|
|
|||||||
| Home | Forums | Rules | Articles | Store | Gallery | Blogs | Register | Donations | FAQ | Calendar | Search | Today's Posts | Mark Forums Read | Search |
| Digital Line Level DACs, Digital Crossovers, Equalizers, etc. |
|
Please consider donating to help us continue to serve you.
Ads on/off / Custom Title / More PMs / More album space / Advanced printing & mass image saving |
|
|
|
Thread Tools | Search this Thread |
|
|
#1 |
|
diyAudio Member
Join Date: Jul 2004
Location: colorado
|
Hi everyone. As many of you probably know, PS Audio is one of the manufacturers out there supporting I2S digital output/input format as a better alternative to S/PDIF.
The I2S format has separate clocks and data and lower jitter and better performance if handled properly. I2S is the native format inside every CD player. If you are using a separate transport, the native I2S inside the transport must be converted into S/PDIF format to get out and into your DAC. This is not the best way to do this. When we released the PerfectWave DAC and Transport a year ago, we decided to include I2S as a standard interface between the two machines. We also decided that the best way to transfer the I2S data was over an HDMI cable. There are many great HDMI cables available on the market and it's an excellent interface cable. Our engineering department spent a great deal of time developing a "standard" that works quite well and several other companies have already adopted this standard. Our hope is that more and more manufacturers will include a similar interface on their devices. I figured the DIY guys would be the logical choice to get this ball rolling so I have attached the schematic for the interface. You are free to use this in any way you see fit. I would appreciate a bit of publicity on this, but nothing more. If you have questions, direct them to me paul@psaudio.com and I'll do my best to pry someone out of engineering to help. Good luck and have fun. Drop me a note if you do this. The results are noticeably better than running through S/PDIF or AES/EBU. Paul |
|
|
|
|
#2 |
|
diyAudio Member
Join Date: Jun 2008
|
Nice of you guys to share....!
jk |
|
|
|
|
#3 |
|
diyAudio Member
Join Date: Jul 2004
Location: colorado
|
Our pleasure. Always happy to share.
|
|
|
|
|
#4 |
|
diyAudio Member
Join Date: Nov 2004
Location: Central Ohio
|
how is the I2C bus used in the link?
|
|
|
|
|
#5 |
|
diyAudio Member
Join Date: Jan 2008
Location: Virginia
|
That's nice, but it wold be more usefull if it had capability for surround sound too.
IMO the clocks should be combined in just one, there is no need to dedicate 3 separate differential channels for that. Other than simplify the schematics that is... I would use the 'Clock' pair for general clock and the 'Data' pairs for the 3 audio data necessary for 5.1 surround. |
|
|
|
|
#6 |
|
diyAudio Member
Join Date: Nov 2004
Location: Central Ohio
|
The I2S bus needs all four of those signal pairs. The MCLK is 256X the sample rate, The LRCLK tells you left or right channel 1X sample rate, and the bit clock is 64X the sample rate.
That one could possibly be derived from the MCLK. So, you would still need a minimum of three. It would be better to add another HDMI cable for multichannel. |
|
|
|
|
#7 |
|
diyAudio Member
Join Date: Jan 2008
Location: Virginia
|
You need just the bit clock, the rest can be derived (XTAL clocks, triggered by BCLK in the other end for MCLK), eventually with some encoding for the LRCLK in the stream...
There are only 4 pairs in a HDMI or in a CAT5E cable. If some could work around the three clock issues... that would be awesome. But not on the RIAA "guidelines" I guess. Last edited by SoNic_real_one; 7th April 2010 at 04:44 PM. |
|
|
|
|
#8 | |
|
diyAudio Member
Join Date: Nov 2004
Location: Central Ohio
|
Quote:
![]() now you're designing a new interface that's not I2S format ![]() if each sample has 24 bits of data, then you do have 8 trailing bits available in each word that could identify the channel. and a few other parameters. |
|
|
|
|
|
#9 |
|
diyAudio Member
Join Date: Jan 2008
Location: Virginia
|
No, I was thinking in leaving the data untouched, but encoding differently the clock.
Something easy to do at DYI level like voltage level shifting for L/R channels. After all is a low freq clock. MCLK I guess is easy to accurate regenerate from bit clock? PLL maybe? Hmmm, sorry, didn't want to hijack this thread. I'll think about it in a month, now I take some exams
|
|
|
|
|
#10 |
|
diyAudio Member
Join Date: Jul 2004
Location: colorado
|
I'd be careful of this approach. PLL's introduce jitter and are not desirable if you can avoid it. What we have shown here works extremely well.
|
|
|
| Thread Tools | Search this Thread |
|
|
|
|
||||
| Thread | Thread Starter | Forum | Replies | Last Post |
| PS Audio PS Elite Integrate Amp | jimmy50 | Solid State | 14 | 5th May 2012 06:07 PM |
| Audio Standards | korneluk | Tubes / Valves | 4 | 7th August 2008 04:16 PM |
| Bad standards are better than no standards | Aengus | Analogue Source | 16 | 13th December 2006 11:56 PM |
| Freeview Audio standards | AndrewT | Digital Source | 1 | 8th September 2005 09:23 PM |
| RIAA Audio Standards | TJtsj | Analogue Source | 24 | 13th July 2003 04:31 PM |
| New To Site? | Need Help? |