SAA7210 to PCM56, a Dutch-American connection

[Bernhard]

Originally posted by rfbrw


Only 4. Well, I suppose we've all got to start somewhere.

2 x 32 PCM61, is that Your's ?

There are a few reasons why I'm at 4 for the moment.

One is perfectly good but MSB adjust is not stable over time.
Perhaps I should try a more stable power supply...
4 gives more stable MSB adjust if only one adj. is used.
To avoid the pot and the drift, I try to cancel MSB errors by combination of 4 preselected chips.
Ideally it would be possible with only two chips per channel.

If it will still be better with MSB adj. , I will go for 32 per channel SE or 2 x 16 differentially.

Any suggestions what is better ? Diff. needs another op amp / transformer / instrumentation amp.

By the way, class is better than mass.
32 per channel is still not enough if performance of chips is unknown.
If 32 untested chips are thrown together, it is likely that the result is worse than one good & adjusted chip alone.

I use only chips that are preselected to be very good with MSB adjust.

[tubee]

Ok Rfbrw, you're the expert.

The datastream from SAA7210 has other info besides audio, will check it up in datasheets/service manuals.

Still LE has to be at the right moment for PCM56: Latch must have a falling edge at the moment LSB has been put in dac's register. Thats imo the problem with 1 channel in my current setup, a small timing error.

A simple glue logic schematic is easy to alter, i am not waiting for a no-funtional logic schematic with a lot of chips for first experiments.

These are my first steps on I2S to PCM matery. Have tried for a while to delay I2S to four 1451a dacs, but didn't get any sound out of it.

Nice dac btw!

[rfbrw]

Originally posted by Bernhard



2 x 32 PCM61, is that Your's ?

There are a few reasons why I'm at 4 for the moment.

One is perfectly good but MSB adjust is not stable over time.
Perhaps I should try a more stable power supply...
4 gives more stable MSB adjust if only one adj. is used.
To avoid the pot and the drift, I try to cancel MSB errors by combination of 4 preselected chips.
Ideally it would be possible with only two chips per channel.

If it will still be better with MSB adj. , I will go for 32 per channel SE or 2 x 16 differentially.

Any suggestions what is better ? Diff. needs another op amp / transformer / instrumentation amp.

By the way, class is better than mass.
32 per channel is still not enough if performance of chips is unknown.
If 32 untested chips are thrown together, it is likely that the result is worse than one good & adjusted chip alone.

I use only chips that are preselected to be very good with MSB adjust.

No not mine, it is Japanese. The multiple chip nostalgia business has never produced the kind of sound I like even though I couldn't tell you what that sound is. It is much more fun doing things with a FPGA and a pair of good dacs like the PCM1738E.

[rfbrw]

Originally posted by tubee
Ok Rfbrw, you're the expert.
The datastream from SAA7210 has other info besides audio, will check it up in datasheets/service manuals.

It has nothing to do with expertise. I have the very same datasheets you do and all I can do is relate that to what I see on the 'scope and previous experience.

Nice dac btw!

Alas, it isn't mine.

[jean-paul]

Originally posted by rfbrw
It is much more fun doing things with a FPGA and a pair of good dacs like the PCM1738E.

If you like the PCM1738 I can send you my last ones if you send me a PM.

[tubee]

It's a pitty i never did the 2nd electronics evening course, and the 1st one was 19 years ago. Still have to learn a lot in special in digital.
I see digital as analogue, but then in a switching way.

Did some research and indeed the data is audio from SAA7210, and subcode is available on other pin. Sorry for wrong comments here.

Tomorrow will try to solve some on the logic.

http://en.wikipedia.org/wiki/FPGA

Rfbrw, did you some special things like DSP with FPGA?

[jean-paul]

Originally posted by tubee

I see digital as analogue, but then in a switching way.

IMO a right view.

[Zoran]

Hi Bernhard
Thanks for posting conn. dia of Your working set up with PCM56...
I will try it...
Are You using the M0 to M3 pins connected like in the first diagram?
Did You try maybe the other way to convert I to V like I-buffer?

[tubee]

The multiple chip nostalgia business has never produced the kind of sound I like even though I couldn't tell you what that sound is. It is much more fun doing things with a FPGA and a pair of good dacs like the PCM1738E.

All matter of taste.

I am in for a honest sound for CD replay, and can not achieve this with my CD-upsampling facility of DVD/SACD player. PCM56 is coming close, but it can only with proper other things like CDM, I/V etc.

Btw what dac uses North Star, also PCM1738?


Bernard, why did you parallel the PCM's, to get K-grade dacs (or better)?


Something else: Jocko dislikes the TDA1541 (not only) because its running hot. PCM56 is a nice cool runner. And the AD844 is not getting warm either.
I was lucky i didn't drill the holes yet in cdp-cover for cooling !

[Bernhard]

Originally posted by tubee

Bernard, why did you parallel the PCM's, to get K-grade dacs (or better)?

See post #11

Did you try passive I/V or make any new listening tests ?