Advice on COAX-to-TOSLink converter

Why are you presuming that you will need to convert? How do you know you won't increase jitter in the datastream to the point where the NAIM DAC won't be able to buffer it?

quote:

Originally posted by likesmusic:
Why are you presuming that you will need to convert? How do you know you won't increase jitter in the datastream to the point where the NAIM DAC won't be able to buffer it?

@likesmusic:
It is indeed an assumption on my side, but I assume that a COAX from a non-earthed, non-Naim source might be better connected by a non-electrical connection.

The question / request for advice is about is it good to do so?

Is there indeed a risk of increasing the problem by using a converter?

Are there converters out there that don't have or minimise this risk?

Is there a way to determine for sure what would be the best option?

At this moment I'm just a beesnest full of questions?

What would Naim's advise be on this issue of connecting these type of sources?

-
aleg

Personally I would think a direct co-ax connection would be better than converting to Optical via another device, however the M-Audio CO2 device is only around £35 and there are cheeper devices available from Maplins so (when the DAC is launched) it won't cost much to try out.

Eloise

quote:

Originally posted by Eloise:
Personally I would think a direct co-ax connection would be better than converting to Optical via another device, however the M-Audio CO2 device is only around £35 and there are cheeper devices available from Maplins so (when the DAC is launched) it won't cost much to try out.

Eloise

Thanks for your replies.

It will be a matter of trial and error then!

No clear-cut answer on this issue, but one tends to lean towards direct co-ax connection.

-
aleg

One would hope both ways of connecting the DAC would sound the same.

I think the rule is to try it and see what happens. Jitter isn't the issue here. It's noise that may come along for the ride from a poorer PS, ground or other source. Higher quality sources may still be better even with proper isolation etc. Processing noise or beats from other parts of the supplying device is not seen as jitter or corrected. In fact, it can't be, at least not without some sort of interpolation.

js, i don't undertand what you are saying. What does "beats from other parts of the supplying device" mean? What are 'beats' in this context?

noise spikes

js, what sources in your experience put such noise spikes in the s/pdif data stream? How do you measure these spikes?

It's easy to see spikes and ripple in the DC supply from standard switching supplies. Seen lots of graghs. Dirty supplies affect the stream via supply as voltage vs ground is the reference for any signal. Processing can also affect the stream but with generally a lower level fuzzy look when sharing supplies but in certain instances can also spike up a bit. There are switching supplies that can look pretty good on paper but I generally find them not as good as usually excessive filtering is used with it's own effects. Doesn't mean it can't be done and I've heard some decent reports on a few pieces (though expensive) that use them.

js, of course there may be spikes or ripple from some power supplies but what evidence have you that these cause a loss of data in the s/pdif stream?

quote:

Originally posted by likesmusic:
js, of course there may be spikes or ripple from some power supplies but what evidence have you that these cause a loss of data in the s/pdif stream?

I can imagine that while a signal is being processed as 0's and 1's that a power spike could cause a 0's voltage level to increase so it becomes (mis)interpreted as a 1's voltage level.

Likewise also a low-level ripple could raise the base voltage level creating a situation that small variations in voltage level sooner cause a 0's voltage level to cross the threshold and become interpreted as a 1's voltage level.

At least that is the way I interpret js' explanation.

So I can imagine that not-very-stable voltage levels for digital circuits might be able to cause such type of effects. If they actually do I cannot tell.

-
aleg

Data loss and noise aren't the same thing. When you oversample which all do, it's pretty easy to sample noise whether spikes or ripple that either gets interpolated into the signal if brought to 44k or becomes part of the oversampled high frequency stream and decoded as what it is. The data is still in there but with artifacts. It's semantics if you view this as data correct or not. That's worst case but you can also have the analog circuitry affected by this noise dirtying the ground which again is the referrence point for signal. If your ground changes, so does your signal. Don't think that this can't also affect the dig portion. Clocks need very clean and stable power to do their best.

Aleg, I thought your reasoning in the MP3 thread spot on.

js, it is hard for me to understand whether you are speculating, imagining, or have hard evidence.

So, can you give me a single example of a source you have measured where the switched mode power supply causes incorrect data to be received via s/pdif? How did you measure this?

and for my next challange........ enough already, please.