Drive NOS AD1865/62,PCM1704/02/63,TDA1541 from FIFO: Universal I2S-PCM driver board

Jump to Latest
iancanada

iancanada

Member
Joined 2009
twitter.com/iancanadaTT
Paid Member
apoo said:
Hi Ian,

is it possible to use your I2S to PCM board to drive this DAC? http://www.dddac.com/documents/dddac1794_nos_ver30.pdf

The schematics of the actual board is here:
dac_delay_large.png


apoo
Click to expand...

iancanada said:
Hi apoo,

Nice project!

From the spec, PCM1794 is not a traditional R-2R DAC, and it accept right justified format which is not the 'PCM' format. So, I don't think you need this I2S to PCM converter.

It's very nice you run it at NOS mode :).

Ian
Click to expand...

vzs said:
PCM1794 is configured for Digital filter bypass - Mono (see specs Table 2.)

He has to delay both channels in order to have 24bit right justified of both: delay right channel by 7 cycles and left by 39 cycles to have the LSB bit of both data channels clocked in when right channel's latch goes low.
31-7 = 24
63-39 = 24
(31 and 63 because I2S format has data delayed by 1 cycle).

Right-channels's I2S latch signal (LR) latches both L and R DACs.

Having reviewed how that works I'm 99% sure I2S-PCM board can be used to bypass Doede's delay line.

But why on earth use PCM1794 when you can use any good old R2R DAC with this I2S-PCM board?

Zsolt
Click to expand...

Hi Apoo,

Maybe I didn't get your point. Running PCM1794 at NOS mode is a special configuration, can your post the timing to drive it?

Here is the principle waveform of I2S to PCM board (half speed mode), you can figure out if it is working for your application.

Ian
 

Attachments

  • Principle2013-08-07.png
    Principle2013-08-07.png
    54.8 KB · Views: 1,237
P

PET240

Member
Joined 2010
vzs said:
But why on earth use PCM1794 when you can use any good old R2R DAC with this I2S-PCM board?

Zsolt
Click to expand...

Cause I don't have one......actually I do have a TDA1541 KillerDAC (google) coming, also have a Teradak Chameleon 16 chip TDA1543 Dac as well, heavily modified. I would like to play with something computer audio related so I can fiddle with higher sampling rates. Also have a Buff III to build a tube output stage for.
Yes I'm one of those tube guys.....

Appreciate you looking into the timing thing Zsolt, much appreciated!

Cheers,

Drew.
 
Last edited:
dvb projekt

dvb projekt

Member
Joined 2009
Paid Member
Hi Ian,

because i use two Red Baron´s parallel, i added two additional U-FL connectors for DR and DL.


An externally hosted image should be here but it was not working when we last tested it.




Now the digital input section is completed!


An externally hosted image should be here but it was not working when we last tested it.



Wave I/O for USB input --> SPDIF/AES input board --> FIFO --> Isolator board --> Dual Clock board with Crystek CCHD-957 clocks (powerd exclusive from a A123 Systems LiFePo4 cell) --> I2S to PCM V2.0 Module --> Red Baron

What i could say after 2 hours of listening, i will never go back to the I2S input! :D

All of you who have pre ordered the board will have MUCH fun!!!

Cheers,
Oliver

:wave:
 
iancanada

iancanada

Member
Joined 2009
twitter.com/iancanadaTT
Paid Member
dvb projekt said:
Yesterday i´ve got Ian´s I2S to PCM V2.0 module to do some pre testing with my Red Baron V5.0.

An externally hosted image should be here but it was not working when we last tested it.


The simultaneous (offset binary) mode runs perfect.

Big thank´s to Ian! :wave:
Click to expand...

Hi Oliver,

Your system look great! Congratulations!

The jumper settings are correct. I think you solved all problem by now.

I was surprised how nice the TDA1541A can be when it's playing with the daughter board at OB mode.

vzs found even with half-speed mode, TDA1541A already been running over clock at 352KHz which I confirmed working. I'll try 384KHz to see if I'm reaching the physical limitation of TDA1541A.

Thank you so much for you efforts and your good news.

I'm very close to launch the board. It seems I still need to speed up.

Regards,

Ian
 
Last edited:
qusp

qusp

Member
Joined 2009
Oliver and Ian, wouldnt it be recommended to replace the 33Ω resistor network on DR/DL with small discrete resistors, so that you can have the correct termination on them with double the lines? am I incorrect, or should the line now be terminated with 66Ω if running 2 parallel on the same driver?
 
P

PET240

Member
Joined 2010
Just to ask a silly question and to be sure I understand as the search functions don't work so wel on an iPad, this board will be capable of running a pair of TDA1541's in dual mono? Ie one chip does left and the other right and output a balanced signal from the DL and DLn outputs for left and vice versa for right?

And Tazzz, I would assume that this will indeed remain proprietary. Personally what Ian does is quite amazing and is probably under appreciated by those not involved in the world of embedded IO control and custom programming. Not saying this is you, but I would estimate if one of us were to pay Ian an hourly rate to develop this item in work hours you wouldn't be short of a bill in the $10-15k mark with pcb design, programming and testing refining, bug resolution, manual writing, quality control and testing. Really is very involved and we get a bargain.
Bravo Ian, really is very appreciated!

Thanks,

Drew.
 
palmito

palmito

Member
Joined 2010
PET-240 said:
Not saying this is you, but I would estimate if one of us were to pay Ian an hourly rate to develop this item in work hours you wouldn't be short of a bill in the $10-15k mark with pcb design, programming and testing refining, bug resolution, manual writing, quality control and testing. Really is very involved and we get a bargain.
Bravo Ian, really is very appreciated!
Click to expand...
And the support!!!!! Ian's support, considering this is DIY stuff and how busy he is, is nothing short of amazing. My additional $0.02...
 
dvb projekt

dvb projekt

Member
Joined 2009
Paid Member
iancanada said:
Unfortunately I can’t find any working software based real time up-sampling apps which can go up to 384 KHz so far for both Mac and PC, any recommendation?

Ian
Click to expand...

Hi Ian,

for the MAC Pure Music could do that including DSD file playback!


  • Up to 384 kHz sample rate
  • Selectable REAL-TIME 64-bit upsampling
  • DSD (2.8 and 5.6 MHz) file playback on any DAC (real time PCM conversion)

Just update the firmware of your Wave I/O to the newest one, witch support 32/384 kHz ;) (for WaveIOs with RevisionID 0×0331 or higher)

Cheers,
Oliver

:wave:
 

Attachments

  • WaveIO_32Bit_384Fs.zip
    18.9 KB · Views: 90
Last edited:
dvb projekt

dvb projekt

Member
Joined 2009
Paid Member
I do some testing yesterday and could confirm that the I2S to PCM module runs great.

Software setup:
iMac OS X 10.8 – Pure Music v1.89g R3 - iTunes

Hardware setup:
WaveI/O with latest SW (32bit/384 kHz support) – FIFO – Isolator - Dual Clock with Crystek CCHD-957 (LiFePo4 cell powered) – I2S to PCM convertor

DAC setup:
Jumper: short JOB
2* Red Baron V5.0 (NOS TDA1541A) parallel with simultaneous mode

192 kHz file - ok
44.1 kHz file upsamling to 192 kHz - ok
DSD 2.8 MHz file converted to 176.4 kHz - ok
DSD 5.6 MHz file converted to 176.4 kHz - ok
352.8 kHz file - SI570 clock needed
384 kHz file - SI570 clock needed
44.1 kHz upsampling to 384 kHz - SI570 clock needed
DSD 2.8 MHz & 5.6 MHz file converted to 352.8 kHz - SI570 clock needed

@ Ian

Perhaps you could do some upsampling listening tests (44.1 kHz files) between
Dual Clock with Crystek CCHD-957 - 192 kHz and SI570 clock - 384 kHz.

It would be very interesting if the upsamling to 384 kHz would bring the SI570 above the Crystek... :rolleyes:

Cheers,
Oliver
 
iancanada

iancanada

Member
Joined 2009
twitter.com/iancanadaTT
Paid Member
dvb projekt said:
I do some testing yesterday and could confirm that the I2S to PCM module runs great.

Software setup:
iMac OS X 10.8 – Pure Music v1.89g R3 - iTunes

Hardware setup:
WaveI/O with latest SW (32bit/384 kHz support) – FIFO – Isolator - Dual Clock with Crystek CCHD-957 (LiFePo4 cell powered) – I2S to PCM convertor

DAC setup:
Jumper: short JOB
2* Red Baron V5.0 (NOS TDA1541A) parallel with simultaneous mode

192 kHz file - ok
44.1 kHz file upsamling to 192 kHz - ok
DSD 2.8 MHz file converted to 176.4 kHz - ok
DSD 5.6 MHz file converted to 176.4 kHz - ok
352.8 kHz file - SI570 clock needed
384 kHz file - SI570 clock needed
44.1 kHz upsampling to 384 kHz - SI570 clock needed
DSD 2.8 MHz & 5.6 MHz file converted to 352.8 kHz - SI570 clock needed

@ Ian

Perhaps you could do some upsampling listening tests (44.1 kHz files) between
Dual Clock with Crystek CCHD-957 - 192 kHz and SI570 clock - 384 kHz.

It would be very interesting if the upsamling to 384 kHz would bring the SI570 above the Crystek... :rolleyes:

Cheers,
Oliver
Click to expand...

Hi Oliver,

Good job! Do you mean you ran your Red Baron 1541A at 384KHz with FIFO KIT and the PCM daughter board? I confirmed 352.8KHz by myself. TDA1541A will be over clocked at both cases according to the spec. It seems Philips engineer left some space there for our audiophiles:).

Thank you for the files, I'll upgrade my WaveIO for more test,

Regards,

Ian
 
Last edited:
P

PET240

Member
Joined 2010
The TDA1541 should accept 176.4 happily as the 7220A/B chip 4 times oversampled with an 11.2896 clock.
So Ian with the half speed mode does that mean the clock is divided by two?
If so with an 11.2896 clock does the TDA1541 require this in half to accept the 352kHz signal?
 
V

vzs

Member
Joined 2005
dvb projekt said:
Hi Ian,

with the dual clock board and the 22.5792 MHz and 24.5760 MHz clocks, i have no output with 352.8 kHz & 384 kHz sampling in half speed mode.

I think you need different clock speed. That´s why i wrote SI570 clock needed. :rolleyes:


:wave:
Click to expand...

Insert an 45.xx and 49.xx in the dual clock board and set it to double-speed mode (read its manual) so you can play 352.8 and 384khz.

zsolt
 
You must log in or register to reply here.
Top Bottom