DAC with DIR9001 & PCM1792A (fist digital project)

dock

Member
2008-02-21 9:33 pm
DAC with DIR9001 & PCM1792A (fist digital project)

Hello everybody,


I'm an real newbie/rookie to digital circuits.
I'm trying to build a DAC myself with following components:

- DIR9001 from TI (digital audio interface receiver)
- PCM1972A from TI (24-bit, 192KHz sampling, advanced segment DAC)
- NE5534 from TI (audio output opamp's)
- LT1028 from LT (ultralow noise opamp's)

for input:
- SN74HCU04 from TI (hex inverter)

for powersupply:
- TL780 series from TI
- µA78M00 series from TI

I have tried to do some homework:
I have designed (actually stolen most of the parts
;) from others.

But i still have some questions before i start to make a print. And probably there are some (huge) mistakes in my circuits. (and in the brain too).

Let's just start with what i came up with.
(little note, in the main circuit , everything is hooked up the way i think it shoud.)

--> Powersupply:

5V, 3.3V analog and 3.3V digital.

http://dockonline.110mb.com/powersupply.JPG

First question about this circuit, are there mistake's? Can i connect the grounds together? or not? (with a ferrite bead?)

--> Input circuit:
from SPDIF
I stole this one from Heiko Purnhagen, thank you for this! (
link)

And i made this link

One question already here. are those capacitors
(10 and 0.1 µF )correct to "isolate" (not correct english) the ic from other ic's ?

--> Reciever circuit:

Basically out of the PDF from TI. link to pdf from DIR9001

i made this out of it: link
questions here , are digital and analog ground directly connected to each other?

--> The DAC circuit:

Also most of it taken from the pdf form TI. link to pdf

my version of the story: link


Putting it all together , the main circuit ((Best viewed on a 110 inch plasma screen) (with 8.1 surround))

What do you think of it? My main concern's are that i don't understand all the functions of the pins of the DIR9001 and PCM1792A. I hope they all would autodetect input characteristics.

Will it work? Are there mistakes, and where? Are there easy improvements?
I'm still working on a +- 15 voltage supply for the opamps.

Anny remark is welcome !
Thank you verry mutch for your time and patience. I will work the project further out with all your remarks...

Best regards,


Dock
 
Don't use the summing circuit shown after the i/v converters with the PCM1792 use this instead:

http://www.diyaudio.com/forums/showthread.php?postid=1444038#post1444038

Read the last few pages of the zhaolu thread, suffice it to say the implementation shown in the app notes doesn't sound very good.

The Zhaolu uses a similar input receiver to the one you show, my big concern with this design is that the input impedance may be influenced by the available open loop gain since this thing is being used as an inverting amplifier. Higher gains result in a lower impedance shunting your 75 input resistor. I'm a bit distrustful of this topology. Hopefully jockohomo can weigh in on this.

You can also use the PCM1794 or PCM1798 which both operate in hardware mode and don't require a cpld or micro-controller to work properly.

Also the 5534 family of devices while nice sounding in no way represent anything close to SOTA for op-amps. I'd look at the LM4562 and its siblings.. Also BB2604/2107/2132 are worth considering.
 

dock

Member
2008-02-21 9:33 pm
Thanks for the responses already !

Controller sounds pretty difficult, so i'm planing to use the PCM1794.

Leon, are you so kind of sharing your design with me/us?
I have seen your printlayout on your website, but it isn't easy to read all components.

Also i'm considering the other opamps.
I'll read that topic this weekend. (have to work(hard) all week !)

Thank you all !
 
FWIW, and I am not totally sure about this, but I suspect the diode from the input circuit back to the input is reversed. Simulations indicate it works much better that way, but overall I'm uncomfortable with this circuit - I think a high performance comparator might just be better.

The Zhaolu 2.5A I have been tweaking uses a very similar input receiver, and I suspect its performance leaves something to be desired particularly at higher sample rates where the performance of the dac seems to degrade. This is idle conjecture, until I get in there with my scope and look around I can't say for sure.
 
leon said:
I'm still expirimenting so I'm not confident enough to publish my design.

If you noticed I use 74HCU04 but I also included a voltage comparator to expiriment with in the input stage.

You might want to consider sharing it anyway, some of us might see things that you can improve and we might also learn something as well.

Somewhere around here is my old dac thread, the dac unfortunately was stillborn.. I'm currently tinkering with the Zhaolu and buying some boards from Twisted Pear Audio. Their spdif receiver module uses the Wolfson WM8804 spdif transceiver module which is both a rather good receiver and includes a transmitter module - it can even be used for reclocking the spdif data stream if desired.