Grounding of analog and digital components

Hi,

I am designing a circuit but confused about the grounding. I have read analog and TI application notes but still have some confusion.

I have one DSP board and with it I plan to connect my own designed board. In my board, I plan to include:

1) DAC for audio output
2) ADC for analog input
3) Audio Amplifier (7W)
4) Opamps

My supply is +12V and from that I will generate different supplies using DC-DC converters. +12V will be directly connected to the Audio Amplifier.

Now, I think I should separate the ground of Audio amp. from rest of the circuitry as it will create problems for ADC and DAC. I also intend to have a ground plane on the other side of the PCB which will be the ground for the rest of the components.

Obviously I have to connect the +12V supply ground to the other ground at one point and I think I will use 0 ohm resistor for that.

From design point of view, what else can you guys recommend?

Thanks

This is probably one of the biggest areas of discussion. There is no absolute right answer. It used to be that separate analog and digital ground planes were recommended. Now, it seems that a single ground plane is often used. You've got the problem of using DACs and ADCs which are both analog and digital devices, but the general recommendation is that they be connected to the analog ground. The method I've often used with success is a single ground plane with cutouts to localize the analog and digital return currents. The fundamental approach you have to take is to keep digital and analog return paths separated; remember that currents want to take the path of least resistance, so even if you've got a single ground plane, your digital return currents won't 'wander' into the analog area if they've got an easier path to the supply return. This article might help a bit:https://www.analog.com/en/analog-dialogue/articles/staying-well-grounded.html

Unless you really need the efficiency, I would not use DC-DC converters, since the switching noise they generate is very good at finding its way into the audio.

crutschow said:

Unless you really need the efficiency, I would not use DC-DC converters, since the switching noise they generate is very good at finding its way into the audio.

Click to expand...

Yes. I was going to say that, too. However, I HAVE used switchers with low-level signals, but it's just another layer of effort.

A switcher with a high enough frequency and no
subharmonic oscillation problems ought not bother the
audio. But that's an "if". I'd definitely stay away from a
non-fixed-frequency (PFM / hysteretic) converter style.

You might consider Class D with its frequency synchronized
to the power converter (if any) rather than a linear power
amplifier. Put all the noise where you know where it is, and
deal with one problem instead of N. Put all the chop -way-
out of band and maybe you can filter / ignore it adequately.
I know audio purists like to sneer at Class D but there's a
lot of work going on and products coming out, that try to
push the audio-quality-envelope. An eval kit could be a
nice cheap one-off vehicle, wrung out ahead of time. It
may come at near BOM cost and save you much effort.

Think in terms of DC and HF AC current loops when you
work on your grounding scheme. And be prepared to tear
it all up several times looking for the quietest configuration.
If you lay out your own board, make ground regions "X-acto
knife friendly" and put gold lands where you can reattach
regions with copper braid, etc. without ruining the whole
thing. Make decoupling and signal loop areas very small
to minimize self-EMI magnetic pickup.

If you can "float" the analog entirely, let it get its ground
from the signal source (there are small, isolated, high freq
DC-DCs which might do) and pass signal across boundaries
in the digital domain (there are now nice digital isolators)
this may be the cleanest for you.

You might consider a split ground plane with all digital grounds on one plane and all analog grounds on the other.
Typically the split would be done under the ADC and DAC chips (some ADC and DAC chips have separate analog and digital grounds).
You can connect the two planes together at one point using a ferrite bead to suppress any high frequency currents between the planes.
Localized 100nF cap suppression across each digital chip power to ground pins also helps minimize the high frequency digital noise from straying too far.

Use separate GND planes for each block then tie them together on a single common GND point where is also the supply voltage GND.
Star connection will be best possible solution in that such cases.( separated GNDs, digital and analog circuits,power consuming circuits,transient creative dirty devices etc.)

crutschow said:

Typically the split would be done under the ADC and DAC chips (some ADC and DAC chips have separate analog and digital grounds).
You can connect the two planes together at one point using a ferrite bead to suppress any high frequency currents between the planes.

Click to expand...

Although the ADCs have digital and analog ground pins, the mfr ALWAYS recommends they both be tied to analog ground. Then why do they label one pin digital ground? I have no idea.

For purposes of not contaminating the on chip analog
ground by digital section current spikes against bond
wire and lead frame inductance. Tying them off-chip is
better than tying them on-chip. Let the digital section's
ground bounce around by 100mV, who cares? Let the
ADC ground bounce the same, when you're looking for
1mV LSB settling in 100nS, you're pretty well screwed.

A digital section with a common clock can easily pull
100mA peak current. Way worse, if you have tens of
CMOS outputs looking at tens of pF of load apiece and
slapping them at nS risetimes with their 25mA drive
current. But anyway, say 100mA current impulse is 1nS
up, 1nS down, no flat top. Say your bond wire is 1mm,
so about 1nH, you have L*dI/dt of 1E-9*(100mA/1E-9) -
100mV of on-chip ground bounce.

All reasonable-ish IC numbers. Best keep that bit o'
nasty to yourself, Mr. Digital Section. Blow it out
your own pipe, not mine, says Mr. Prancey Analog.

Of course making the ground that serves both, stay
clean is now on you. But you have resources (like
thick copper ground plane, high value low ESR caps
and routing freedom) that the IC (with its spindly little
legs and even thinner wire bonds, if it's not a bumped
chip-scale package) does not.

My question was rhetorical. The point was, if you want both pins tied together, then give them names like GND1, GND2. Calling them analog and digital ground just adds to the confusion.