Yes, it's possible. Most of you audiophiles listen to music, but some of us play games too. 🙂 The gamecube outputs digital audio. Data, serial, and LR clock is outputted through the digital AV port. Some smart Japanese person figured this out (site is in japanese, altavista will not translate it for some reason) and used a TC9231N Digital Audio transmitter from Toshiba with a 12.2880MHz clock to produce an S/PDIF signal. I have redrawn the design with a CS8405Am but the only thing I need to know if the serial audio output is I^2S, Left or Right Justified, but since I cannot find the datasheet anywhere for the TC9231N (I emailed Tosiba Semi but they haven't gotten back to me yet) I am unsure and the only way I'll figure it out is with trying all 4 serial input formats the CS8405A will operate with. If anyone has the datasheet for the TC9231N, PLEASE let me know. The actual S/PDIF circuit is here that uses the TC9231N. I just need to figure out the output format based on how the serial format pins are wired. If anyone has it, let me know.
🙂
🙂
Just wondering if there has been any update on this, or if anyone has any more idea. Very tempted to try this (especially if it is an I2S output) but would ratehr get an idea of chances of success first!
Well, I did get farther. Toshiba faxed me the datasheet, but the circuit with the Crystal part never worked. I did something wrong, but haven't figured it out. Pioneer has the Toshiba part, but for $43. I ordered the part and was forced to eat rice and generic pasta for a week. I built the circuit, using a $56 SDIP to DIP adapter that never really put the pins in a DIP footprint, just something screwed up. I built it according to the schematic. I connected it up and I was suprised to hear a wonderful sound comming from my DAC. I played Wario Land 4 and it sounded great. The circuit uses little power. The regulator feels room temp. I added coax output as well as RS485 output, and even with these extra circuits (they don't draw much anyway) it still stays cool and still sounds great.
Granted, this converter cost me around $150, but it was well worth the money. It runs well, sounds good, and is very reliable. The designer, oscar, is a genious.
The RS485 output I was talking about is the TTL S/PDIF signal connected to an RS485 transceiver configured as a transmitter. The type of connection I use with it is standard straight through phone cable. It sounds great actually. My DAC has another RS485 tranceiver configured as a receiver. It outputs TTL. Sounds great
Granted, this converter cost me around $150, but it was well worth the money. It runs well, sounds good, and is very reliable. The designer, oscar, is a genious.
The RS485 output I was talking about is the TTL S/PDIF signal connected to an RS485 transceiver configured as a transmitter. The type of connection I use with it is standard straight through phone cable. It sounds great actually. My DAC has another RS485 tranceiver configured as a receiver. It outputs TTL. Sounds great
This is neat. If anyone's got a scope, please probe what you see on the A/V connector and figure out a couple things:
- wether the I2S audio output is 16 or 32 bit (bck=32fs or 64fs)
- as asked before, wether the audio output is I2S, left justified, right justified, etc.
- edit: and clock polarity! find out of LRCK and SDATA changes on the rising/falling edge of BCK.
With this known, it'll be straightforward to make something like this work with a TI DIT4xxx or Cirrus CS840x chip - I'll throw a design together if requested. Only thing is, the 128 or 256Fs signal has to be phase locked to the incoming BCK or FS - which can be done with a PLL/VCXO, or even a cheap integrated IC like an ICS525.
I'm not sure if the Toshiba chip worked without phase locking out of coincidence or if it's an actual 'feature' of that chip.
- wether the I2S audio output is 16 or 32 bit (bck=32fs or 64fs)
- as asked before, wether the audio output is I2S, left justified, right justified, etc.
- edit: and clock polarity! find out of LRCK and SDATA changes on the rising/falling edge of BCK.
With this known, it'll be straightforward to make something like this work with a TI DIT4xxx or Cirrus CS840x chip - I'll throw a design together if requested. Only thing is, the 128 or 256Fs signal has to be phase locked to the incoming BCK or FS - which can be done with a PLL/VCXO, or even a cheap integrated IC like an ICS525.
I'm not sure if the Toshiba chip worked without phase locking out of coincidence or if it's an actual 'feature' of that chip.
emuman100 said:
I pasted the link into Google's search bar and it offers the translation if that is still helpful for you.
Google Translation
Sorry to resurrect an old thread...
Hey, all, sorry to ressurect an old thread, but I'm interested in building one of these circuits for my gamecubem - I've found all of the chips and regulators except for the crystal - if anyone knows where to find one of these, I'd appreciate it ;D
THANKS!
Hey, all, sorry to ressurect an old thread, but I'm interested in building one of these circuits for my gamecubem - I've found all of the chips and regulators except for the crystal - if anyone knows where to find one of these, I'd appreciate it ;D
THANKS!
- Status
- Not open for further replies.