effect of HF/VHF noise on DAC output

[miero]

I have a question about presence of HF/UHF noise on the DAC output.

Can such noise somehow reduce performance of following amplifier stages?

Here are my measurements of the Buffalo 3 DAC with 4-Channel S/PDIF Level Converter and comparison of noise from Coax and TOSLINK connection.

I've reduced voltage of S/PDIF convertor to 4.53V that improved a noise spectrum. The noise seems to be generated comparator on S/PDIF convertor and also by TOSLINK module itself and it is somehow modulated to the DAC output.

Can you perform similar measurements on your digital source and post the pictures for comparison?

 

[DF96]

How much of that noise is really on the DAC output, and how much is just cable or ground pickup?

RF can affect amplifiers. The popularity of badly-made DIY or 'high-end' cables suggests that some people prefer the resultant sound. They regard the slight increase in modulation noise as extra 'detail'.

 

[miero]

Above measurements was done by probe connected to the output RCA connector. The FM band appears always when the probe is attached to any piece of larger metal. The optical TOSLINK does create ground loop, but noise is present and radiated into whole chassis.

 

[spiri]

I had similar trouble with open inputs on a single high speed comparator with the circuitry of the demo boards of ess. I measured the oscillation with my lecroy 314 but i have no picture.
An open input and also a short cable without signal at the input and a oscillation occurs. Although the internal hysteresis should be large enough, but both inputs driven from the same bias circuit with only an RC for signal biasing seems too risky. Good decoupling of the supply voltage and the bias voltages can help but I did not test a quad comparator.
To rule out that it comes from the comparators, switch the 100n at the unused inverting inputs to ground.
I hope this helps.

 

[marce]

EMC

 

[miero]

To rule out that it comes from the comparators, switch the 100n at the unused inverting inputs to ground.

There is already 100n between ground and inverting input -- http://www.twistedpearaudio.com/docs/digital/spdif_4_schematic.pdf

I'll try to replace coaxial cable between the module and the toslink/coax connector by a cable with better shield. This one seems to have sparse shielding.

 

[spiri]

Excuse my english, I meant to replace the capacitor by a short-circuit on the unused inputs.

 

[abraxalito]

Can such noise somehow reduce performance of following amplifier stages?

For sure, though the magnitude of the effect depends on the following amplifier. If it has effective ultrasonic filtering (by far the majority don't) then it'll be largely immune. The subjective effect of such noise is a reduction in dynamics.

 

[marce]

Try measuring in a faraday cage, you have lots of EMC problems so need to try and measure the equipment without the external influences.