplifier, so that points to the changed component, but it is a system so hard to untangle the 2..
I’ll say. When i 1st wa stwigged to the possibilities (bty the WE old timer engineer) it took about a decade & a half for things to chrysalis, but my understanding is still improving as we see more discussion and more work.
dave
And are we not discussing on eof the possible meachanism that affexts how they sound?
dave
- Status
- Not open for further replies.