I found that simple sapping one component against another does not always work.
For example Opamp swapping :
A circuit that is stable with an NE5532 may not be stable with an AD797 or other higher speed modern Opamps.
In a harsh RFI environment an OPA134 may work better then an NE5532.
Adjusting both tests ( stabilizing the Ad797 ) and removing the RFI in the second texst may give a null results.
I really do not think that the diffence in distortion is detectable in any case when the gain it not too high provided that the circuit listened too is stable and there is no HF ingress.
There may also be diffences in PSU rejection.
Having fixed that may also help to swap the differences.
When there is a diffence it may come from instability, Hf ingress and noise on the powerline and not minuscule differences in distortion.
For example Opamp swapping :
A circuit that is stable with an NE5532 may not be stable with an AD797 or other higher speed modern Opamps.
In a harsh RFI environment an OPA134 may work better then an NE5532.
Adjusting both tests ( stabilizing the Ad797 ) and removing the RFI in the second texst may give a null results.
I really do not think that the diffence in distortion is detectable in any case when the gain it not too high provided that the circuit listened too is stable and there is no HF ingress.
There may also be diffences in PSU rejection.
Having fixed that may also help to swap the differences.
When there is a diffence it may come from instability, Hf ingress and noise on the powerline and not minuscule differences in distortion.
Last edited:
Somtimes tonal balance is so por on some recordings ( thin, sharp, hollow, nasal etc )
that i whould love to have a good tone control like in the old times but that is a nogo in high end of cause.
I do not think about something drastic. May be a Quad like tilt plus tone controls at 50Hz and 12khz.
that i whould love to have a good tone control like in the old times but that is a nogo in high end of cause.
I do not think about something drastic. May be a Quad like tilt plus tone controls at 50Hz and 12khz.
^
I tried a portable device in a Hi-Fi store recently which I had never heard nor seen before, very little information in advance, just the name of it really.
I listened to it and tried to guess which op-amp it was I was listening to, from the line-out. I was quite confident I was listening to the OPA1612.
If I had to say a name, that was it, little question in my mind.
Later, I searched online via my smartphone, for which op-amp it was in it, it said it was the OPA1612.
Yes, just a little story, but it made me feel more confident now, I plan to try more tests like that.
I tried a portable device in a Hi-Fi store recently which I had never heard nor seen before, very little information in advance, just the name of it really.
I listened to it and tried to guess which op-amp it was I was listening to, from the line-out. I was quite confident I was listening to the OPA1612.
If I had to say a name, that was it, little question in my mind.
Later, I searched online via my smartphone, for which op-amp it was in it, it said it was the OPA1612.
Yes, just a little story, but it made me feel more confident now, I plan to try more tests like that.
Last edited:
For me that would be a wrong turning - that type of recording is a treasure for me, because they expose the remaining weaknesses in the playback system - the amplifier, or system, that is a " wire with gain " is a myth, there is always added distortion. And the ones that nominally are the best performers in low distortion figures can be the worst to listen to, because you can now hear the nastier effects of interference artifacts and suchlike, which are masked in more "mellow" playback environments.
Personally, I have found persisting in cleaning up everything in the electronics area gets one past the dilemma you mention ..
I like tone controls, and based on talk on other boards it seems that people are starting to realize how valuable they can be when properly implemented. I've found one of the largest audiophile contradictions in the 90's was the shun of any audio gear with tone controls "because they added distortion", all the while the same person/people who stated such were into high distortion "audiophile approved" brand gear.
Just a little caveat to my story three posts above.
I suppose, you know in theory...... I could have seen which op-amp that portable device was using online somewhere, while reading the internet aimlessly. I could have then remembered that detail, without knowing it.
Then, when I listened to this device, my subconscious tells me it's supposed to sound like OPA1612, while I'm consciously thinking I have no idea what is inside of that device.
That is, once more, a pretty advanced illusion, I don't think that's the case, just saying so for completeness.
It could have been sheer luck as well, in theory, but I was quite sure I was hearing the OPA1612.
Side-note, I don't like that chip very much, I prefer LME49990 for transparency and many other chips for sheer sonics.
Their specs in the spec sheets shouldn't necessarily reflect reality, they may have idiosyncratic errors, yet another theory I'm putting out now just so someone doesn't answer me "you can't hear 0.0000X you psycho basket".
I suppose, you know in theory...... I could have seen which op-amp that portable device was using online somewhere, while reading the internet aimlessly. I could have then remembered that detail, without knowing it.
Then, when I listened to this device, my subconscious tells me it's supposed to sound like OPA1612, while I'm consciously thinking I have no idea what is inside of that device.
That is, once more, a pretty advanced illusion, I don't think that's the case, just saying so for completeness.
It could have been sheer luck as well, in theory, but I was quite sure I was hearing the OPA1612.
Side-note, I don't like that chip very much, I prefer LME49990 for transparency and many other chips for sheer sonics.
Their specs in the spec sheets shouldn't necessarily reflect reality, they may have idiosyncratic errors, yet another theory I'm putting out now just so someone doesn't answer me "you can't hear 0.0000X you psycho basket".
Makes perfect sense to me. By the repetition, you steadily built up a short term memory image of what the music piece was about, which had all the qualities of the best version of it, presumably the 'proper' DAC's rendering,. Finally, the brain was able to overlay the 'inferior' replaying of the piece with the optimum, internally built up version - your mind 'masked' the deficiences of the poorer version, it interpolated the extra qualities it knew should be there.
Sonic memory total is only a few seconds, right?
At least, volume memory is.
I'm not sure about attack and decay memory, time memory, noise memory, pattern memory, music memory, soundscape memory...... ad infinitum.
I hope so!
I started listening to op-amp's and DAC's in 2012, I really like to collect them, however I've had some inner conflict you could say.
Right now I feel like fresh air.
Taking a pause from the internet, I'll upload my theories later.
The truth is not afraid of inquisition.
See you!
I take no notice of repetitively listening to sounds; the first sense, take of what I'm hearing is all-important - the freshness of the sounds to my hearing is what counts. I do Yes/No judgement calls - is the sound acceptable or not? If not, what's the most obvious thing wrong with it? I then ignore everything else until I have an answer to that "most obvious wrong thing" - and once refreshed, go through the whole exercise again ...
What you are describing, Frank, while reasonable, is also a relatively rare case in real life. I find that op amps rather rarely have such big differences between them as to be obvious in the first go.
To me, the proof of the pudding is when you listen to something for some time, where time is measured in days. The reason for thisis that we are not always in the same mood ourselves, and that does influence what we hear. By making the test longer, you allow for such factors, and sometimes, with a close call, that's the only way to make any difference.
On very close calls, I find that sometimes I have to repeat the whole procedure all over again.
To me, the proof of the pudding is when you listen to something for some time, where time is measured in days. The reason for thisis that we are not always in the same mood ourselves, and that does influence what we hear. By making the test longer, you allow for such factors, and sometimes, with a close call, that's the only way to make any difference.
On very close calls, I find that sometimes I have to repeat the whole procedure all over again.
That's right. The look is deceptive (not ugly , not boring, as you say, retro). So very correct to do that. I was very surprised by them. The only speaker that was similar to me was Spendor Prelude made of surplus BC3 drive units. Not like BC1 or 3. Yours I imagine to be more capable as sound levels increase. The Spendor Prelude also had the ability to drive the room. That is you close your eyes and the speakers are difficult to locate whilst music is easy to position. BC1 on a rare occasion could do that . Usually the speakers 1.5 metres from any wall. The older BC1 sounded better. I was told animal glue is the reason. I did two identical total rebuilds on BC1 with even the same drive unit spec. The older pair sounded open and projected well. The newer pair sounded enclosed and boring. Everything was the same except cabinet. Someone said Spencer would have redesigned to suit the modern plywood and he would have realized the change in sound.
Nige
Someone called it the Hook. A little amp called Rational Audio had that. Sad thing is no one would chance buying one. It had built in loudness control. I felt no need to unsolder it. Linn Isobariks were almost as good as Linn believed them to be with that amp. It trounced anything else with them that I heard except the Krell (yes, I can't believe any of that either). They are not easy speakers, dry bass, weird stereo, difficult load, metronome rhythm. Real life can sound like that, they are not a commercial sound. They have big virtues and even bigger flaws. Playing bad recordings is part of what a speaker must do, Isobariks are not happy doing that. The reason is a lack of balance. It is not acceptable that a transistor radio does bad recordings better. Some of that was an amplifier not liking the speaker is my best guess?
If I remember correctly Rational Audio amp was 9 transistor with feed-forward. It was nothing like a Quad 405. It had a DIN plug and socket on the volume control shaft. Pull the control forward to separate it and have preamp power-amp split. The weird bit is the re-coupling was easy and felt like some expensive custom part.
Dejan, the sort of issues I listen for are probably somewhat different from those you do, and I use a specific technique - stress testing. I listen loud, using 'difficult' recordings, and am on the lookout for defective playback - this takes me straight in to where differences are, the deficiencies are usually very obvious. I find if the system is then upgraded to handle and pass that type of testing then more normal playback is a doddle, the "safety margin" assures good results every time.
The only problem I have with your approach is the fact that most of the time, I use very little average power, both because my room is small and because my speakers are relatively effcient.
Ramping up power changes the natural behavior of my own ears, in my belief in accordance with the Fletcher-Munson curves. I have heard a few amps which sounded just fine at those levels, but botched it up at the lower levels I normally use. The effect of changing tonality at low and higher power works BOTH ways, you know.
So, to make sure, I do both tests. I start with normal levels and work my way up to very high levels I actually never use at home, not even when I'm alone and decode to let it rip a bit. My wife is one of those who can hear the paint dry, but oddly enough, if I'm saying something she doesn't like, she switches on her Smith-Papoulos filters at 375 dB/oct and corner frequency of 0.01 Hz.
So, she CAN'T hear me. And of course, it's my fault. Starting low and working your way up has the added benefit of telling me at which point will the amp start to change tonality, which is a nice little pointer where to look. My Marantz 170 DC keeps its cool all the way up to, or very near to, its nominal output, which is s good show and is rare in mass produced products. In other words, my chances of ever hearing it without aiming for it are realistically zero, never going to happen under normal circumstances, not even if I gun it quite a bit.
|
There are differences in PSU rejection, that's the spec called PSRR.
Then there's EMIRR, RFIRR, CMRR, slew rate, settling time, noise spectrum, distortion spectrum, output current.
Then there's rise time, fall time, overshoot, small-signal transient response, large-signal transient response.
Can rise, fall, undershoot, overshoot, transient response, settling time...... theoretically cause sound quality differences?
For example the specs in this paper, page 3, page 9.
http://cds.linear.com/docs/en/datasheet/1363fa.pdf
I just noticed on page 9 there, for high drive current they recommend low ESR, 1uF to 10uF Tantalum capacitors.
Just to note, Tantalum may have low ESR, however according to these papers it has high distortion at 100 Hz, 210 Hz, 310 Hz
http://psykok.dyndns.org/diy/UP/Youpi/PCBs/Capas/EW-WW_CapsSound_Part6.pdf
High EMR at 9 kHz
http://www.hificritic.com/downloads/APassiveRole.pdf
Note, that in the first paper the Evox Rifa Metallized PET measure the lowest in THD numbers.
In the second paper, the Metallized Polypropylene seem to have the cleanest meaurements as well.
Perhaps an amplifier only using metallized capacitors would yield pretty nice results?
Metal-film resistors as well for an all metal design.
Just an idea, not sure how it translates into reality.
|
Last edited:
Kastor, practically everything you mention has been discussed on these pages already, some items even several times.
The thing is, 41 years after the appearance of the Otala/Lohstroh paper in IEEE on TIM, and the attendent TIM-less amplifier, we are still where we were conceptually, with two schools: those who claim much NFB is good, and those who believe that too much NFB is not good and ceratinly not necessary.
Here, John Curl, Demian Martin and me are the loudest proponents of wide bandwidth and low global NFB as the way to go, some are still thinking about it, and the rest are of the much global NFB school.
From this, logically there are a number of approaches to the rest of the list you supplied. Wide bandwidth approach for example doesn't have a problem with very short rise times by definition, and most don't have a problem with reasonably high slew rates, while far too many still cling on to the notion that lower THD means better sound.
Just dig through the log.
The thing is, 41 years after the appearance of the Otala/Lohstroh paper in IEEE on TIM, and the attendent TIM-less amplifier, we are still where we were conceptually, with two schools: those who claim much NFB is good, and those who believe that too much NFB is not good and ceratinly not necessary.
Here, John Curl, Demian Martin and me are the loudest proponents of wide bandwidth and low global NFB as the way to go, some are still thinking about it, and the rest are of the much global NFB school.
From this, logically there are a number of approaches to the rest of the list you supplied. Wide bandwidth approach for example doesn't have a problem with very short rise times by definition, and most don't have a problem with reasonably high slew rates, while far too many still cling on to the notion that lower THD means better sound.
Just dig through the log.
Last edited:
I see, I'll have a look through the thread later. I don't even know what TIM is. 
I believe in the low distortion school then, after all, why else would LME49990 sound better than LME49710?
By the way I like Maserati's too, they look damn good.
I just had an idea, why not make two "The Wire" amplifiers, one with LME49710 and one with LME49990, then blind test them? If this is a viable test, I'll consider this.
I believe in the low distortion school then, after all, why else would LME49990 sound better than LME49710?
By the way I like Maserati's too, they look damn good.
I just had an idea, why not make two "The Wire" amplifiers, one with LME49710 and one with LME49990, then blind test them? If this is a viable test, I'll consider this.
Last edited:
I see, I'll have a look through the thread later. I don't even know what TIM is.
I believe in the low distortion school then, after all, why else would LME49990 sound better than LME49710?
By the way I like Maserati's too, they look damn good.
I just had an idea, why not make two "The Wire" amplifiers, one with LME49710 and one with LME49990, then blind test them? If this is a viable test, I'll consider this.
Briefly, TIM stands for Transient Intermodulation. This is an effect when two or more dynamic signals create sounds which do not exist in the original source material. Since they are dynamic, they do not show up in classic IM measurements because they are not steady state, like pure sine waves, but are much like actual musical signals, which vary over time.
One of the possible reasons for this to happen is excessive use of global NFB. Another is insufficient speed of the Miller compensation, and so forth, there's more.
Since its prime causes have been identified, it is considered that this is no longer a problem these days, although I sometimes wonder.
Same thing about slew rate. It has been discovered that its prime problem is insufficient current to drive the lead compensation circuits; there are other problems also, but they are considered of lesser importance. A very informal standard has been set that if an amp's slew rate is 1 V/uS per every peak volt of output, then slew rate and slew induced distortion (SID) will not be a problem. Yet, right here on the forum, there are people who think even that is not really enough, for example, John Curl, Richard March, myself, etc. John for example thinks 100 V/uS should be standard for amps hoping to be better than most, Richard is not usre if even that is enough, and I think we should get all the slew rate we can without compromising something else.
The point is, when some of these specs were met in the late 70ies and early 80ies, they stopped being even quoted in spec sheets, which could lead one to believe that they were a thing from the past. I seriously doubt that, especially with regard to the multitude of free for all schematics available on the Internet, where it is possible to see with simple mental calc that a surprising number of those projects will fail some basic requirements - and especially the slew rate.
- Status
- Not open for further replies.