Hi everyone,
I need to isolate the spdif signal from a noisy source to a WM8805 transceiver. I have been reading that the toslink connections available nowadays aren’t able to do 192khz anymore. So that leaves me with ‘normal’ spdif.
Since my two boards are in the same enclosure, i was thinking to just use a high speed optocoupler to do the isolation. Something like this https://www.mouser.com/pdfDocs/Renesas_RV1S9960A.pdf
Do you see any issue with that? (Or is there a better solution?)
Ideally, i could also use this method to use a coax spdif from a computer
I need to isolate the spdif signal from a noisy source to a WM8805 transceiver. I have been reading that the toslink connections available nowadays aren’t able to do 192khz anymore. So that leaves me with ‘normal’ spdif.
Since my two boards are in the same enclosure, i was thinking to just use a high speed optocoupler to do the isolation. Something like this https://www.mouser.com/pdfDocs/Renesas_RV1S9960A.pdf
Do you see any issue with that? (Or is there a better solution?)
Ideally, i could also use this method to use a coax spdif from a computer
Why not use a pulse transformer? The issue with optos (and Toslink) is the pulse width distortion - I guess how much jitter results depends on which edges of the input signal the S/PDIF receiver uses to regenerate the bitclock.
What is the noisy source? Maybe the problem can be solved at the root. Or add a SPDIF transformer and solve the noise at the root.
When using a non standard solution standard sources won't be easy to connect. BTW there are Toslink transmitters and receivers that do 192 kHz but it is trial and error.
When using a non standard solution standard sources won't be easy to connect. BTW there are Toslink transmitters and receivers that do 192 kHz but it is trial and error.
Last edited:
I’ve actually bought some ethernet pulse transformers but i am confused how to use them in this application. The differential pairs are the issue for me. I got these Blocked the combination of differential and common mode confused me. Do you have experience with these?
Unfortunately the source is a wifi receiver board and the only way to keep it happy is to put it on an isolated supply behind some emi shielding. I’m currently using i2s to transfer the audio (through a galvanic isolator). I bought the transformers to filter i2s signals but i found that a dedicated isolator chip was easier.
The goal is to add an additional spdif input, so i thought i would use spdif instead of i2s for the wifi board and put the wm8805 between as a mux (connected to the i2s of my dsp). That way i could easily switch between spdif inputs without needing to replace the dsp.
How can i test if the toslink is able to do 192?
Unfortunately the source is a wifi receiver board and the only way to keep it happy is to put it on an isolated supply behind some emi shielding. I’m currently using i2s to transfer the audio (through a galvanic isolator). I bought the transformers to filter i2s signals but i found that a dedicated isolator chip was easier.
The goal is to add an additional spdif input, so i thought i would use spdif instead of i2s for the wifi board and put the wm8805 between as a mux (connected to the i2s of my dsp). That way i could easily switch between spdif inputs without needing to replace the dsp.
How can i test if the toslink is able to do 192?
Attachments
No ethernet transformers are for a different purpose: ethernet.
You can test 192 kHz in your setup and you will learn quite fast if it works or not. Point is that there are various brands making these and none of them are Toshiba. Toshiba stopped producing these after the factory either was ruined by a flood or a fire.
If the transmitter does 192 kHz maybe the receiver does not and if they do both you're a lucky guy 🙂 I used to replace them both for the same brand/series but nowadays I don't bother as coax is better and USB in some cases (but certainly not all).
You can test 192 kHz in your setup and you will learn quite fast if it works or not. Point is that there are various brands making these and none of them are Toshiba. Toshiba stopped producing these after the factory either was ruined by a flood or a fire.
If the transmitter does 192 kHz maybe the receiver does not and if they do both you're a lucky guy 🙂 I used to replace them both for the same brand/series but nowadays I don't bother as coax is better and USB in some cases (but certainly not all).
The is some SPDIF and TOSLINK interfacing info in the SRC4392 data sheet. Begins about half way down page 42 and continuing discussion for a few pages beyond that.
TOSLINK is commonly not supported at 192kHz. Have to look at user manuals, data sheets, etc., and or do some testing.
TOSLINK is commonly not supported at 192kHz. Have to look at user manuals, data sheets, etc., and or do some testing.
Cool. I’m looking for a readout or something to verify. But you’re saying i can restrict the output and just see if it works. I could probably test it with my computer, actually. It has a toslink output. I’ll force the bitrate and listen.
I do 24/96 all the time over fiber in my car system, just haven’t tried to go higher. All my ‘home’ stuff is HDMI on my primary system so most of my toslink gear is pretty dated at this point.
But the other part of my issue is that two of the boards i want to connect are in the same enclosure. Adding toslink connectors and fiber isn’t practical for that connection. That’s where my optocoupler idea came from: the toslink is really just an optocoupler with a fiber in the middle. I figured that there should be an optocoupler that has similar enough specs to just drop in (at least from a physics perspective, the devil is always in the details)
I do 24/96 all the time over fiber in my car system, just haven’t tried to go higher. All my ‘home’ stuff is HDMI on my primary system so most of my toslink gear is pretty dated at this point.
But the other part of my issue is that two of the boards i want to connect are in the same enclosure. Adding toslink connectors and fiber isn’t practical for that connection. That’s where my optocoupler idea came from: the toslink is really just an optocoupler with a fiber in the middle. I figured that there should be an optocoupler that has similar enough specs to just drop in (at least from a physics perspective, the devil is always in the details)
I'd suggest not - the main issue is the primary inductance is showing as 300uH. Whereas for a purpose designed S/PDIF pulse trafo, its roughly 10X that. PE65612 is what I use on S/PDIF.
Access Denied
Perfect. Thanks. I recognize that part. I found a thread where people recommended a newava part as well https://www.digikey.com/en/products/detail/newava-technology-inc/S22083/555461 i’ll pick one up to do some experiments for the future.
For the moment, I think i am going to go a different direction and leave out the second input until i swap DSP to one that has more inputs and can handle multiple clocks (adau1466 or so). I realized that it might not be so graceful to swap inputs and master clocks to change sources.
For the moment, I think i am going to go a different direction and leave out the second input until i swap DSP to one that has more inputs and can handle multiple clocks (adau1466 or so). I realized that it might not be so graceful to swap inputs and master clocks to change sources.