Digital audio from IDE

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
I can't provide a schematic just yet, but here's what would be involved in such a project:

- decently fast microcontroller (ATMega, single-clycle 8051, etc) which connects to CDROM's IDE bus and performs audio read commands.
- clock generation hardware.
- I2S generation hardware, probably a XC9572 or similar CPLD, which accepts parallel audio data from the microcontroller and an audio clock and shoots out serial I2S.
- digital audio transmitter which converts I2S to SPDIF/AES/toslink/whatever.

Actually. a Microchip dsPIC chip has internal I2S generation hardware and likely has enough GPIO to handle an IDE CDROM.
 
If I recall correctly, a lot of the CD-ROM drives I have had in my hands, have a digital audio out connector (located next to the analog out). I do not know the format, but it must be easier than using the IDE channel.

Is there any particular reason for not using this digital audio port?


Jennice
 
Is there any particular reason for not using this digital audio port?


The ttl level spdif cannot be of very high quality. There are all sorts of compromises involved not the least of which is a dubious ceramic oscillator. Of course some cdrom chipsets allow non-oversampled I2S which is much better. Getting audio out of the IDE channel is certainly not trivial.
 
ghetto said:
Does anyone believe there would be any benefit to getting the audio data via IDE if there are already I2S signals available on your particular CDROM?
You can use your own clock - an "IDE-to-I2S" jig can generate its own high quality clock, or use one from an external DAC. The drive will spit out data on demand, it doesn't necessarily have to be at an "audio speed" established by the drive.

It may be possible to hack a CDROM to operate in I2S slave mode... it'd be a fun adventure to open up a drive, look up the datasheets for the chip used, and see if you can get that to work.

But if this works and then your drive dies or wears out, then you're repeating the datasheet hunt once again unless you get the exact same drive to replace it... There's another advantage of a jig - it should work with any cdrom/cd-rw/dvd drive you plug into it.

I'm contemplating doing a design of this now. :D
 
Personally I think an IDE based controller would be awesome. If people are serious about designing such a beast, I'm pretty handy with VHDL, PIC and AVR and could write and test code. Further I think a PCB that included whatever controller is settled on, plus high quality power supplies, user interface (buttons, LCD, RC5), and maybe even a popular DAC would make a great basis for a group buy.
 
ok i am going to follow this thread now too. BUt to say i have 2 cd players right now. one being a cd rom drive from a computer and the other one kinda the same but a manufactured dvd player in side there is a SMPS psu and a DVD rom and a screen ect ect then a dac board. Now when compairing the 2 the cd rom based one with NO IDE conversion sounds way better but useing the other one i can hear the difrence for sure. it spins the cd at hi rpm and converts the digital audio to non digital.. humm is it really worth it to use a cdrom and convert that to digital then to non digital ?
 
jleaman said:
ok i am going to follow this thread now too. BUt to say i have 2 cd players right now. one being a cd rom drive from a computer and the other one kinda the same but a manufactured dvd player in side there is a SMPS psu and a DVD rom and a screen ect ect then a dac board. Now when compairing the 2 the cd rom based one with NO IDE conversion sounds way better but useing the other one i can hear the difrence for sure. it spins the cd at hi rpm and converts the digital audio to non digital.. humm is it really worth it to use a cdrom and convert that to digital then to non digital ?


Errrr, the data on the disc IS digital. There is no conversion.
 
I think a microcontroller such as Mega32 would make a great candidate for this project. Its fast, pretty cheap, available in both SMT and DIP ... plus the free excellent GCC compiler can make the code managable, readable, and portable. I think this chip is probably fast enough to clock out the I2S data by itself, but I admit that I havent run the numbers to know for sure. Any comments?
 
MCLL said:


Thanks rfbrw remind me. Do you have any update ATA Packet Interface for CD-Rom?

If you mean a design like the ones in that looong CDROM thread , then the answer is no. I see no point in yet another design of that type. If you mean a design that uses the IDE port to extract data then the answer is an even bigger no. There are some serious obstacles in the way of such a design.
 
ghetto said:
I think a microcontroller such as Mega32 would make a great candidate for this project. Its fast, pretty cheap, available in both SMT and DIP ... plus the free excellent GCC compiler can make the code managable, readable, and portable. I think this chip is probably fast enough to clock out the I2S data by itself, but I admit that I havent run the numbers to know for sure. Any comments?


At what speed do you intend to extract data off the disc?
 
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.