I2S galvanic isolation: TOSLINK optocoupler or transformator???
I would like to decouple galvancally the I2S source from my DAC, without introducing extra jitter
3 options:
optocoupler
transformator, maybe spdif or ethernet transformator
TOSLINK using 3 fibers
other optical link using high speed toslink (coming from ethernet over fiber equipment)
Any advice?
I would like to decouple galvancally the I2S source from my DAC, without introducing extra jitter
3 options:
optocoupler
transformator, maybe spdif or ethernet transformator
TOSLINK using 3 fibers
other optical link using high speed toslink (coming from ethernet over fiber equipment)
Any advice?
Last edited:
True, everything can add jitter, optos I believe are the worse.
You can also add aDum devices to your list.
You can also add aDum devices to your list.
But does it add sufficient jitter to caue a difference which you can measure or hear?
is the galvanic isolation of the grounds more beneficial than the jitter introduced?
I suppose that I2S over 3 toslinks is still better than SPDIF over 1 TOSLINK?
I suppose that I2S over 3 toslinks is still better than SPDIF over 1 TOSLINK?
IMO that question has missed the point slightly.
The question should be: Is the added jitter worse than the noise removed from the system for the whole system performance? eg Is there a nett gain in system performance?
stistill not implemented, but seems that I2s from sound card are quite polluted with digital noise
Seems? i work as part of a couple of teams that do a lot of audio based design, digital noise is quite often inaudible. With noise you do need to do measurements to determine the frequency and the source of the noise. Then you can work out whether it is serious and how to combat it.
going from the sound card directly to the TDA1543 NOS dac having common ground will not be too bad?
Hi, I have a setup of RPi and asynchronous DAC with is own precise clock. They are linked with I2S and they share PSU. PS connections are isolated inbetween and PSU is very stable on its own. I did not notice any noise, however I would like to add extra isolation to I2S connection. Jitter should not be a problem in my case as I have DAC with its own clock. I am looking to buy this one: (NVE IL715) http://www.nve.com/Downloads/il71x.pdf What is your opionion on my decision? Thanks.
Galvanic isolation is to isolate stray currents and for voltage protection it is not primarily for noise isolation, noise isolation is a far more complex problem to solve and a lot is down to correct layout and correct filtering from source to destination. This is something that I very rarely see done in the amateur world of layout.
The devices themselves all have some coupling capacitance between sides, this is the main coupling method for high frequency noise, so some noise will cross the devices, so more than just slapping an isolating chip down is required. I have to ask is it really necessary to do this, are you likely to gain anything and more important do you have the equipment to test whether your solution works, a decent scope is the minimum you would require.
The devices themselves all have some coupling capacitance between sides, this is the main coupling method for high frequency noise, so some noise will cross the devices, so more than just slapping an isolating chip down is required. I have to ask is it really necessary to do this, are you likely to gain anything and more important do you have the equipment to test whether your solution works, a decent scope is the minimum you would require.
If I understand this, you have a Julie@ PCI sound card from which you wish to tap the I2S and feed to a TDA1543 DAC? I assume the Julie@ is inside a PC?
Ignoring the isolation question for a moment, given that you want to keep I2S runs as short as possible it implies that you will need to install the DAC inside the PC as well? If so, that is a pretty noisy environment period.
If it were me (and I have successfully done something very similar) I would invest in a pair of Twisted Pear Teleporters; these use LVDS to send the I2S over a considerable distance and will give you isolation too. I would also consider reclocking the I2S immediately before the DAC.
I assume you know that you can expose the I2S on the Julie@ simply by removing the top section of the card?
Here's some information on the Teleporters;
http://www.diyaudio.com/forums/twisted-pear/201106-introducing-bit-teleporter.html
Teleporter Digital Transfer Module
Whoops, just noticed how old the original posts were so you can ignore this.
Ignoring the isolation question for a moment, given that you want to keep I2S runs as short as possible it implies that you will need to install the DAC inside the PC as well? If so, that is a pretty noisy environment period.
If it were me (and I have successfully done something very similar) I would invest in a pair of Twisted Pear Teleporters; these use LVDS to send the I2S over a considerable distance and will give you isolation too. I would also consider reclocking the I2S immediately before the DAC.
I assume you know that you can expose the I2S on the Julie@ simply by removing the top section of the card?
Here's some information on the Teleporters;
http://www.diyaudio.com/forums/twisted-pear/201106-introducing-bit-teleporter.html
Teleporter Digital Transfer Module
Whoops, just noticed how old the original posts were so you can ignore this.
Last edited:
- Status
- Not open for further replies.