Yes, I think I gave in detail how the position of mic make a day and night difference.
Nobody succeed in perfect recording. I can record very good sounding music, just not the same as the real thing. Anyone that experienced in recording should know that.
Bottom, I think quite a few people "listen" to the music through their spectrum analyzer rather than by the ears.
I am absolutely surprised people still have not bring the speakers into the picture in this thread other than me. It is the single most distorted component in the system. It is the speaker that define the sound of the system. I am not saying the amp is not important, but amp can only bring the best out of the speaker. All the people here debating on this, but they likely do not even have the same speaker.
Case in point, go to the store, stay with the same amp and switch speakers, you hear a completely different sound. But if you stay with one particular pair of speaker and switch amps, you can hear one better than the other, but they pretty much sounds similar. You really have to nail down your pair of speaker before talking about the amp.
Last edited:
Guys, while we are on this subject. I am looking for an amp that sounds warmer, bigger and more musical. What amp I should study more?
I too like MiiB, I have separate OPS pcb from IPS/VAS, I can easily switch the IPS/VAS boards. I even intend to stack a few of the different IPS/VAS and use a rotator switch to switch them around to listen.
I designed my preamp to be able to switch the output to drive two different power amps and a relay signal to switch the speaker/dummy load between two power amp to drive one pair of speaker using auto relays. I want to have instantaneous switching of two different power amps to one pair of speaker for comparison.
I too like MiiB, I have separate OPS pcb from IPS/VAS, I can easily switch the IPS/VAS boards. I even intend to stack a few of the different IPS/VAS and use a rotator switch to switch them around to listen.
I designed my preamp to be able to switch the output to drive two different power amps and a relay signal to switch the speaker/dummy load between two power amp to drive one pair of speaker using auto relays. I want to have instantaneous switching of two different power amps to one pair of speaker for comparison.
Last edited:
I think it's not going anywhere by comparing playing back recorded sound with live music sound, as we clearly know that Hi-Fi stops at the record producer's monitor room. One can only get as close as possible to what the producer had heard in his pair of monitors at best and nothing further. We perhaps should leave the live music sound out of power amp discussion as it is not what Hi-Fi is after, therefore, it cannot serve as a reference in the discussion.
True, just want to point out to the people that insist on true fidelity that there is nothing fidelity about it.
So we don't know, all that's left are the numbers... All of a sudden I understand why nothing is moving.
It's my impression that, for tubes at least, a persuasive theoretical argument can be made for a touch of GNFB, but not for SET where the problem of not using GNFB was more technology related. The technology now exists to avoid GNFB in SETs without raising distortion.
That roughly translates as I want an amplifier with distortion, lots more disortion.
Now days you can attend a prom and listen to it again as soon as you get home 🙂. It's odd to hear yourself cough!
I do promenade, so do not expect the presentation delivered by the OB unit to sound the same. But not sure what Alan has against violins.
Yes, I did start this in a light hearted way. GNFB or zero FB, nothing more. I seek calm discussion and open minds.
We are talking about subjectives here, because THD is such a woeful numerical appraisal system because it ignores psychoacoustic issues which I believe are pivotal with SS amps. If something cannot be measured, such as image depth or 'engagement', it means that we have not figured out HOW to measure it, rather than the quality does not exist.
For the record, I have no particular issue either way. If there were a predictive way to design an amp that incorporated the best qualities of a SS and a tube amp, then the parameters should be available, but to my mind nothing is certain after all these decades.
Perhaps THD is too coarse. Maybe a forensic analysis of the distortion profile might be useful, and smart guys like Jean Hiraga suggested a linear decreasing profile is better. This is not quite what MiiB suggests, but it's maybe better than nothing.........
Hugh
We are talking about subjectives here, because THD is such a woeful numerical appraisal system because it ignores psychoacoustic issues which I believe are pivotal with SS amps. If something cannot be measured, such as image depth or 'engagement', it means that we have not figured out HOW to measure it, rather than the quality does not exist.
For the record, I have no particular issue either way. If there were a predictive way to design an amp that incorporated the best qualities of a SS and a tube amp, then the parameters should be available, but to my mind nothing is certain after all these decades.
Perhaps THD is too coarse. Maybe a forensic analysis of the distortion profile might be useful, and smart guys like Jean Hiraga suggested a linear decreasing profile is better. This is not quite what MiiB suggests, but it's maybe better than nothing.........
Hugh
My apologies for the misrepresentation.
Can we say you believe moderate amounts of feedback are preferable to no feedback, in contrast to Putzeys who has said if you use feedback, more is better?
That's OK. Gives me an excuse to toot my horn 😀
You can say that, I think. To me moderate amounts of feedback would be
something like 20 dB or so. It does presume that the circuit is fairly linear
beforehand.
If you build Class D amplifiers, it seems you need lots of feedback, at least
that's what I discovered when I built this one back in the early 80's.
Attachments
I am more looking for what brand of amps fit the bill so I can study their schematic.
I just don't like orchestra music, don't like classical music.
Keep tooting, love to hear from one that is known to produce the tip top designs. Your words are the golden words.
I almost going to bid on ebay one of the Threshold amp to use it as the standard to judge my designs.
GNFB isn't the only, isolated, method of linearising an amplifier; a lot has to do with the choice of individual stages and how to make them as linear as practical in order to avoid using a high degree of GNFB which can bring in a new set of problems if applied inappropriately. There are known benefits, and traps, in this design approach.
Marshal Leach used cascoding in the input stage of his low-TIM amplifier (and in the VAS on his high power amplifier), and didn't use active current sources and mirrors; he liked the sound of the amplifier without the latter technique. He did use a little feedforward, but I think this was for high frequency stability, and the commonly used emitter degeneration in the diff amp input, which amounts to localized feedback.
A practical Class AB output stage can't be completely cleaned up in the Lin topology partly because of stability issues. Error correction, which seems to be a nested feedback/feedforward? loop, is used in the Benchmark AHB2 amplifier to achieve extremely low distortion at all power levels and high frequencies. It's also a very energy efficient design; I think the output stage is really Class B and it uses an active switching power supply. I've only read one review of this design, so I don't know how generally well received it is.
I'm studying CFA design as well; it doesn't handle feedback in the same ways as VFA designs. I doubt CFA will displace VFA by any means, it's just a different approach to design that need careful study of how to employ it.
Class D is yet another approach, but has a plethora of 'gotchas' and I haven't had any interest in it.
Marshal Leach used cascoding in the input stage of his low-TIM amplifier (and in the VAS on his high power amplifier), and didn't use active current sources and mirrors; he liked the sound of the amplifier without the latter technique. He did use a little feedforward, but I think this was for high frequency stability, and the commonly used emitter degeneration in the diff amp input, which amounts to localized feedback.
A practical Class AB output stage can't be completely cleaned up in the Lin topology partly because of stability issues. Error correction, which seems to be a nested feedback/feedforward? loop, is used in the Benchmark AHB2 amplifier to achieve extremely low distortion at all power levels and high frequencies. It's also a very energy efficient design; I think the output stage is really Class B and it uses an active switching power supply. I've only read one review of this design, so I don't know how generally well received it is.
I'm studying CFA design as well; it doesn't handle feedback in the same ways as VFA designs. I doubt CFA will displace VFA by any means, it's just a different approach to design that need careful study of how to employ it.
Class D is yet another approach, but has a plethora of 'gotchas' and I haven't had any interest in it.
It seems fairly standard to criticize the use of feedback by bringing up the ultra low distortion,high feedback, poor sounding amps produced during the "distortion wars"...without mention that these amps were not necessarily very linear to start with....
Seems class d has come a long way since the 80's.....
well as there is no IEEE standard for 'added warmth' quite how high a distortion profile you need is hard to tell. I would start by borrowing a SET amp from someone and see if that is distorted enough or if you want more.
one of the firstwatt designs (forget which) can be tuned for distortion so you can adjust to taste. That might also be a very good starting point if you discover you only want a smidge of extra harmonics.
Fine you don't like classical, but why then use that as an example?