1. This site uses cookies to help personalise content, tailor your experience and to keep you logged in if you register.
    By continuing to use this site, you are consenting to our use of cookies.

    Dismiss Notice

Testing audiophile claims and myths

Discussion in 'Sound Science' started by prog rock man, May 3, 2010.
First
 
Back
905 906 907 908 909 910 911 912 913 914
  1. 71 dB
    I am feeling the same… …I don't get what exactly this discussion is about. Who cares what the voltage is as long as we don't overdrive any input causing distortion and also have high enough signal level to not have signal-to-noise issues? Any reasonably designed piece of audio electronics should have plenty of safety margin in this regard so that distortion doesn't happen in the input stage.
     
  2. KeithEmo
    Let me try to clear up a few things.

    A balanced signal does not include a "hot" and "cold" signal - even though they are sometimes described that way.
    The (+) signal is a full level in-phase signal and the (-) signal is a full level signal that is inverted (exactly 180 degrees out of phase with the first).
    At the input those two signals are subtracted (remember that subtracting a negative value is the same as adding a positive value).
    The result is an output signal that is TWICE the level of the in-phase input signal.

    The idea is that any noise that impinges on both wires at the same time will be added to both.
    Then, when you subtract the two signals....
    The desired (+) signal, minus the desired (-) signal, gives you TWICE the desired signal.
    But the undesired noise signal (on the + line), minus the undesired noise signal (on the - line), cancels out... and the noise mostly goes away.
    This is the main reason why balanced connections are used.
    (Certain distortions that might arise in the output and input circuitry may also cancel out.)

    While it's true that some people consider the levels involved "so obvious they don't bother to mention them"....
    It's also true that, since the main purpose of a balanced connection is NOT "to give you a higher level signal", not everyone treats it the same.
    On some equipment, the two lines of a balanced connection do each carry a signal at the same level as a single unbalanced connection...
    However, on a lot of gear, especially consumer gear, the balanced outputs and inputs operate at the higher "pro level", while the unbalanced inputs and outputs operate at a lower "consumer level".
    And, on a lot of home audio gear, both lines of the balanced inputs and outputs operate at the lower "consumer level", which still provides the benefit of noise immunity.
    And, because, of this, in some gear, the two signals in the balanced connections are subtracted as they should be, then the result is divided by two, to produce a lower level "safer" signal.

    You will also find similar disagreement on what the "standard level" of an unbalanced signal "should be"...
    There are "standards" that call for 0.770V, 1.0V or 2.0V ... and there is no single obligatory standard.
    Therefore, most outputs are designed to be able to deliver enough voltage to satisfy any of those requirements...
    And most analog inputs are designed with enough gain to work with most of those levels as well.
    (You will also find that most consumer gear specifies an "output level" and a "MAXIMUM output level" - which is much higher - to cover that range of possibilities.)

    Most two-channel analog preamps are set up in such a way that their analog inputs can handle much higher voltages.
    In a typical analog stereo preamp, you have a passive Volume control first, followed by an active gain stage, which together can tolerate very high input voltages with no problem.
    HOWEVER, in typical home theater gear, the input is immediately digitized, at which point you may have to worry about overloading the input ADC circuitry, which may occur at only 2V or so.

     
  3. goodvibes
    The great majority of analog preamps have a line stage (often a unity gain buffer) before the V control and additional gain after. Not arguing here, just clarifying. Too many on this board believe that balanced is inherently better but it's not. just a different way of doing things. Common mode rejection doesn't tend to offer much in home kit.For balanced to be better or even as good, the up and down phase would need perfect symmetry which is more difficult to achieve in the dynamic scheme of an amplifier than many appreciate.
     
    analogsurviver likes this.
  4. analogsurviver
    @goodvibes : Correcto mundo. Coming from analog, could not agree more.

    @KeithEmo : Coming from analog, I am acustumed to having AMPLE input overload margin. Not tube ample - not THAT high - but can't possibly disagree with the fact that preamps and power amp front ends running on VERY high power supply voltage rails ( dwarfing the bog standard +- 15 VDC most IC based preamps operate on ) do sound better in real use. If nothing else, this capability comes handy when dealing with a scratch or pressed defect on the analog record - the output of which will most likely clip the low power supply voltage powered circuits. Two of the solid state preamps I like both operate at +-45 + ( few volts above that , less than 50 ) supply rails - and do not encounter any of the stridency most IC based preamps fell prey to when having to deal with ample input overload.

    It is far lesser concern with digital, but still having the capability "not needed/required" ( not just for say 1-2 dB above 0 dBFS ) on board does work better in the end.
     
  5. goodvibes
    Lots of headroom is good but so is stiffness of supply. If it's not bouncing I really don't think those sorts of margins are required. ICs have tiny tracks. I have often found that even when using some DAC chips etc, one channel of a stereo chip per side is preferred to running one chip for 2 channels. I think the issue with ICs (and some are good) is more about current than anything else. Trade offs because they can also be very fast and refined. There's a reason beyond costs that everyone including the old guard uses them in certain areas of low level circuitry.
     
  6. analogsurviver
    Well, correct again. I grew so acustomed to power supplies that still do not flinch even when the circuitry they power is grossly overloaded. It works much better than when the power supply can get modulated by the music it is supposed to back up flawlessly no matter what.

    Yes, the tracks in ICs are tiny - and they do limit the current delivery. And the reason why even the most rabid dicrete design lovers would use ICs is the fact that they are, by the very definition - matched for electrical parameters. It is on the same die of silicone, after all ... Getting a discretely build circuit to anything the precision even common ICs have long ago taken for granted is no laughing matter - it does take lots of money and lots of time - which ultimately translates into yet more money.

    ICs are more reliable. Any piece of electronics that is in off condition cold but turns into an oven when actually running non-stop, is a time bomb. There are only two kinds of these: those that did fail and those that WILL. Class A amps are the most notorious for this - not to even start about valve/tube variety.

    A clever designer would know the pros and contras and use the most (cost) effective solution in order to reach the goal envisioned at the beginning of the project. Valves/tubes are oft maligned in this thread; there are applications and uses they can not be ( yet ...? ) bettered by any other known technology. Audio included.
     
  7. KeithEmo
    That makes perfect sense to me.

    With analog sources it's quite possible for a noise like a tick or pop to very briefly exceed the maximum level of the normal signal - by a very significant margin.
    A tick or pop coming from a vinyl album may also contain very high levels of very high or even ultrasonic frequencies (this is how many "tick and pop removal" methods recognize them).
    (With digital signals this is essentially impossible - since there is a specified "hard maximum signal level" at "0 dB" - the highest number that can be encoded.)
    Therefore, it makes perfect sense that such a signal, when passed through a preamp or amplifier with lots of extra overload margin, might sound noticeably better.
    (I would expect this to be most significant for the phono preamp itself...)

    Analog circuitry often responds unfavorably to overloads, and may ring or otherwise distort for some time after being overloaded, which can make the overload far more audible than a simple "loud tick".
    And, if so, then having a substantial overload margin would certainly reduce or prevent this.
    (The input filters that precede A/D converters are also often rather sensitive to overload.)
    And, yes, many circuits that use lots of feedback, like amplifiers using op-amps, tend to respond more dramatically to overload than other types, when driven outside their normal operating range.
    Amplifier circuitry can be designed to respond "gracefully", even when it does overload, but this is something that is often not given much attention, and may account for significant differences in sound.
    (After all "you're not supposed to overload it".)

    This may well account for something that many people who listen to vinyl have noticed (I've never heard a name assigned to it but I always think of it as "the dirty windshield effect".
    If you have dirt and bugs on your car windshield, they obviously reduce your visibility to a degree, until you remove them.
    However, because your eyes have a limited depth of field, when you're watching the road, the dirt and distracting flecks on your windshield are out of focus to your eyes.
    Because of this, even though collectively they reduce visibility, you tend not to notice individual bits and bugs... unless you deliberately focus your eyes on the glass.
    Similarly, if you play a vinyl album with lots of ticks and pops, even though you hear the noise, it often seems to be "on a separate layer from the music"....
    Because of this, even when there is a significant amount of noise, we often find it "easy to hear the noise as distinctly apart from the music".
    This effect then seems to disappear when we digitize that music (if we fail to remove the ticks and pops first).
    It could well be that, by perturbing both the analog and conversion gear in the signal path, and producing distortion that extends past the actual time of the tick itself, the noises are being made more audible.

    (Just for the record, I personally have always found ticks, pops, and surface noise to be a distraction that I am simply unable to ignore... which is why I prefer digital audio to analog.)

     
  8. KeithEmo
    I would agree - provisionally.

    Some audio circuit designs are very sensitive to even minor variations in supply voltage.
    However, some circuit designs are virtually immune to them, and can tolerate huge variations in supply voltage with to measurable change in performance.

    The specification that describes this is called "power supply rejection".
    One thing many designers overlook is that the specifications of a chip, by itself, do not necessarily describe how it will perform in a particular circuit.
    For example, most IC op-amps, as a separate device, have spectacularly good power supply rejection...
    However, that doesn't necessarily mean that they will be unaffected by power supply variations when used in a particular circuit topology.
    (Some circuit topologies are themselves inherently sensitive to variations in supply voltage.)
    Conversely, some of the reference voltages inside a DAC are super-critical of power supply fluctuations and noise...
    However, because of this, many DAC chips include internal precision regulators to generate and regulate those reference voltages.
    (And, sometimes, when designers bypass that internal circuitry in an attempt to "improve it", they compromise this internal control.)

    It's also worth noting that many DAC chips in specific offer a special "mono mode".
    In this mode, a full stereo DAC chip is operated with both channels "cross connected", with a resulting improvement in performance.
    This mode uses both of the stereo outputs on the DAC chip to derive a single output channel.
    (With these, the benefit is derived from the special operating mode, and not from the fact that the two chips are physically isolated.)

    However, in general, using two separate chips certainly isn't going to hurt... and sometimes it helps...

     
First
 
Back
905 906 907 908 909 910 911 912 913 914

Share This Page