Quote:
Originally Posted by klingenfelter /img/forum/go_quote.gif
Thanks for the explanation Audioengr
I assume you mean ground loops caused by return circuit paths both through the audio cable and the chassis ground through the power cord. I'm using the optical S/PDIF connection for audio data to the Receiver. Doesn't this eliminate the ground loop or are you talking about something else?
|
This is what I am talking about. The ground from the power outlets is connected through the S/PDIF cable usually, forming a loop. Only in the best equipment usually has S/DIF isolated with a pulse-transformer.
Quote:
| I gather what you are saying is that if the bit stream high to low and low to high transitions have timing dither due to noise or clock drift, they can affect the DAC output in the Yamaha causing distortions. Is this correct? |
Correct, jitter causes frequency modulation in the analog output. It is generally the low-to-high transitions, or rising clock edges that matter.
Quote:
| I looked at the Audio Trends UD-10 unit and it looks like it accepts the PC data stream via the computers USB port. Then with a clean power supply and accurate clock, corrects the timing dither and resends the data to the Receiver. Is this correct? This unit also has an optical link to resolve ground loop problems. |
Correct, the clock in the USB to S/PDIF converter will reduce jitter of the PC clock to some extent. It is a digital locked-loop or DLL typically. The quality of the oscillator in the converter is critical, as is the power supply. Battery is best. The selection of the chip doing the DLL is critical as well. Devices that use the TAS1020 are typically the best.
Quote:
| The Receiver in my rig is a Yamaha RX-Z9, their top of line digital sound processor. It mathematically generates numerous acoustic signatures of sound halls around the world and generates the front effect speaker sound for so called 9.1 surround (front reflections from the back of the stage). My bet is that this unit performs the UD-10 type cleanup function internally before presenting the data to the DAC’s, but without studying its schematics and such it’s not possible to know how it handles these digital inputs. |
Unlikely that it "cleans-up" as you say. These DSP functions just add effects to the data. The jitter is still a big problem. Distortion in receivers like this is usually quite high. They pack way too much in one box and do not do an excellent job on any individual piece. There are usually cheap electrolytic caps that the signal must pass through etc...
There are exceptions, such as the Panasonic SA-XR55 and XR70. These are not perfect either, but at least they are digital in to digital out, so no distortions are introduced in the analog circuitry because there isn't any. The right power and output filter improvements to these receivers can make them approach world-class performance.
Quote:
With all money tied up in this rig perhaps its best just eliminate possibility of the problem.
I’m an electronic engineer but from some years ago when audio was all analog. So I can understand but am unfamiliar with the digital PC functions, but I do know about switching PS noise and clock transition dither, and am well aware of the noisy electronic environment in the PC. This environment would be very difficult for processing analog signals. Please comment technically on the above thoughts. Money is not the issue but I don’t like buying a component without knowing what I’m getting. |
The PC environment is not a good place for generation of S/PDIF output or a DAC to give you analog outs. Way too much electrical noise and clocks are very jittery, sometimes intentionally, to reduce emissions by "spread-spectrum" techniques.
Putting the conversion to S/PDIF in an external box is superior because it separates the clocks and power systems from the PC. Elimination of S/PDIF is even better. For music this is possible, but usually not for movie sound tracks. I2S is the interface that the D/A chip uses, so eliminating S/PDIF and going straight I2S from either USB or networked (Ethernet) devices is a big advantage in reducing jitter even more. It is possible to get almost perfect digital output with inaudible jitter using either USB or networked devices such as the Squeezebox or Sonos. In all cases, the computer should be viewed as only the data storage and transfer device. The timing information should be added in an outboard device. When this is done right, the performance can be superior to even the most expensive transports and CD players, at a fraction of the cost.
Steve N.
Principal Engineer
Empirical Audio
