Mixing A/D power domains

Tip / Sign in to post questions, reply, level up, and achieve exciting badges. Know more

cross mob
lock attach
Attachments are accessible only for community members.
Anonymous
Not applicable
        Hi, I have separated power domains for analog and digital with a common ground on a isolated copper area where the battery plugs. I'd like to check if in this case would be a good idea to connect a voltage divider to Vdda using an open-drain pin to ground (as seen on attached pic), since this would make the return current flow through the noisier and possibly shifted digital and battery grounds. Is it worth connecting the bottom node to Vssa using the analog globals for the sake of signal integrity? Even though this would add around 1k resistance to ground. Am I worrying too much? Does it make a noticeable difference? Thanks!   
0 Likes
6 Replies
Anonymous
Not applicable
                
      

How about this? Use ADC as differential mode.

      

 

      

Image-1

      

insamp1.JPG

      

 

     
    
   
0 Likes
Anonymous
Not applicable
        Hi PSoC73, thanks for the reply, I forgot about that, let's say that I will probably use the ADC in differential mode which will attenuate these effects I describe like the ground shift. But my concern is being able to connect and disconnect the voltage divider dynamically to avoid wasting power and I'm still not sure if mixing analog power with digital GPIOs is a good idea. Thanks again for the advice.   
0 Likes
Anonymous
Not applicable
        This method is just adapt with rail-to-rail range. AI_Pos and AI_Neg must stay within Vdd to Vss. If you want more spacious range, I recommend you to proper analog isolator.   
0 Likes
Anonymous
Not applicable
        But, More stable way is Isolate after ADC stage. Use a digital isolator, that is accurate. Do not use photo-couppular, It's my experience.   
0 Likes
ETRO_SSN583
Level 9
Level 9
250 likes received 100 sign-ins 5 likes given

Your question is also a f() of the resolution and accuracy you are

   

looking for. Clearly digital ground will have noise on it, and ohmic

   

drop. Noise you can take out of A/D result by filter/averaging. However

   

it has to be uncorrelated noise to be effective. And digital noise has

   

strong correlated components in it typically. Ohmic drop by cal routine.

   

And all of this a f() of resolution you are seeking.

   

 

   

Possible references that may be of use -

   

www.google.com/url   AN60263

   

www.cypress.com/      AN66444

   

 

   

Regards, Dana.

0 Likes
Anonymous
Not applicable

I'll try your suggestions to see how they work for me. Thanks a lot for your replies.

0 Likes