deriving i2s signal from the emu1212 for the buffalo dac
Jan 18, 2009 at 12:32 PM Thread Starter Post #1 of 5

minivan

100+ Head-Fier
Joined
Jun 15, 2007
Posts
473
Likes
11
the em1212 sound card consist of 2 cards, digital (1010) and analog (0202),
both are linked by a ribbon cable.
i had read in the past about signals sending from the digital and analog card are i2s, since i can't find any info on any forum about tapping the i2s signal from the emu1212, by getting some ideas from this thread
http://www.head-fi.org/forums/f46/ea...d-card-242285/
i had mapped out the i2s signal from the emu-1212 sound card.
on the analog card where the dac chip reside, it's a cs4398 chip, by looking at the data sheet and using a multimeter , this is how i derive the i2s signal
ribbon cable socket(pin5) to dac (pin 3) SDIN (serial data in)
ribbon cable socket(pin 8) to dac (pin 5) LRCK (word clock)
ribbon cable socket(pin 13) to dac (pin 8) DGRND (DIGITAL GROUND)
ribbon cable socket(pin 10) to dac (pin 4) SLCK (serial bit clock BLCK)
ribbon cable socket(pin 9) to dac ( pin 6) MCK (master clock) (no need for buffalo dac)

the ribbon cable on the emu1212 is similar to the one on the floopy disk, so i just took one from an old pc, cut it at one end and soldered it onto some shield audio cables. according to the manual of the buffalo this is how u connect the i2s signal

from sound card to buffalo
cable 5 to D2 on buffalo
cable 8 to d1 on buffalo
cable 10 to dck on buffalo
cable 13 to g on buffalo


i used foobar kernal streaming playing flac,
for emu patchmix software setting , i use
wave 1/2 to out 1/2 i/o card ,44.1khz

if someone know a better way please let me know, i am still not very familiar with patchmix

the sound quality does sound better then spdif, just like other said it has better air and space between instrument.

i still have some problems with this setup, when i try run 96khz, all i get is pop and click, it seem like the buffalo does not like it.

i measured the voltage reading of the :data, word clock, bit clock, all giving me a reading of about 3v AC, and about 1.5 v DC, what does the 1.5 dc reading mean? is it good to run dc signal into the buffalo dac?

other info i found out:

audio data come out from a chip lcx244, this is a buffer chip.

there's a receiver chip called ak4112, this is a receiver chip, looking up the data sheet it seem like it can run in slave mode, so may be there's a possibility of running the emu1212 digital card in slave mode with the dac on master mode?

now that the analog card 0202 is not being used, it's possible to made a dac out of it by using a receiver, psu ,etc.
 
Jan 18, 2009 at 2:35 PM Post #2 of 5
hi minivan,
i am using juli@ I2S with stunning results.
i have 2 spare EMU 1212 and one question:
How You are counting the Pins ?
is the red cable No. 1 ?
Thanks
alfred
 
Jan 18, 2009 at 6:52 PM Post #3 of 5
Yes, the red cable is №1.
Doesn't your patchmix software complain about absence of 0202 unit? Taking I2S signal off my 1010 is what I'm thinking about since 0202m is gone. Think 0200NOS.
 
Jan 18, 2009 at 10:11 PM Post #4 of 5
i can see the led of the 0202 light up in the patchmix, it was fooled into thinking the dac is the 0202

make sure you guys used well shielded cable and ground all the i2s signals.
if any1 want to try this and have an oscilloscope , would be great if they post some pic of the signal coming out of the 0101.
a cheap multimeter is all i have atm.
 
Jan 18, 2009 at 10:30 PM Post #5 of 5
Quote:

Originally Posted by minivan /img/forum/go_quote.gif
i can see the led of the 0202 light up in the patchmix, it was fooled into thinking the dac is the 0202


Great. On my Linux machine I can set every output I want not even having a single DAC or ADC connected. It's good that Windows systems work with I2S as well.
Quote:

make sure you guys used well shielded cable and ground all the i2s signals.
if any1 want to try this and have an oscilloscope , would be great if they post some pic of the signal coming out of the 0101.
a cheap multimeter is all i have atm.


I2S output isn't very interesting. You just get standard sequences - see CS4398 datashet. As for shielding you'd better be aware of capacitances it adds - it may degrade the signal to unreadable. I believe that the standard cable is - there are lots of ground wires. Not to say that similar cables have been used for transferring much more data than digital audio output requires.
BTW, I'd use an internal "0200". Additional supply filtering and some optional shielding may be enough.
 

Users who are viewing this thread

Back
Top