TDA1541 "Fizz"

Status
Not open for further replies.
TDA1541 "Fizz"

Gentlemen, I have a problem!

I have just built a DAC comprising a CS 8412 and a TDA1541A in what I suppose is the datasheet configuration.

It works, but there is a background “fizz” sound. It is pretty low-level, but I’m not talking about some sort of audiophile “Golden Ears” effect. It is pretty obvious!

OK. I know you need some information -

About the sound:

The sound is not a buzz from the power supply – I know what a 100Hz ripple sounds like!

It is not “hiss”. I really think “fizz” is the best description. It is modulated by the programme material. It is at its most obvious during low level – but not silent - material.

I wonder if it might be quantization noise. Could I have a few of the lower bits sticking?

About the circuit:

It really is datasheet stuff -

It is built on “Vero” and the earthing strategy is a bit schoolboy, but I expected that it would work and that multiple regulators, star grounding etc would yield “audiophile benefits”?

I know it’s not a lot to go on, but does anyone recognise the symptoms?

Nick
 
My old dac boards where also done on vero boards with local regulators (lm317/337) and stargrounding and also had some problems.

In my dac it sounded like fast morsecode at very low level with a hich pitch tone 1k-3khz.

Never found out what it was. My current dac boards using the same 1541 doesn't have this problem.

So it must have been something stupid I did or didn't do.
 
Grounding

:smash:

Sounds like a grounding issue to me. Could you explain your ground layout, or better draw a picture?

I've had these problems before, ranging from midly annoying (like your problem) through to big crackles through the speakers (due to long ground paths and sharing of signal with power grounds)

:hot:
 
Hi

Zodiac was spot-on. I broke a couple of loops in the grounds and the background noise went away.

Now I want to improve the grounding further but I don't know whether the 3 power supplies for the DAC IC should be decoupled to the analog or the digital ground.

My guess is that the + and - 5 are digital and that the -15 is analog - does anyone know??

Nick
 
Hi,

When you look at some design on the nets, you can find +5 and -5 are digital and -15 is analog. That's were i got the info.

Best layout as far as i know:

+5 and -5 decoupled to dgnd (obvious)
-15 to agnd (also obvious)
DEM caps to agnd (can also be found on the net)

Connect dgnd and agnd with a trace under the DAC as short as possible.

Use a starground on the agnd pin (PS, outputs, DEM caps, trace to dgnd)

Use a groundplane for dgnd. Then use a separate PS for digital and analog supply (2 transformers), so there is no groundloop.

I am using this setup without problems. Comments 😕

Greetings,
 
Guido, thanks.

I have modified the circuit, separating the digital and analog grounds completely. Instead of joining the two grounds under the DAC I have taken all the grounds back to the point where the PSU is connected.

No difference in sound, as far as I can tell ;-)

Nick
 
nwilson said:
Guido, thanks.

I have modified the circuit, separating the digital and analog grounds completely. Instead of joining the two grounds under the DAC I have taken all the grounds back to the point where the PSU is connected.

No difference in sound, as far as I can tell ;-)

Nick

Mmm

From an AD document "grounding in high speed systems", it is posted here somewhere. I editet the text a bit, removed the explanation:

Inside an IC that has both analog and digital circuits, such as an ADC or a DAC, the grounds are usually kept separate to avoid coupling digital signals into the analog circuits.....However, in order to prevent further coupling, the AGND and DGND pins should be joined together externally to the analog ground plane with minimum lead lengths......The name "DGND" on an IC tells us that this pin connects to the digital ground of the IC. This does not imply that this pin must be connected to the digital ground of the system. :smash: :smash:

BB things the same way, read the PCM63 datasheet.

Greetings,

PS, there is one improvement i know of, but minor. C who finds it :angel: And who finds the things i did not see :cannotbe:

Anyway, the intention with that pcb is that all PS'es are separate (digital, analog per dac, normal digital circuits, clock and that the analog outputs of the two dacs are not joined (transformer coupled and then joined.

Greetings,
 
Guildo,

Thanks again for replying. I have read similar advice (link the analog and digital grounds under the IC) in application notes. However, the writers of the notes always assume that "their" chip is the only one in the system with both analog and digital grounds. You cant very well join the grounds under all the ICs (or can you?).

Nick
 
Status
Not open for further replies.