TDA1543 to replace saa7321

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.

Bas

diyAudio Moderator Emeritus
Joined 2003
Hi All,

I've been tweaking my Philips cd634 wich had saa7321 and saa7310. Now I made a simplest tda1543 DAC wich I feed directly with the saa7310 with 2k3//150nF as Vref resistor and like in the datasheet an opamp stage with opa2134. The tda has it's own power supply with oscon and bypass cap. It sounds great and very open with superb dynamics but produces an awefull amound of background noise in my headphones and thru the speakers. The original opamp stage is still in the player but replaced by again an opa2134 and also the headphones opamp is 2134.
How to get rid of the noise? Any thoughts? Drop me a line!


Bas.
 

Bas

diyAudio Moderator Emeritus
Joined 2003
Hi Abraxalito,

No I haven't removed the 7321, it's still in place. I could lift the power supply and some other pins to make shure there is no 40mA's lost.
You say grounding is the problem but I can't figure out where. The tda and opamp ground comes from the groundplane on top of the board........
If I remove the tda board completely there is even more background noise........any tricks I could try?



B.
 
Just had a look at the datasheet for the SAA7323 (upgrade part to the '21) and it seems you can't remove it completely because its the clock master for the whole player. But they have separated out the digital supply to the clock (pin 23) from the rest of the digital stuff (pin 22). So you could try just gently lifting pin 22.

The grounding is a problem primarily because there's a groundplane shared between digital and analog ground. So noisy digital power supply currents get coupled into supposedly clean analog grounds.
 
Problem is that there's no isolation of grounds within the IC - I think all the noise will be coupled into the substrate. Lifting the ground for the digital supply with an inductor probably just forces the noise to take another route through the IC substrate. So I don't think that's worth trying.
 
Sounds like then that some other circuits depend on that digital power. So reconnect pin 22 and we'll try some better filtering on that supply.

Remove resistors 3661 (4R7) and 3662 (33R) and capacitor 2758 (56p). In place of the 33R fit a ferrite bead with two or three holes through it, loop some bare wire through the holes with at least 360 degrees of looping. We need something lossy at HF, a ferrite bead is better than an inductor for that. Then see if the noise you're getting changes when you short out that ferrite bead. If it doesn't then my original hunch wasn't correct and the noise is arising some other way.
 

Bas

diyAudio Moderator Emeritus
Joined 2003
Why do you think it's from that PS connection? You think it's digital noise or analog? I could just add a ferrite bead with some wire thru instead of the inductor for the test.
Let you know what the outcome is when I have the time to mod.
TNX!
 

Bas

diyAudio Moderator Emeritus
Joined 2003
No succes! I have removed the components and fitted in a choke but it seems to become worse in the new situation.
I'm used to silence in the 600 series from Philips. I've already tweaked 600, 614, 502, 710. Some fitted a new clock in, some made NOS. I have a 634 spare and can compare to the original. The sound of the tweaked 634 is so much better but the original is much more quiet in noise. The tweaked 634 is the best player I have at the moment, not in noise but in sound.
By the way, I did more to it. Muting transistors out, output caps replaced, power supply caps replaced by low ESR ones and bypassed by the orange Philips 150n's, molded the X-tal in Shin Etsu glue, damped the lid etc.
 
Why do you think it's from that PS connection?

Its the one that powers the fast digital circuits, the datasheet says there's 40mA of current going in. The bigger the current, the bigger the noise current potentially.

You think it's digital noise or analog?

Noise is noise. Can't understand how 'digital' noise could differ from the 'analog' variety.

I could just add a ferrite bead with some wire thru instead of the inductor for the test.
Let you know what the outcome is when I have the time to mod.

In your subsequent post you refer to 'a choke' - is this a slip or did you really fit a choke as opposed to a ferrite bead? Assuming you did use a ferrite then the next noise source to explore would be the other digital chip, the 7310.
 

Bas

diyAudio Moderator Emeritus
Joined 2003
Hi,

I wrote choke but mentioned ferrite bead. Just a bead with some teflon wirewrap wire thru.

How come the non modded cd player is quiet and the modded is not? It might have something to do with my mod right?

I lifted p10 an p44 earlier and they are still lifted, maybe that is the problem?
 
How come the non modded cd player is quiet and the modded is not? It might have something to do with my mod right?

That would be my best guess yeah. What other mods have you done and not mentioned so far?

I lifted p10 an p44 earlier and they are still lifted, maybe that is the problem?

They seem to be just analog outputs. Not likely to be noise sources.

OK then if digital noise is noise born of digital circuits, then yeah it seems you have a digital noise problem here.
 

Bas

diyAudio Moderator Emeritus
Joined 2003
I cut traces from WSI CLI and DAI from I2S bus just after the PCF3523 that would go to the 7321 so the inputs from the 7321 are floating.........eeehhh what else, cut traces from p40 and 14 and one cap came off 2656 100p near p40.
Fast shottky type diodes in power supply BYV26's that's about it I guess...........

The opamp stage after my tda1543 has just one feedback resistor of 2k7, no cap used parallel, can't be the prob. cause I lifted the opamp out already, same noise.

I discovered the excessive noise whilst listening to a noisy recording. After listening to a silent recording or no recording at all, there was still the noise.
 

Bas

diyAudio Moderator Emeritus
Joined 2003
Got it now. I again disconnected my dac board with opamp stage and listened very carefully and most of the noise is gone now. I guess it has some to do with the Vref adjustment. I have 2k7 from Vref to GND and because of the distortion on the output I experimented with a cap parallel. I guess there is 150n parallel now but that might not be the good thing to do. I also bring Vref to the opamps like in the datasheet. Is there a better way to make it work? Passif? Maybe 2k7 as feedback resistor is not enough?
 

Bas

diyAudio Moderator Emeritus
Joined 2003
Made passif I/V converter like Doede Douma in his ddddac. 1xtda1543 5V supply 1k as Vref and 1k as L/R output resistor to GND. Sounds awesome!
Mute doesn't work properly now, I already cut out analog mute transistors. Any other possibility to make mute work properly again?
 
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.