At least PMA can hear the different of cymbal recording at 44.1 and 96 and you don't. You will call PMA an alien?
Maybe PMA just a normal person, and you are....
PMA could hear the difference between a 96/24 file downgraded to a lossy MP3 and back to the same 96/24 format as the master file via 44.1/16.
In effect he was listening between a lossless PCM master file vs a compressed lossy MP3 file.
So please stop telling nonsense that he could hear the difference between 44.1 and 96.
Hans
I've done that, but measurements with the cap multiplier in circuit and out of circuit show no difference ((zero). And I could not hear any difference either. I will try to dog up the measurements - done about 3 years ago.
you need to decouple the supplies well. I in any event as a matter of course use a 10-22 ohm I each supply rail with a 100uF to ground = heavily damped supply rails with little of no HF on the rails. Most small signal stuff is running class A (loads are light) so higher order harmonics are also off the rails.
Simply said: music is not a continuous signal in a way that it is changing all the time in amplitude and frequency.
Each time something changes, depending on the rate of change it will induce the Gibbs effect to some degree.
Hans
When it comes to low noise regulators, I did some acid tests and this appears to beat an LT3042 on noise, hard to believe, but it's what the meters say.
D-Noizator: a magic active noise canceller to retrofit & upgrade any 317-based V.Reg.
You will find noise and PSRR measurements I hastily did towards the end of the thread. I did not expect too much, so it is not scaled and the noise amplifier acted as a high pass. However, since it directly compares this circuit against an LM317 and an LT3042, I think it all points into one direction: too good to be true, but hard to deny.
When I come back from my tour through Indochina, I will make a proper setup. That'll be toward the middle of March.
D-Noizator: a magic active noise canceller to retrofit & upgrade any 317-based V.Reg.
You will find noise and PSRR measurements I hastily did towards the end of the thread. I did not expect too much, so it is not scaled and the noise amplifier acted as a high pass. However, since it directly compares this circuit against an LM317 and an LT3042, I think it all points into one direction: too good to be true, but hard to deny.
When I come back from my tour through Indochina, I will make a proper setup. That'll be toward the middle of March.
My comment really was a subtle dig at the fact that we don't see square waves with very fast rising edges in nature.
I cant read the amplitude details om Scott's plot and it seems that its really very much a 2nd order effect.
No doubt, some will claim its audible without any formal DBT.
Thanks for the link to that thread. Post #20 mentions the factor of regulator linearity, referred to in terms of THD in the post.
https://www.diyaudio.com/forums/pow...ofit-upgrade-317-based-reg-2.html#post5642205
Interesting to find here and there little bits of evidence that others have considered that same issue as applied to audio and found it can have some audible effect.
Last edited:
Jack just mentions its hard to measure THD effects of a superreg
This sentance does not parse.
BTW I note you DIDN'T quote post 19.
Sorry. Regulator linearity was something I was thinking about as it affects analog audio circuits. I have seen a small number of other people mentioning it as a possible neglected factor in regulator design for audio. Still just an idea, nothing more.
I use those buffered LT3042 regulators for some digital audio stuff. Very good for that IMHO.
Whatever it is about regulators and analog audio circuits, in some cases choosing the right regulators can make a lot of difference in fine tuning sound quality. For one easy example, maybe think of amplifiers using folded cascodes. Not great PSRR, but they can sound very good if properly powered. A circuit with higher PSRR may be more detached from its power supply which is convenient for modular design thinking, but it may not end up sounding better to picky humans who don't know what's inside the box and don't care, they just want sound they like.
And that is relevant to DIY audio because...........?A circuit with higher PSRR may be more detached from its power supply which is convenient for modular design thinking, but it may not end up sounding better to picky humans who don't know what's inside the box and don't care, they just want sound they like.
Because we decide what goes in our boxes, and maybe we agree with other picky listeners about the sound.
Who don't know or care what's in the box, ie what they are really listening to?
In the in end if the design is good they enjoy pretty low distortion. One thing that is carefully controlled is the HD spectrum underlying what distortion there is. It is almost exclusively very low order, and pretty low level.
It seems to be something that some picky listeners are sensitive to, I would guess due to the IMD spectrum produced. Just a guess though.
Last edited:
First, you missed the regulator type, it's likely about the LT1963. Yest this one is also noisy as hell, at least for the fixed voltage versions, up to 3.3V
Secondly, it would be a good idea to get past the first paragraph in the data sheet; page 9 shows the noise for various models; the adjustable version (the only one that can get to 5V output) is shown with a noise 2x...3x lower than the fixed vesions (exercise, find out why).
Third, you missed the rest of the dynamic parameters in the data sheet, in particular the transient response.
And finally, the LT1963 qualifies as a rather poor LDO. If one really wants a decent 5V LDO (assuming it is required, case in point, it is not) there's lot of better fish in the pond.
I would like to know what distortions I prefer the sound of, I would find it hard to care much about the views of people who don't care, but that's just me, I'm not trying to sell anything, that includes my personal preferences.
Sorry, my compiler aborted after the first 3 words in the phrase. WTF is a "regulator linearity" and where/how is it specified?
In art, and 'the art of electronics' is actually to some extent an art, some people's preferences seem to have appeal to many other people. Very hard to say why some art is deemed great and other art middling, although we can always think up imagined reasons. Something about how it registers in the brain, it would seem.
- Status
- Not open for further replies.