Non Oversampling DAC-complementing CD-PRO

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
After succesfully finishing my CD-PRO, next obvious step is to add a DAC using I2S. My first choice is AD1865 (after reading many comments from Elso.;) ). I finished a DAC before, based on BB1704 chips and the results are really impressive.

Since AD1865 is not directly I2S compatible I have to implement AD application note found somwhere on a forum (unless somebody did it already;) )

Any suggestions for I/V stage? If not, I'll probably use Pass Dac 1 output stage. Since I want to place everything directly under transport I don't have much space to deal with tubes (unless I provide auxilary output and use tubes outside the player).

The DAC will be truly balanced and chips will be parallelled. Guess I have enough?;)
 

Attachments

  • ad.jpg
    ad.jpg
    32.4 KB · Views: 7,509
Well, if i can talk about any DAc, i can about an PCM 63-K.
In 1991 i had a well designed one and i had to give up the
illusion it still is top of the art.

Even the pulse ones from Sony are superior meanwhile (most in bass response)

There was a picture-disc player that outperformed it long ago.
It was a cheap 1541 design - not claimed High-End - but better!

I don´t know the 1702, but the PCM 63 is a waste of time today i believe.


Wombat
 
NON-OS AD1865 DAC

Hi Peter,
If you want to implement the AD1865 in the CD-PRO you do need glue logic. I am not so happy with AD application note. The most simple circuit; minimal component count; is with a shift register. 74HC164.
If you wish to construct a separate DAC component you could use the CS8412. The latter chip can be put in mode 6, 18 bits output mode and you only need a inverter to get a inverted LATCH signal.
After much experimenting I gave up the idea of a balanced DAC with the AD1865N-K, simply because it sounds better non-balanced. Also found that of the two methods to get separation into left and right audiochannnel: stopped clock technique and inverted LATCH; the latter gave the better sound. I did not hear the time difference or delay between the channels that is obvious with inverted LATCH.:)
 
I wouldn't like to use CS8412 (although I have one), because in my opinion the whole advantage is to use I2S interface and feed it directly to the DAC. Best sound, is what they say.;)

As to PCM1702, I've built already a DAC implementing parallel 1704 chips, using Dave's board. The sound beats my air bearing TT.;)
 

Attachments

  • dac.jpg
    dac.jpg
    94.3 KB · Views: 7,081
The circuit posted does not address the interchannel delay and a balanced dac built without addressing the interchannel delay is unlikely to perform at its best. If the stopped clock method of addressing said delay is not to your liking, one can always go down the large shift register route keeping the continuous bitclock.

ray.
 
Interesting schematic.

But maybe I'm not seeing it correctly.

Looks like you are shifting SDATA 16 bits (2X 74HC164) so that the left channel's LSB and down-going edge of the inverted LATCH coincide. From the above I assume that it is for use with 16 bit I2S. I do believe there is a 24 bit variant, too.

But the AD1865 has an 18 bit input.

Won't the AD1865 still accept the last 2 LSBs intended for the opposite channel and add them above the MSB of the channel before the LATCH is enabled?

Don't you have to stuff an extra 2 LSBs for each channel, in there somewhere, so the 1865 has an 18 bit input register full of good 16 bit data? :confused:

Steve
 
If the CD pro has.....

SAA7372 chip. You should be able to useit in the EIAJ mode. Bypassing the I2S issue. If I remember there's a CDrom EIAJ mode that doesn't use the onboard interpolation. It's been a few years since I've played with the Pro kit. Just something to consider. CDROM/EIAJ mode with a inverted latch, for NOS cd playback?
 
Peter,

I am working on getting two TDA1541's in dual mono. That is one does left and one does right channel output. Output can be balanced by inverting one of the channels or not. In the latter case both outputs can be paralleled. This is more or less the same then as many TDA1541 designs with two in parallel. Only now left and right channel can have their own powersupply..

I posted this a while ago in this forum. Used quite a few shift registers then, but it is working fine! This is going to be only one in the future. Chipcount: 74HCT74, HEF4517 (replacing eight 8-bit registers) and a 20V8 GAL. Not too bad i think.

Maybe something to look at if you go down the TDA1541 route.
You have to check the output of the PRO mechanism. I am using the output of an old decoder in an old CD player (SAA7210 chip) and it is using 32 bit per frame, so 16 bit data and 16 bit dummies. The size of the shiftregisters depends on it.

I know not everybody likes the 1541. For me, this is the first attempt at creating my own DAC, so don't shoot me for using this chip. I was able to get two crowns for a reasonable price and this is just a start for me. On the other hand, some people like it very much.. I have no meaning yet, only have one old player with one old TDA1541 which is not modified.

Greetings,

Guido
 
NON-OS DAC

Hi Peter,
See for example:
http://members.xoom.virgilio.it/hi_fi/com.htm and
http://members.xoom.virgilio.it/hi_fi/cd960.htm
<B>AND</B>: ( this applies to the Philips CD960 that uses the same IC's.)
Modidification to NON-Oversampling including installation of KWAK-CLOCK [version with single supply]
Color codes of wires of KWAK-CLOCK:
Red = +5V
Green= ground
Blue= clock output of KWAK-CLOCK
Remove/desolder C511; C512 & X501 (11.2896 Mhz crystal) All are on board
PD16; the board with the large IC SAA7220 (Q503); just near this IC.
Solder 11.2896 MHz crystal to KWAK-CLOCK PCB.
Connect the green wire of the KWAK-CLOCK to the negative side of C514.
Remove /desolder entire IC SAA7220 (Q503) 24 pins.
Connect the red wire to pad of pin# 24 of IC SAA7220 (Q503)
Connect blue wire to pad of pin# 9 of IC SAA7220.
With a piece of insulated wire connect pad of pin # 1 with pad of pin# 18 of IC SAA7220.
Same for pad of pin# 2 with pin# 16.
Same for pin# 3 to pin# 15 .
Please note these instructions are different from Marco Amboldi's as no tracks are cut in my modification making the modification reversible, not permanent.
Also note that with the latter modification the digital ouput is lost. You can not use a outboard DAC component. The KWAK-CLOCK output is directly connected to pin 26 of SAA7210 (digital signal processor)

:)
 
Thanks Elso. I've got your schematic for KWAK-CLOCK2, that's the current version. I understand, I remove the crystal (11.2896MHz) from DSP board and use it with KWAK-CLOCK2.

Send me info regarding pricing for board and/or a whole kit.

You said: >The KWAK-CLOCK output is directly connected to pin 26 of SAA7210 (digital signal processor) <

Is that correct? Or is the output of a clock connects to pin 19 of SAA7210?

They recommend OP 42, OPA 604 chips for analog output. Is there any better alternative?

Thanks again for any info.

PS. One reason I'm doing that, I'd like to compare TDA1541 with AD1865 chips.:)
 
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.