Modular multi channel DIY AD/DA Box

GroupDIY Audio Forum

Help Support GroupDIY Audio Forum:

This site may earn a commission from merchant affiliate links, including eBay, Amazon, and others.
Raphael,

I'm really impressed with your dedication and progress with this, and have been reading the thread with great interest! I look forward to seeing the project completed :)

Keep up the good work!
 
Hi Raphael,
                  I note looking at the circuit there is no input decoupling, and the inputs will sit at Vbias. Have you thought of that? Are you transformer coupling the input?
 
audiovisceral said:
Does Mikkel's board allow SMUX?

Yes. Both the transmitter and receiver should work fine for SMUX also.

Best regards,

Mikkel C. Simonsen
 
radardoug said:
Hi Raphael,
                  I note looking at the circuit there is no input decoupling, and the inputs will sit at Vbias. Have you thought of that? Are you transformer coupling the input?

Hi radardoug,

well it is the idea f the project to give everyone the maximum degree of freedom. So everyone can make a choice which input decoupling is wanted (Transformer, Caps, etc.).

Raphael
 
Igor said:
Nice PCB layout. Analog and digital grounds use same symbol at schematic, but they separated at pcb, right? :)

That old recommendation of split ground planes has been hopefully discarded. Even the most recent AKM and Analog Devices data sheets say don't do that!

With proper attention to layout, your mixed-signal board with a single unsplit ground plane can be as quiet as necessary.

-a
 
HI,

I've some good news: The first AD-channel is working. Sounds ok so far. I simply connected the I2S-output of the PCM4204 to the I2S input of a SHARC evaluation board and used the DAC on this board to connect a loudspeaker to it. The ADC input was connected to s sinus wave generator and I herad a nice sinus. Checking the important places with an oscilloscope the results were ok.
Next task now to stuff the other channels on the board and a new 12 channel AD-converter is born. ;)
So I'm coming closer to the moment when I put everything into one box and mount it in a rack...  ::)

Raphael
 
If I add the SMUX parts to the OptoRec/OptoGen boards are there any testers? I don't have any "real" equipment to test with...

Best regards,

Mikkel C. Simonsen
 
mcs said:
If I add the SMUX parts to the OptoRec/OptoGen boards are there any testers? I don't have any "real" equipment to test with...

Best regards,

Mikkel C. Simonsen

Hi Mikkel

I'm equipped to test SMUX here, I'm currently running an Apogee AD-16X
through an RME ADAT card which is working great.

I'd be happy to help however I can.

Mark
 
So does anyone have an idea for a clock solution yet? I was thinking about running two 4204's, two 4104's and using the wavefront OPTO boards to interface with a lightbridge. Mikkel said that his opto boards could support S/MUX (i think) so I want to run 8/8 i/o at 24/96. If i read the datasheets correctly, i/we can run 1 4204 as "master," utilizing the SCLK input to create the BCLK and LRCLK signals. (Those pins in master mode are outputs) We can use those outputs to sync the other adc running in slave mode. ( where those pins are inputs.) The same idea applies to the 4104's...i.e. 1 master, 1 slave. Or, can we use the BCLK/SCLK output from the master 4204 to time the 4104's as well. In any case i believe the opto parts utilize a BCLK input so we can time those as well. Someone suggested a clock on a daughterboard, and as i have no other option for clock generation, this will have to become a reality for me. The datasheet for the 4204/4104 states that SCLK be a multiple of Fs. (sample rate). Giving us dual rate sampling mode, at 96kHz; 256Fs= 24.576MHz for SCLK. So, any ideas?
 
24.576MHz is the clock you want for 96kHz and 48kHz. If yer running at 88.2 or 44.1, then the clock needs to be the "standard" 22.xxxx MHz. Those clock frequencies are always shown in the converter datasheets. Yer on the right track now. I have a 4204 EVM board which I plan to work on soon based on this thread. Keep up the good work. DW.
 
TheGuitarist said:
Whats the quality of the AD conversion? Is it the same chip as some other converter out there?

Well, I think the chips do play in the same league like many professional boxes do. Eg. the ADC in a Digi 003 is a CS5361 (as far as I know) and this chip is rated at 114dB DR and -105dB THD+N. A Digidesign 192 I/O is spec'd at 118dB DR and -109dB THD+N. A PCM4204 is spec'd at 118dB DR and -105dB THD+N per channel without tricks like paralleling ADC-channels. These numbers just to give you an impression. Don't forget it is not the chip only the gives you a clean sound. You have to look on the complete circuit.

Raphael
 
Completely right, Raphael. It is much more critical what happens before the converter chip and in the clock. I have no doubt that a very professional AD can be done with the chips mentioned.

Mikkel, great you joined the thread and generously offer help for an s/mux extension for your circuit!
Unfortunately I might not have the time to be a good 'guinea pig' this time, but I always wanted to see how this might be done with your optorec board (which is great)!!!

To be on the save side there should be an option for two crystals on the clock board, so it would be possible to get all the industry standard sample rates 44.1,48,88.2,96.

Kind regards,
Martin
 
As for the 48/96khz clock I would suggest following TCXO for low price and wide availability: http://search.digikey.com/scripts/DkSearch/dksus.dll?Detail&name=535-9318-1-ND
Corresponding 44.1/88.2 clock would be a 11.059MHz quartz, due to lack of 22mhz availability from the series at digikey.
http://search.digikey.com/scripts/DkSearch/dksus.dll?Detail&name=535-9308-1-ND
I guess 20ppm might fit the task well. These quartz's already have been used in some audio clock upgrade stuff with good results.
Hm, for clocking DA I would still prefer src upsampling and using only the 96khz internally, so just one quartz...

Kind regards,
Martin

PS: SORRY, I mixed the series up, these are the TCXO's at +-2.5ppm, they are only slightly more expensive i.e.quite inexpensive related to others or the 'voodo'<=1ppm modules:
http://search.digikey.com/scripts/DkSearch/dksus.dll?Detail&name=535-9618-1-ND
 
Here is a pic of a 12 channel ADC-board. Fully stuffed except the input attenuator because I need to get the potentiometers with the next order from another distributor. For testing I replaced it with a fixed setting on an extra board.
I'm now putting everything into one box and doing some tasks like a master clock for it.

Raphael
 

Attachments

  • DSC_0371.JPG
    DSC_0371.JPG
    241.6 KB · Views: 404
Looking great  ;)
Fully stuffed except the input attenuator because I need to get the potentiometers with the next order from another distributor. For testing I replaced it with a fixed setting on an extra board
In which stage is potentiometer? I can't see it on schematics or I missed something?
Regarding sound, do you plan to add some sound samples ?
Thanks for constant progress on this and stay sharp  :)
 
rkn80 said:
Input stage, P101 etc. that are potentiometers, aren't they? ;)
I imagine that as a trimm . Sorry.. BTW I don't see any reason for input pot on the ADC stage (just adds some noise)  but it's OK if U like it  ;)
 

Latest posts

Back
Top