A S/PDIF solution for the =
DDDAC1794
There was still a lot of request for good old =
S/PDIF, so I=20
designed a "S" version...
I assumed high res audio was only about PC and Mac, =
so I=20
never bothered to implement S/PDIF right in the =
beginning. But=20
I learned that there are still quite some CD Players =
in use or=20
other media players (or PCs which use S/PDIF output). =
As=20
S/PDIF is not that difficult to implement, I took this =
up and=20
started designing again. As you know the KIT is =
already=20
available for some time. I just had no time to update =
the Web=20
Site ;-)
These were the design requirements I wanted to =
implement in=20
the new mainboard:
- It needed to be fully compatible with the "USB =
only=20
mainboard" and the DAC Modules, electrically and=20
mechanically=20
- The S/PDIF functionality=20
- Use best available technology (is a bit =
arbitrary of=20
course ;-)=20
- Sampling rates Up to 192kHz (Optical 96kHz)=20
- Option for a 12MHz XO Clock (like the ones =
from Guido=20
Tent)=20
- Easy to switch between USB and S/PDIF source =
- Improve a few things on top:=20
- LED signal part can be broken off for use at =
the front=20
panel of the chassis=20
- Improve buffering of signals to be able to =
drive more=20
boards (so far 11 is done)=20
- Improve timing of I2S signals with flip flop =
in stead=20
of the 1k resistor solution for data-delay=20
- Make logic voltage 5 Volt for higher =
compatibility=20
inside the design
The only important selection was the receiver =
itself....=20
There are a few choices, like DIR9000 CS8416 and the =
Wolfson=20
WM8804. I have been listening around and for arbitrary =
reasons, I thought the Wolfson WM8804 would be the =
best choice=20
for me. I bought a few chips, made some breadboard =
tests and=20
voila it woked within a few weeks. Let me explain a =
few things=20
I did in the new mainboard...
You can =
download the=20
full schematics (S/PDIF and USB mainboard version) =
from the=20
download section or push the button below:
Theory of operation of the S/PDIF mainboard
First part was the implementation of the WM8804. It =
is done=20
of course in hardware mode. In the schematic below you =
recognize the option for the XO clock. Further more =
there is a=20
set of resistors, which are used to configure the =
WM8804 at=20
start up. Of course I need to Wolfson chip to output =
I2S=20
signals based on S/PDIF input signal. All easy, after =
studying=20
the somewhat complex data sheet..... In the circuit =
clip below=20
you see I have chosen to separate the power supplies =
for the=20
chip to maximize performance. In case of a Tent Clock=20
installed, the two ceramic Cs and Crystal should be =
removed=20
(or not installed). The 75 Ohm resistor (R1) is not on =
the=20
mainboard. It is meant to close the impedance of the =
coax=20
cable, so best position is directly at the RCA=20
connector...
<=
/A>=20
I had two problems which needed to be solved: =
First.... The=20
WM8804 is outputting 3,3 Volt logic and I wanted to =
control=20
the DAC boards (and WaveIO) with 5 Volt logic. =
Secondly I=20
wanted to be able to switch between USB (WaveIO I2S =
signal)=20
and the S/PDIF I2S signal.
After some reading and checking I came up with the=20
following solution. The 74VCHT244 is a bus buffer =
which=20
converts 3-5 Volt logic to 5 Volt logic. As it can go =
in=20
tri-state (high impedance on the bus) I saw a kind of =
two way=20
switch in front of me :-) Have a look at the circuit =
clip=20
below and you will understand the beauty of this =
solution.....=20
- The inputs are switched without use of =
mechanical=20
switches - The strong output buffers can have the =
clock=20
signal drive more DAC boards (the USB only board can =
do max 4=20
DAC modules) - All is 5 Volt logic now at the =
output=20
=20
The final difference with the USB only mainboard is =
the=20
improved data delay. Remember? The Data needed to be =
delayed=20
half a clock cycle. I did that with a 1k Ohm resistor =
at the=20
data input. I was not sure how well this worked at =
multiple=20
boards (now 11 have been tried ;-) so as I was =
designing, I=20
simply implemented a standard flip flop to delay the =
data=20
signal with an extra half cycle. This means, that when =
data is=20
being clocked into the DAC, the state is rock solid at =
clock=20
transition... =20
<=
/TD>
| The Board itselfCompared to =
the "red=20
USB only board, you can easily recognize the =
section,=20
where the WM8804 chip is and the extra power =
supplies=20
(3,3 volt)=20
At the Left is now the connection for the LED =
indicators. It can be easily broken off for =
front panel=20
use, or just let it stick of course
This example shows a 4 deck version. Several =
8 Decks=20
have been implemented without any drive issues. =
I know=20
of two cases where 11 boards were stacked with =
great=20
success !
As mentioned, the board is fully compatible =
with the=20
old DAC boards, as well the somewhat revised =
second=20
version of the DAC module, which is now being =
delivered.=20
You can freely use them through each other. So =
easy of=20
DIY is guaranteed :-) |
The=20
DDDAC1794S Next to S/PDIF functionality, =
several=20
practical improvements=20
|
Keep reading ...... the story continuous with =
installing=20
all software ( Programs=20
and Drivers )
|