Jump to content
IGNORED

Understanding USB


Recommended Posts

I know some basic electronics, and would like to hear more from others that understand USB and it's shortcomings.

 

I understand there are 4 pins (DATA+, DATA-, 5V, GROUND).

 

My understandng is that the issue with usb noise is on the 5V bus or pin.

 

My "thinking" is that the two pins for data, even if they had noise on them, that it should be simple to transfer a bit perfect 1s & 0s via those 2 pins. 

I mean all you are looking for on those pins is (ON OR OFF), and any simple design should be able to transfer that data perfectly because you are just looking for a one or a zero and any noise should be able to be filtered very easily.

 

If the 5v is not used in the case of isolation or external 5v circuitry, I do not understand what the issue is?

 

ANY source should be able to provide "perfect" binary data to the dac....and if the 5V is properly isolated, then the source REALLY does not matter?

 

All this hoopla around stopping services, different OS's, what computer, or other media device really should make NO DIFFERENCE whatsoever?

 

Obviously the DAC circuity will, but the hardward source really should not be an issue?

 

Link to comment
10 minutes ago, beerandmusic said:

My understandng is that the issue with usb noise is on the 5V bus or pin.

 

My "thinking" is that the two pins for data, even if they had noise on them, that it should be simple to transfer a bit perfect 1s & 0s via those 2 pins. 

I mean all you are looking for on those pins is (ON OR OFF), and any simple design should be able to transfer that data perfectly because you are just looking for a one or a zero and any noise should be able to be filtered very easily.

 

If the 5v is not used in the case of isolation or external 5v circuitry, I do not understand what the issue is?

 

 

A couple of points where it looks like you've not quite understood the issues. Firstly, USB noise is on all the pins of the cable, not just the 5V. And secondly, the noise is nowhere near high enough amplitude to result in digital data becoming flipped.

Link to comment
2 minutes ago, opus101 said:

 

A couple of points where it looks like you've not quite understood the issues. Firstly, USB noise is on all the pins of the cable, not just the 5V. And secondly, the noise is nowhere near high enough amplitude to result in digital data becoming flipped.

 

I did state that even if there was noise on the data, that should be able to be reconciled very easily, since all you are looking for is on or off, and any circuitry should be able to provide bit perfect data very easily....if this wasn't true we have would have huge issues with transferring data of all kinds (not just music).  That technology has been perfected with the advent of usb.

Link to comment

^^^  I would be shocked if any DAC engineer would suggest that there is difficulty in managing a bit-perfect source....I could understand issues with converting the bit-perfect source into accurate analog, and how noise introduced into the dac could cause some issues, but that there is not ANY difficulty getting the bit-perfect data.

Link to comment
2 minutes ago, Ralf11 said:

then there is the question of what any noise does...

 

and how much noise it takes to eff up whatever

well with my basic electronics knowledge, noise should not affect square wave data in ANY way.

And if the DAC uses its own 5V circuitry, i find it difficult to believe that there is any issue with the SOURCE.

I do believe that DACs will have their own signature sound based on implementation, but that noise from the source should no longer be an issue.

Link to comment
5 minutes ago, hdo said:

Only poorly designed DAC will blame USB!

This i agree with...and the thought of having to be a "special cable" is totally insane.

I keep thinking about how data can be transferred perfectly to an external hard drive 100% of the time, that if noise made a difference, then we would have garbage data when it is transferred.

 

We are talking square waves, on or off, 1s and 0s, and this is very old and mature technology.

 

Link to comment
9 minutes ago, Ralf11 said:

what if the noise adds to the '0' of your square wave to bring it up to a 1 ??  or vice-versa

 

what if noise -somehow- alters the timing?

 

then we would have spaceships tended for mars landing on the moon.

Seriously....if you can copy a data file a million times to a flash drive, and did a checksum a million times, it would be perfect every time.

 

 

Link to comment
10 minutes ago, sandyk said:

 

 

We are exporting an analogue representation of the binary data, and it will be far from a perfect square wave when received. There will also be system noise riding along with the Analogue waveform.

There are many 100s of C.A. members that have found that it is far from as simple as you seem to believe it is.

What you are in effect claiming, is that all the members who use Regens, improved PSUs and USB cables with better isolation  between the Data and PSU wires are (that nasty word again), DELUSIONAL ! 

 

Have you even bothered to read any of the stuff posted by well respected E.E. John Swenson in the Uptone section of C.A. ?

 Believe what you want, it's your loss when you choose to ignore the 1,000s of confirming posts on the subject by fellow members.

 

 I am out of this, yet another,  in the series of  stupid "Anti Audiophile" threads.

BYE !

 

I suggest that you cut beerandmusic some slack.

 

He isn't "anti-audiophile".

 

He is simply, and I say this with great respect and love, "clueless" (;)) when it comes to technical matters like this and I believe that he started this thread out of a genuine desire to increase his own understanding.

 

I personally admire his willingness to put himself out there like this.

 

Sometimes it's like someone took a knife, baby
Edgy and dull and cut a six inch valley
Through the middle of my skull

Link to comment
34 minutes ago, sandyk said:

 

 

We are exporting an analogue representation of the binary data, and it will be far from a perfect square wave when received. There will also be system noise riding along with the Analogue waveform.

 

When you say you are exporting an analog signal, Are you sure about that statement?

Do you also suggest when you transfer a data file from your pc to a hard drive, that it is an analog signal?  I believe, but maybe i am mistaken, that it is a binary finle, not an analog file as you suggest.

 

Link to comment

The issue with USB boils down to noise contamination of the DAC. Noise is generated via the following avenues:

 

1. Noise transmitted from the sending device. The power network inside a computer is a hellstorm of noise, and that gets transmitted along with the USB signal and 5V power wire.

2. Self-noise generated by the USB receiver's controller. USB audio data comes in as lumps (packets) at regular intervals, causing the controller to perform bursts of processing. Also, the crappier the USB signal, the harder the controller has to work to fix it.

 

There is no such thing as "digital audio". What we have are square waves being transformed into sine waves. The human brain can perceive moments of sound far beyond its capacity to perceive frequencies of sound. Even a slight ripple in the square-to-sine wave transformation process can be perceived by humans.

Link to comment
1 minute ago, GUTB said:

The issue with USB boils down to noise contamination of the DAC. Noise is generated via the following avenues:

 

1. Noise transmitted from the sending device. The power network inside a computer is a hellstorm of noise, and that gets transmitted along with the USB signal and 5V power wire.

2. Self-noise generated by the USB receiver's controller. USB audio data comes in as lumps (packets) at regular intervals, causing the controller to perform bursts of processing. Also, the crappier the USB signal, the harder the controller has to work to fix it.

 

There is no such thing as "digital audio". What we have are square waves being transformed into sine waves. The human brain can perceive moments of sound far beyond its capacity to perceive frequencies of sound. Even a slight ripple in the square-to-sine wave transformation process can be perceived by humans.

When you say there is no such thing as digital audio, the data is stored on the media as digital (1s and 0s).  The sine wave is a result of the DAC.  It is digital data until it is converted.  It may be digital with noise, but it is not analog.

Link to comment
3 minutes ago, beerandmusic said:

When you say you are exporting an analog signal, Are you sure about that statement?

Do you also suggest when you transfer a data file from your pc to a hard drive, that it is an analog signal?  I believe, but maybe i am mistaken, that it is a binary finle, not an analog file as you suggest.

 

All signals are analog in nature, its the binary interpretation of an analog signal which gives rise to digital.

Link to comment

From wiki

A digital signal refers to an electrical signal that is converted into a pattern of bits. Unlike an analog signal, which is a continuous signal that contains time-varying quantities, a digital signal has a discrete value at each sampling point.

 

Take music out of the equation.  Let's look at the transfer of a digital file from a pc's internal hard drive to an external usb hard drive.

You have a hard drive that has 1s and 0s on it.  Let's assume a digital file with a value of 5 (one zero one),  you transfer those 3 bits on off on via the usb connector to the external hard drive.  it looks like a near perfect square wive.

 

image.thumb.png.5ca232b17820254d08a9ebd0c9a7fe42.png

 

it may have noise (jaggies on the wave), but it can be transmitted perfectly 100% of the time regardless of the noise on the square wave.

 

It is digitally stored, transferred digitally, and stored on it's destination as a perfect 5 (1 0 1).

 

If it was that difficult to transfer the data perfectly, then we would be in heaps of trouble at the bank.

 

I do agree the only issue may be that the DAC may have issues with the D->A circuitry during conversion, but that is the only thing i could see as being an issue.

 

Link to comment
13 minutes ago, beerandmusic said:

When you say there is no such thing as digital audio, the data is stored on the media as digital (1s and 0s).  The sine wave is a result of the DAC.  It is digital data until it is converted.  It may be digital with noise, but it is not analog.

 

There is no such thing as digital audio. 1s and 0s do not appear anywhere. What we see are square waves. Square waves are 100% analog in nature. There is no fundamental difference between a square wave and a sine wave. What you're referring to as "data" is, in fact, counting a period of time between rises and falls in a square wave. What we see in computer chips are nothing more than very densely packed arrays of various circuits designed to perform advanced counting procedures on square waves. Square waves are useful because they provide a large range of power fluctuation within the moment of time between changes while allowing a circuit to reliably "guess" the value. The square wave will go through the laborious sine wave transformation process every step of the way passing along any noise that gets into power network and that which eventually gets into the output circuit. 

 

Here is another figment of our imagination:

image.thumb.png.c89fc839a97a26cd9d65fd442aa6b693.png

There are no such things as "circles". This shape exists solely in our minds as an imaginary concept -- a way for us to generalize information we perceive.

Link to comment
4 minutes ago, GUTB said:

YOU>>>

There is no such thing as digital audio. 1s and 0s do not appear anywhere.

There is no fundamental difference between a square wave and a sine wave.

 

The source file is stored in 1s and 0s.

 

Agree to disagree...they are worlds a part.  A sampling of a square wave is on or off, a one or zero, where a sample of a sine wave (or analog signal) has unlimited variations in value....Tell me you are not an engineer?!

Link to comment

Just to emphasise what opus101 is saying, all electrical signals are always analogue - it's just in computers, digital processing that the signals are treated as representing data, 'pure' information. In audio the only thing that really matters is how our ears perceive the sound - this is an analogue process, not digital. And, unfortunately, it turns out that our ears are very sensitive to disturbances in those analogue waveforms which can be caused by those parts of circuit processing which are called, "digital".

Link to comment
5 minutes ago, fas42 said:

Just to emphasise what opus101 is saying, all electrical signals are always analogue - it's just in computers, digital processing that the signals are treated as representing data, 'pure' information. Since in audio the only thing that really matters is how our ears perceive the sound - this is an analogue process, not digital. And, unfortunately, it turns out that our ears are very sensitive to disturbances in those analogue waveforms which can be caused by those parts of circuit processing which are called, "digital".

 

Even in music, it is digital (and useless from an audio perspective) until it is converted.

The file that makes up the music is stored and transferred digitally as a digital signal.

It is not an analog signal until it is converted.

Link to comment

Create an account or sign in to comment

You need to be a member in order to leave a comment

Create an account

Sign up for a new account in our community. It's easy!

Register a new account

Sign in

Already have an account? Sign in here.

Sign In Now



×
×
  • Create New...