XMOS-based Asynchronous USB to I2S interface

Just to make sure : If I read a 32/384 file from my PC, and feed a DAC that is only say 18/96 or 16/44.1, it is the WaveIO that makes the conversion ?
No :) There are NO sound processing techniques involved here as long as I'm aware off. If you'll play 32/384 KHz music file then WaveIO will output the streams at 32/384 KHz but it depends on your DAC how will handle the extra bits of data. WaveIO's output protocol (I speak about I2S ports) is a standard 64fs I2S one so, for a 24 bit resolution music file, it will output 24 bits starting with MSB first and fill the rest of extra bits with zeroes. Same for 16 bits but I didn't personally test the configuration (so theoretically I can be wrong :) ). As I said, it all depends on your DAC chip and how it will handle the entire I2S bus signals...
I hope it helps,
Lucian
 
S/PDIF Attenuators

Hi Lorien

I have been reading about the benefit on jitter by inserting RF Attenuators in the S/PDIF output line to the DAC. Mainly advocated by JKeny with his Hiface conversions.

Mainly the benefit comes I gather when the S/PDIF Output voltage exceeds the S/PDIF standard (?0.5v p-p). Some converters on the market appear to be outputting up to 5v!

Before trying this out on my WaveIO, please could you let me know what is the WaveIO Isolated S/PDIF Output voltage into say a 75 or 50 ohm load?

Has any WaveIO owner tried RF attenuators with their WaveIO?

Many thanks
Jonathan
 
Yes..

Hi Lorien

I have been reading about the benefit on jitter by inserting RF Attenuators in the S/PDIF output line to the DAC. Mainly advocated by JKeny with his Hiface conversions.

Mainly the benefit comes I gather when the S/PDIF Output voltage exceeds the S/PDIF standard (?0.5v p-p). Some converters on the market appear to be outputting up to 5v!

Before trying this out on my WaveIO, please could you let me know what is the WaveIO Isolated S/PDIF Output voltage into say a 75 or 50 ohm load?

Has any WaveIO owner tried RF attenuators with their WaveIO?

Many thanks
Jonathan


Jonathan: I think it is important to think about why these attenuators might a difference, and whether or not the difference is actually an improvement in signal transmission. The theory is that an inline attenuator will attenuate any reflections on the line enough that those reflections will not cause jitter in the SPDIF receiver (and that since, in the case of the Hi Face, the signal is already too high in voltage that the attenuator can be used without making the signal too weak). Whether or not a reflection will cause jitter will depend on the time at which the reflection arrives, and its magnitude. If the reflection arrives at the same time as the actual signal wave form, the receiver will be "confused" and jitter will result. But, better to have "no" reflections in the first place by using a good cable, and making sure terminations are all true 75 ohm...

In any case, without looking at the SPDIF transmission with a really, really good scope (and knowing how to measure high speed signals), adding an attenuator and expecting an "improvement" is kind of like playing darts blindfolded: you do not really know if you are hitting the target. You might "change" the sound, and it might sound "better" to you, but that "better" sound (in a given flawed system) might actually be the result of higher jitter.
Of course, it is good to know the specification of the SPDIF output of the Wave IO. I suspect that Lorien has built it to conform to the SPDIF spec.
 
Sensible advice Barrows. Thanks.

I think the reflection issue is really the one I was trying to improve and I suspect however good the cable and terminations (and they are unlikely to be perfect in reality) that will always be a potential issue.

For a few quid though, I might try to find the treble 20!
 
Hi Lorien, I received the WaveIO some days ago and modified them with a Lundhal LL1572. It was not so easy because the pin layout is rotated by 90 degrees compared with the Murata transformer. I did the same with the Metrum Octave and spend both real 75 Ohm BNC connectors and a profesional 75 Ohm video cable. The WaveIO is fed by a small regulated power supply with batteries. I'm very satisfied with the sonic result and the quality of the board. Thanks a lot for this beautiful product! Georg
 
Member
Joined 2004
Paid Member
The level on an spdif link may influence the line receiver in another way. Most receiveres are compatible with signals from min spdif (.25v ) to AES full level (8v) without failing. But the comparitor at the heart of the receiver most likely will degrade with overdrive. If you look at the specs for fast comparitors you will see the response time slow as overdrive increases. This is because the internal circuits need to remove stored charges to switch.

It would make sense that the jitter could be lower with less drive. I would put the attenuator at the receive end which will reduce reflections from receive mismatch whic would be more of an issue. However the cable should be well shielded. SPDIF can be a major noise generator with fast risetimes.
 
The level on an spdif link may influence the line receiver in another way. Most receiveres are compatible with signals from min spdif (.25v ) to AES full level (8v) without failing. But the comparitor at the heart of the receiver most likely will degrade with overdrive. If you look at the specs for fast comparitors you will see the response time slow as overdrive increases. This is because the internal circuits need to remove stored charges to switch.

It would make sense that the jitter could be lower with less drive. I would put the attenuator at the receive end which will reduce reflections from receive mismatch whic would be more of an issue. However the cable should be well shielded. SPDIF can be a major noise generator with fast risetimes.

DEMIAN - Thanks for the helpful advice.

Would you suggest any particular type of attenuator for this application? RF?

LORIEN - would you be kind enough to let me know what is the nominal spec of the WaveIO S/PDIF output voltage?

Many thanks
Jonathan
 
Hi Lorien, I received the WaveIO some days ago and modified them with a Lundhal LL1572. It was not so easy because the pin layout is rotated by 90 degrees compared with the Murata transformer. I did the same with the Metrum Octave and spend both real 75 Ohm BNC connectors and a profesional 75 Ohm video cable. The WaveIO is fed by a small regulated power supply with batteries. I'm very satisfied with the sonic result and the quality of the board. Thanks a lot for this beautiful product! Georg
Thank you Georg, I'm glad that you managed to wire your Lundahl trafo to your WaveIO card! I am sorry about the pinout reversal of Murata trafos but at the designing time, I was hit by the same 'issue'. :confused:

Thanks Lucian for the firmware update. Sweet..
Nice to see it working. In the weekend I'll try to find some spare time and update the infos on my fist post.

LORIEN - would you be kind enough to let me know what is the nominal spec of the WaveIO S/PDIF output voltage?
Many thanks
Jonathan
I tried to take accurate measurements but there's no chance for me here because there are a lot of interferences produced by surrounding antennas (even with filters applied). Anyway, my USB scope shows @ WaveIO's isolated SPDIF output ~1.0Vpp but was taken without any SPDIF receiver connected to it (output float). I didn't took the chance to move further as the results were already quite unintelligible. If anyone else has a better opportunity to run few measurements then it would be nice!

Hi Lorien.
Any news regarding your new interface? :) I'm very curious.
:)
Speaking of that... :p The boards are in manufacturing process right now and I'm eagerly wait to have the result. I was quite surprised to see the layout when googling for pictures of WaveIO's implementations :) Meanwhile I made few changes to it but nothing too serious. Be sure that when I'll have it fully functional I'll make a new thread for it here.

Kind regards,
L
 
S/PDIF Output Voltage - Attenuator

LORIEN - thanks for trying. So it might be 0.5v PP over the standard S/PDIF spec of 0.5v? That's not a criticism, but I was keen to try an attenuator and if the output was say 0.5v or less, then there is a danger of reducing the voltage to unacceptable level.

DEMIAN - if we assume an output voltage of say 1.0v PP, would you recommend what value attenuator I should buy to bring it down to 0.5V?

Many thanks both.

Jonathan
 
Hi Lorien,
Can you let me know if there will be many issues getting I2S working with the Najda board ? I am looking for a USB I2S solution for Najda.
I believe you and Nick are working on a soultion.
Regards Peter.

Hi, the issue with I2S overload of DSP_1 overload on Najda board has been solved by Nick! I am running it now without any problems:)

Regards

Sigi
 
Hi - I am interested in trying the Najda board with my WaveIO. Would appreciate your feedback and review of SQ and ease of configuring the Nadja.

Many thanks
Jonathan

Hi Jonathan,
I am using the Najda board as replacement of my super tweaked DCX 2496. SQ of Najda in my system outperformed the DCX easily. I had no time to make a blind test between SPDIF and WaveIO - USB inputs with Najda. My first impression was that specially the highs sounded more detailed! Najda is very easy to configure, after starting a new setup with Najda control. Most features are self- explanatory. Excuse my sometimes unusual english, it's not my mother language!:D
Regards

Sigi