Improvements to TDA1541 dac output stage?

[sharpi31]

I'm looking at buying a TDA1541 dac kit and would be grateful for some advice. The schematic shows two opamp stages for each channel. Is this optimal? Could the scheme be simplified/improved?

Thanks in advance :-)

[sharpi31]

Could I just replace the output stage with this simple stage from the datasheet?

[oshifis]

My philosophy is the simpler is the better. You can get rid of the 2.2nF capacitors, as well. The next step will be to go for no-oversampling (removing the SAA7220), then rolling the opamps... This is a never-ending game.
Don't forget the output of the opamp will be at DC so a blocking capacitor is needed. It is in the signal path, so...

[sharpi31]

The first schematic I posted is incorrect - here is a corrected version:

[sharpi31]

I plan on using LME49710 opamps instead of the OPA2604 shown.

Would sound quality improve if I just use one opamp per channel without the 2.2nF cap?

What is the second opamp on each channel for? Is this a low pass filter for rolling off the >20KHz digital rubbish? If so, I think I can live without it (I'm used to TDA1543 non-oversampling dacs without filtering).

[poynton]

First opamp is the i/v converter, 2nd opamp is the lowpass filter.

Improve the power supply - separate the analog and digital 5v etc

Replace the clock with a better version

Use a tube i/v and output


Where do you stop??

Andy

[philpoole]

My philosophy is the simpler is the better. You can get rid of the 2.2nF capacitors, as well. The next step will be to go for no-oversampling (removing the SAA7220), then rolling the opamps... This is a never-ending game.

Be careful! My first DAC was a TDA1541A, and I simply implemented that opamp circuit without the capacitors and it worked great. BUT, I was using a slow opamp, and when I upgraded to anything just reasonably faster, the sound would collapse after a few seconds, and the opamp would get too hot to touch. I reckon the opamp was forced into oscillation. I was really surprised it was that cap at first. I spent a lot of time reworking the power rail decoupling and grounding and stuff like that before I noticed the missing cap was the problem.

Also, I implemented NOS and it sounded awesome, but it sounded much better when I added an SAA7220P/B. But this was a DAC, not in situ, in a CD player, so my recloking was perhaps different.

Sorry, Oshifis, I very rarely disagree with you!

But, there's loads you can do

[Mooly]

If you are doing any mods and component changes around the output OpAmps and filters than at the very least you have to do some simple measurement's to back up the changes, even if it's just a 20-20khz sweep from a test CD to make sure you havn't altered the response.

Have a read at Marantz CD50 Mod that is just stunning
Post #10 I actually measured the effects of this one.

[sharpi31]

My inclination would be to drop the SAA7220 and low pass filter stage, to give a non-oversampling dac with active I/V (LME49710).

Any thoughts?

[poynton]

Originally posted by sharpi31
My inclination would be to drop the SAA7220 and low pass filter stage, to give a non-oversampling dac with active I/V (LME49710).

Any thoughts?

The schematic shown in post 4 does not use the SAA7220 and so is already NOS.


Andy