Hi,
Yes, we discussed that the very low SPL of your tests meant that where essentially meaningless, as at such low levels you would not encounter such levels of HD from most amplifiers and because your speakers have (relatively) high distortion even at these low levels which may cause other interactions.
I will repeat, distortion audibility depends on SPL, frequency and precise nature of the non-linearity. I desire only to reduce HD below audibility for the Amplifier when driving an "average" sensitivity (87dB/2.83V/m) Speaker, I see no benefit from going further.
To your ears and requirements it may be so, I do not find agreement from my requirements and ears. I find that minimising looped feedback agrees more with what I desire than maximising it.
Ciao T
Yes, we discussed that the very low SPL of your tests meant that where essentially meaningless, as at such low levels you would not encounter such levels of HD from most amplifiers and because your speakers have (relatively) high distortion even at these low levels which may cause other interactions.
I will repeat, distortion audibility depends on SPL, frequency and precise nature of the non-linearity. I desire only to reduce HD below audibility for the Amplifier when driving an "average" sensitivity (87dB/2.83V/m) Speaker, I see no benefit from going further.
As for sounding better , who knows , but i can assure you that it will not sound worse , for sure....
To your ears and requirements it may be so, I do not find agreement from my requirements and ears. I find that minimising looped feedback agrees more with what I desire than maximising it.
Ciao T
A complete amplifier where only the VAS is modded to see this
stage contribution to total THD.
Output stage is lateral fet class A biaised and all versions
where built using the same input stage and triple output
pair 2SJ48/2SK133.
Surprisingly , reducing degeneration and using the resultant
gain for NFB will yield way lower THD and IMD , that is ,
linearization of the VAS using a said ratio of degeneration
will not be as efficient as using this available NFB in a larger loop ,
global or local , so your quest for better VAS linearity using degeneration
may well be a dead end....
Last edited:
Hi,
But that is the answer to a different question.
I am not sure why this would be surprising.
May I repeat that I do not consider low THD (and related IMD) a valid design goal, beyond a certain point. I appreciate that you do, which is fine as far as it goes.
However i am asking about the taste of mangos and oranges and you keep talking about apple's...
You may notice that I have a local loop AND degeneration for the VAS...
Alas, again you mistake my intentions.
I do not want "better", I want "adequate for my own requirements".
Ciao T
But that is the answer to a different question.
I am not sure why this would be surprising.
May I repeat that I do not consider low THD (and related IMD) a valid design goal, beyond a certain point. I appreciate that you do, which is fine as far as it goes.
However i am asking about the taste of mangos and oranges and you keep talking about apple's...
You may notice that I have a local loop AND degeneration for the VAS...
Alas, again you mistake my intentions.
I do not want "better", I want "adequate for my own requirements".
Ciao T
Hi,
Neither do J-Fets and vertical Fets...
One does have to watch the fet's parasitic capacitances though.
Ciao T
Neither do J-Fets and vertical Fets...
One does have to watch the fet's parasitic capacitances though.
Ciao T
Exactly. There is always a compromise between distortions of unloaded stage VS distortions of the stage loaded on non-linear impedance. Everything reflects, but reflects differently. Even tubes do, especially output tubes that work on edge of possibilities, through grid currents and Miller capacitances.
Last edited:
So let s say that my own requirements mandate parity with CD HD levels
rather than with the ones of 128 Kbits MP3....
dvv: "Some, notably the Brits (Naim, Cyrus, etc), will offer you add-on power supplies which will improve your sound - but by the time you have your end price, you discover you would have been better off buying a say Accuphase and having it all inside just one case. As I see it, they are charging you extra for what should have been there in the first place."
Bingo.
Bingo.
Hi,
Please tell me where you got your speakers with -90dB THD at rated power, 20Hz-20KHz from. I want a pair.
You do have speakers that meet what your requirements mandate, right?
Ciao T
Please tell me where you got your speakers with -90dB THD at rated power, 20Hz-20KHz from. I want a pair.
You do have speakers that meet what your requirements mandate, right?
Ciao T
Just for a note, the sim for the DH-120 which has a cascode VAS, no buffer, but was fed with a diode-bjt ccs, adding the fet cascode ccs gained 10dB in simulated distortion reduction. Much more that the gain in the IPS.
Hi,
Really. You actually measured this? A THD of -90dB (0.003%) at rated power?
Ciao T
I would lend you mine, but you have a small room ....
Really. You actually measured this? A THD of -90dB (0.003%) at rated power?
Ciao T
C'mon T you cant be this easy ...... 
Anyway, we can tolerate higher THD values from speakers than we can from the electronics, same as Tubes vs SS , analog vs Digital . I'm saying we don't need speakers with an .003%thd @ rated power to hear the difference in the amplifier , as described by Wahab, especially an SS one.
Anyway, we can tolerate higher THD values from speakers than we can from the electronics, same as Tubes vs SS , analog vs Digital . I'm saying we don't need speakers with an .003%thd @ rated power to hear the difference in the amplifier , as described by Wahab, especially an SS one.
Speakers apparent average fidelity shouldnt be a cover to justify
poor performing amps.
If such a logic would be applized to other parts of the audio chain
i guess that LPs would be enough , perhaps even high end K7 desks
a la Nakamichi 700/1000 would have been adequate , rendering
digital formats useless and in fact just a commercialy hyped format.
Yet , in listening tests those high distorsion speakers still show
some difference between theses audio formats......
Hi,
No I am brutally serious.
That's rubbish.
It is merely proof that THD is meaningless as measure of quality.
All evidence we have points to the fact that distortion audibility depends on SPL, Frequency and precise transfer function (which is for example resolved by a FFT as a series of harmonics). So THD is completely meaningless.
To say that we can tolerate more THD from Speakers than from electronics is actually saying that two have different transfer functions for their THD and one transfer function is more audible than the other.
If your speaker, solid state and tube electronics had the same precise transfer function they would have the same distortion audibility for harmonic distortion, if the dynamic transfer function was also identical and several others (suspected/guessed/assumed) then we would even expect identical sound.
But that is another story.
So my point stands.
If someone insists on -90dB THD from an amplifier he must also insist on this from the speaker, otherwise it is gulping down the camel while straining out the gnat...
If a Speaker on the other hand is allowed several percent THD at full power, then demanding that an amplifier have 0.003% THD at full power is a classic case of doublethink, of holding two completely opposite and mutually exclusive views (in this case that of "high THD is GOOD" and "High THD is BAD") at the same time and fully believing completely in each one when called upon and never noticing the total conflict.
My design goals are very low order higher harmonics and reasonable (inaudible) levels of 2nd to 5th harmonics at the expected SPL's, as that effects distortion audibility, not low THD as such.
Ciao T
C'mon T you cant be this easy ......
No I am brutally serious.
That's rubbish.
It is merely proof that THD is meaningless as measure of quality.
All evidence we have points to the fact that distortion audibility depends on SPL, Frequency and precise transfer function (which is for example resolved by a FFT as a series of harmonics). So THD is completely meaningless.
To say that we can tolerate more THD from Speakers than from electronics is actually saying that two have different transfer functions for their THD and one transfer function is more audible than the other.
If your speaker, solid state and tube electronics had the same precise transfer function they would have the same distortion audibility for harmonic distortion, if the dynamic transfer function was also identical and several others (suspected/guessed/assumed) then we would even expect identical sound.
But that is another story.
So my point stands.
If someone insists on -90dB THD from an amplifier he must also insist on this from the speaker, otherwise it is gulping down the camel while straining out the gnat...
If a Speaker on the other hand is allowed several percent THD at full power, then demanding that an amplifier have 0.003% THD at full power is a classic case of doublethink, of holding two completely opposite and mutually exclusive views (in this case that of "high THD is GOOD" and "High THD is BAD") at the same time and fully believing completely in each one when called upon and never noticing the total conflict.
My design goals are very low order higher harmonics and reasonable (inaudible) levels of 2nd to 5th harmonics at the expected SPL's, as that effects distortion audibility, not low THD as such.
Ciao T
I see no particular problem (other then still more work) in making a cascode with parallelled transitors.
Nelson Pass did it 20 years ago.
I've been fooling around with a prototype circuit all afternoon. Sheesh, it actually works.
Wahab, your view of THD & IM levels is just fine, go for it. Personally, I have experienced far too many cases in which amps with LOVELY specs sounded poor, and I own an amp with such specs at a level where you would probably demand its quick demise, yet it manages to sound surprisingly good. I refe to Harman /Kardon 6550 integrated amp, which I purchased in December 1993.
It's spec sheet says it has a THD factor of 0.3%. Not a typo, no zeroes lost in translation, really one third of one percent. Then it adds insult to injury and delivers only 50/70 wrms into 8/4 Ohms (I guess Wayne is gathering a posse by now ), but that's all one can REASONABLY expect from a SEPP designs, using a single pair of Toshiba's 2SC3281/2SA1302.
Yet, it manages to convey a lot more "air" and "space" than quite a few nominally "better" amps.
To make matters worse, they use just 17 dB of global NFB, sending the water down to Thorsen's mill.
Its circuit is not unlike Thorsten's, there are similarities, but their input stage is plain, ol' differential pair of plain ol' bipolars - and that's it. Aha-ha, BUT their degeneration on that same input pairs is big - collector resistors 1.2 kOhms, emmitters resistors 220 Ohms. The second stage is EXACTLY like Thorsen's, and the driver stage is a two transistor Darlington circuit, again with local NFB. This drives the said Toshiba transistors.
But it has a drive energy like you wouldn't believe, it's really, really fast. All from a single pair of 15,000 uF capacitors of South Korean origins. No fancy capacitors inside. Or other parts, for that matter. But the engineering is kinda good, it seems. Even if some parts, trimmers and pots, are probably the worst I have ever seen, yet this too is an H/K hallmark, been for years.
Nelson Pass did it 20 years ago.
I've been fooling around with a prototype circuit all afternoon. Sheesh, it actually works.
Wahab, your view of THD & IM levels is just fine, go for it. Personally, I have experienced far too many cases in which amps with LOVELY specs sounded poor, and I own an amp with such specs at a level where you would probably demand its quick demise, yet it manages to sound surprisingly good. I refe to Harman /Kardon 6550 integrated amp, which I purchased in December 1993.
It's spec sheet says it has a THD factor of 0.3%. Not a typo, no zeroes lost in translation, really one third of one percent. Then it adds insult to injury and delivers only 50/70 wrms into 8/4 Ohms (I guess Wayne is gathering a posse by now
), but that's all one can REASONABLY expect from a SEPP designs, using a single pair of Toshiba's 2SC3281/2SA1302.Yet, it manages to convey a lot more "air" and "space" than quite a few nominally "better" amps.
To make matters worse, they use just 17 dB of global NFB, sending the water down to Thorsen's mill.
Its circuit is not unlike Thorsten's, there are similarities, but their input stage is plain, ol' differential pair of plain ol' bipolars - and that's it. Aha-ha, BUT their degeneration on that same input pairs is big - collector resistors 1.2 kOhms, emmitters resistors 220 Ohms. The second stage is EXACTLY like Thorsen's, and the driver stage is a two transistor Darlington circuit, again with local NFB. This drives the said Toshiba transistors.
But it has a drive energy like you wouldn't believe, it's really, really fast. All from a single pair of 15,000 uF capacitors of South Korean origins. No fancy capacitors inside. Or other parts, for that matter. But the engineering is kinda good, it seems. Even if some parts, trimmers and pots, are probably the worst I have ever seen, yet this too is an H/K hallmark, been for years.
Fine. Please explain to me how a single pair of 6,800 uF caps satisfies all the requirements of a nominally 2x50 Wrms into anything lower than 7.9 Ohms. I would really like to know.
Or how a single pair of 12,000 uF caps satisfies 5 x 105 Wrms.
Because that's what you are getting in modern day AV receivers (as if any other type still existed), and sadly, in many wanna-be High Fidelity (not to say High End) products.
While at it, kindly explain to me why would companies like Naim, Cyrus, Linn at al. have made such inroads with add-on power supplies, which turn their products from standard to above standard, mostly by adding banks of capacitors.
I understand a budget product cannot have inside what a High End product does, I understand they have to design them with price in view first, but come on, let's not overdo it. Or we'll sonn be in the "P.M.P.O" territory, you know, when they give you a tiny 12 V transformer and claim "1,200 Watts PMPO" - literally.
Hi,
We are not talking about APPARENT FIDELITY, we are talking about distortion.
How, if very high THD (usually well above -60dB @ 1 Watt input) does not produce a poor performance speaker, can you claim the same distortion produces a poor performing amplifier?
You are applying double standards. This is neither logical nor based in evidence.
Poor performance is poor performance ever which way.
So if -90dB THD is necessary for high fidelity, it is necessary also from speakers, otherwise -90dB THD is not necessary from an Amplifier either.
Simple logic demands this, otherwise all you expressing is an irrational belief, not truth, science or fact, which you are BTW totally welcome to do, as far as I'm concerned.
I respect any-ones delusions, as long as they respect what they think are MY delusions (I consider them evidence based truth, often bleeding self evident, but I do not feel evangelical or inquisitorial about this).
Ciao T
We are not talking about APPARENT FIDELITY, we are talking about distortion.
How, if very high THD (usually well above -60dB @ 1 Watt input) does not produce a poor performance speaker, can you claim the same distortion produces a poor performing amplifier?
You are applying double standards. This is neither logical nor based in evidence.
Poor performance is poor performance ever which way.
So if -90dB THD is necessary for high fidelity, it is necessary also from speakers, otherwise -90dB THD is not necessary from an Amplifier either.
Simple logic demands this, otherwise all you expressing is an irrational belief, not truth, science or fact, which you are BTW totally welcome to do, as far as I'm concerned.
I respect any-ones delusions, as long as they respect what they think are MY delusions (I consider them evidence based truth, often bleeding self evident, but I do not feel evangelical or inquisitorial about this).
Ciao T
Last edited:
Hi,
Make the transformer voltage high enough to serve whatever voltage you need to deliver 50W into 8 Ohm and whatever you want into lower impedances. You could even make the supply voltages switchable.
As long as there is enough voltage to not drop the output stage "out" even lower capacitance works fine.
The Goldmund Mimesis 8 uses a pair of 4700uF capacitors per channel and it delivers 300W into 2 Ohm (185W into 8 Ohm) at 1% THD. There is no trace of increasing LF distortion caused by "too small capacitors".
Goldmund Mimesis 8 power amplifier Measurements | Stereophile.com
So 6,800uF is ample for a 50W Amp, at least if it is only about delivering power...
Ciao T
Make the transformer voltage high enough to serve whatever voltage you need to deliver 50W into 8 Ohm and whatever you want into lower impedances. You could even make the supply voltages switchable.
As long as there is enough voltage to not drop the output stage "out" even lower capacitance works fine.
The Goldmund Mimesis 8 uses a pair of 4700uF capacitors per channel and it delivers 300W into 2 Ohm (185W into 8 Ohm) at 1% THD. There is no trace of increasing LF distortion caused by "too small capacitors".
Goldmund Mimesis 8 power amplifier Measurements | Stereophile.com
So 6,800uF is ample for a 50W Amp, at least if it is only about delivering power...
Ciao T
I am totally convinced that waveform matters, including change of waveform with loudness, decay, reverberation, and so on. The whole waveform of the whole sound. Fourier transform is only an instrument to analyze some types of distortions. It can be completely useful, but in such case all spectrum of signal has to be analyzed, not only harmonics generated from certain level of sinewave signals (well, sinewave in terms of Fourioer analyzis has no other members except fundamental frequency - related only and only whan the signal is infinitely long). Castrated analyzis can give castrated results only. Envelope matters as well. Changes of envelope, intermodulations between everything in the sound picture that change waveform. And scale of changes that matters for hearing them is different from seeing them on oscilloscope screen. When mix of soft and loud sounds, or overtones of sounds, that change in time is heard, everything matters: soft overtones and their change, as well as loud tones and their change, while soft tones and overtones may be less visible on oscillooscope than loud tones.
- Status
- Not open for further replies.