I think it was the Wiki author or other source who was trying to manipulate the results there. You don't find too many audio distortions with H6 @ 5% and H2 only 0.2%.
BTW You'd better get the range back together if you expect supper any time soon
Jan, I have to agree there. It is most difficult to find rationale in these shifting, uncertain and uninformed statements. Syn08 had a couple of acronyms for guys with this problem generally; MAAW comes to mind. 'Cryptic, but the "middle aged" bit get's all of us sometime.
Last edited:
MiiB,
The question is how to accomplish that in power amplifiers, distortions have a strong propensity to rise with frequency, are rarely linear, rather nonlinear symmetrical and nonlinear asymmetrical, comprising a higher level of much more annoying high frequency intermodulation products. Solid state devices (especially bipolars) don`t exhibit a tube-like nonlinearity.
The single-ended topology produces the lowest distortion and the most pleasing sound spectrum. I don`t reject other topologies, I`m using other topologies, but I have to be honest with myself and with others.
Single Ended/Push Pull Amplifier
There are surely better sources of knowledge.
Yes, it´s exactly what I´m saying.
1% THD is preferable over 0.001% THD if the distortion is mainly low order and the high order distortions are small, although it´s a thousand times higher level. The motto is: level is nothing, spectrum is everything.
The question is how to accomplish that in power amplifiers, distortions have a strong propensity to rise with frequency, are rarely linear, rather nonlinear symmetrical and nonlinear asymmetrical, comprising a higher level of much more annoying high frequency intermodulation products. Solid state devices (especially bipolars) don`t exhibit a tube-like nonlinearity.
Hopfully, you will see the point some sunny day.
The single-ended topology produces the lowest distortion and the most pleasing sound spectrum. I don`t reject other topologies, I`m using other topologies, but I have to be honest with myself and with others.
I ignore and avoid tricks, tricks hit back hard. By Ciuffoli`s concept nothing is gained. How could I argue with that man about concepts?
Single Ended/Push Pull Amplifier
There are surely better sources of knowledge.
Then a simple amplifier with these specs (simulated) must be something good...
Harmonic Frequency Fourier Normalized Phase Normalized
Number [Hz] Component Component [degree] Phase [deg]
1 1.000e+03 2.271e+00 1.000e+00 -0.06° 0.00°
2 2.000e+03 2.384e-07 1.049e-07 -35.72° -35.67°
3 3.000e+03 2.562e-08 1.128e-08 22.05° 22.11°
4 4.000e+03 8.751e-11 3.853e-11 -112.44° -112.38°
5 5.000e+03 2.749e-10 1.210e-10 -88.62° -88.57°
6 6.000e+03 1.549e-10 6.820e-11 4.71° 4.76°
7 7.000e+03 2.362e-10 1.040e-10 109.99° 110.04°
8 8.000e+03 1.296e-10 5.706e-11 162.53° 162.58°
9 9.000e+03 1.373e-10 6.044e-11 -8.09° -8.04°
Total Harmonic Distortion: 0.000011%
Harmonic Frequency Fourier Normalized Phase Normalized
Number [Hz] Component Component [degree] Phase [deg]
1 1.000e+03 2.271e+00 1.000e+00 -0.06° 0.00°
2 2.000e+03 2.384e-07 1.049e-07 -35.72° -35.67°
3 3.000e+03 2.562e-08 1.128e-08 22.05° 22.11°
4 4.000e+03 8.751e-11 3.853e-11 -112.44° -112.38°
5 5.000e+03 2.749e-10 1.210e-10 -88.62° -88.57°
6 6.000e+03 1.549e-10 6.820e-11 4.71° 4.76°
7 7.000e+03 2.362e-10 1.040e-10 109.99° 110.04°
8 8.000e+03 1.296e-10 5.706e-11 162.53° 162.58°
9 9.000e+03 1.373e-10 6.044e-11 -8.09° -8.04°
Total Harmonic Distortion: 0.000011%
I have newer read his site so I would not know how good his other concepts are.. And why should I judge this concept by what other idea he has been working with.. bit like looking at the messenger and thus rejecting all. I like the open minded approach where all new ideas could yield something spectacular..
And frankly the differences on the two concepts have been supported by quite a few good people in here..
Member
Joined 2009
Paid Member
I had a quick look at that link - it's good to see somebody having fun with these experiments. I think I've seen some very similar ideas explored in some depth before on the tubes-valves forum.
But I agree with WuYit in that there are subtleties not always appreciated. For example, the idea that a push-pull transformer can be used as a single ended transformer by simply ensuring the dc currents balance out is, I suspect, not a complete solution and we might be fooling ourselves if we thought they were one and the same. One of the key things about a SE output transformer is that it remains constantly magnetized by the standing dc current flowing through it and therefore the signal traverses a different part of the hysteresis curve than would be the case for a push pull output transformer. So in at least this one regard I don't view what Ciuffoli has done here is equivalent to a SE tube amplifier even though it may technically be classed as operating single ended.
But I agree with WuYit in that there are subtleties not always appreciated. For example, the idea that a push-pull transformer can be used as a single ended transformer by simply ensuring the dc currents balance out is, I suspect, not a complete solution and we might be fooling ourselves if we thought they were one and the same. One of the key things about a SE output transformer is that it remains constantly magnetized by the standing dc current flowing through it and therefore the signal traverses a different part of the hysteresis curve than would be the case for a push pull output transformer. So in at least this one regard I don't view what Ciuffoli has done here is equivalent to a SE tube amplifier even though it may technically be classed as operating single ended.
MiiB,
AC means currents and voltages that change direction and polarity, respectively, over time. Any point in a circuit that does not change potential, like the power supply rails, is an AC ground.
Electrons don`t care about pompous concepts and don`t read schematics, they only will return to their source.
AC means currents and voltages that change direction and polarity, respectively, over time. Any point in a circuit that does not change potential, like the power supply rails, is an AC ground.
Electrons don`t care about pompous concepts and don`t read schematics, they only will return to their source.
Thd spectrum can be everything provided the level is high
enough to be heard.
0.001% is 100db BELOW the output level, so in essence, you are
saying that you can hear the noise of a CD while the amp is running
at say a comfortable level...
At such THD ratios, the harmonics are buried in the amp s and source
noises, and can be measured by sophisticated instruments only using
an averaging to extract the said harmonics from the dominating
noise level...
wahab,
I just was using Mike´s figures for demonstration purposes, the 10% : 0.01% ratio would have been more realistic. The message should be clear anyway.
Bonsai,
I just was using Mike´s figures for demonstration purposes, the 10% : 0.01% ratio would have been more realistic. The message should be clear anyway.
That`s an unfortunate formulation. There´s an audibility aspect, but not in the way you are suggesting. The audibility level depends on harmonic order; that is high for low order distortion and low for high order distortion, the latter is always accompanied by a much higher level of nasty intermodulation products. Dynamically, the simulated or measured static distortion will rise dramatically and unpredictably (along with the dynamic noise floor) loosing entirely its already negligible validity. This applies for amplifiers with global feedback.
Bonsai,
I meant something else. Missing odd order harmonics would be equally undesirable. A healthy spectrum and low THD are incompatible.
Hi Bonsai, Wuyit is somewhat an extremist audio amateur.
Wuyit, noise wise there s a particular situation where feedback based
amps produce huge amount of noise, but it s just not important..
Suppose we feed the input with a square signal.
During the rise time, there s no NFB, and the amp
is at his maximal OLG, so the EIN is amplified in a
large magnitude, but then , this rise time being very short,
the equivalent noise density vs time is very low.
Provided the input signal slope doesn t exceed the amp s
maximal slew rate, the NFB will be fully functionnal and the
gain will stay at the closed loop gain value...
As for your exemple of THD , 10% is just unlistenable,
whatever the THD content, and 0.01% , be it with the less
musical harmonics only, is infinitly better.
So you started from a bad exemple and ended with a worse one.
Wuyit, noise wise there s a particular situation where feedback based
amps produce huge amount of noise, but it s just not important..
Suppose we feed the input with a square signal.
During the rise time, there s no NFB, and the amp
is at his maximal OLG, so the EIN is amplified in a
large magnitude, but then , this rise time being very short,
the equivalent noise density vs time is very low.
Provided the input signal slope doesn t exceed the amp s
maximal slew rate, the NFB will be fully functionnal and the
gain will stay at the closed loop gain value...
As for your exemple of THD , 10% is just unlistenable,
whatever the THD content, and 0.01% , be it with the less
musical harmonics only, is infinitly better.
So you started from a bad exemple and ended with a worse one.
Last edited:
It´s not about rise time.There are quite many types of static noise identified, but the dynamic noise is clearly worse as it directly modulates the signal and is raised by the very same feedback mechanism that converts benign low order harmonics into high frequency dirt, resulting in a hard, cold and lifeless sound deprived from its harmonic content, and not a single spectral component being nearly correctly represented in time.
Well, you have apparently made your choice.
For an amplifier... but not for a loudspeaker.
OS
10% THD is virtually a 20db S/N ratio, be it a speaker s THD...
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.