Actions
Forum Thread Details
  • State Not Answered
  • Replies 3 replies
  • Subscribers 463 subscribers
  • Views 393 views
  • Users 0 members are here
Related

Need strategies to implement a virtual ground

Catwell
Several instances in the past I have come across the need for virtual grounds, often in audio and communication situations. Working with datasheets and application notes, on one particular audio design, I could not find anything in their designs that would give me the virtual ground that was constantly being referenced. The virtual grounds were being referenced everywhere, and other engineers that were working with me wanted to just tie those grounds to the return path of the circuit. I differed on their opinion, and two prototypes were made.

One where all the virtual grounds were tied to the negative terminal of the supply, the other engineer's idea, did not function at all. I watched them tinker with that PCB for almost a week with no results.

The one I had made used this sort of set up at the source:
image
And it, for the most part, worked and played sounds. However, I was not satisfied with the result, past lording my triumph over my fellow engineers. I felt it was an ad hoc sort of Band-Aid or the project.

What are the absolute standards in adding a virtual ground to a system? What are the best practices and references on the subject?

Cabe
Parents
  • 0
    In some circumstances, careful design can eliminate most need for a virtual ground. This is especially useful for AC coupled circuits. I have used dc coupling in a sequence of amplifiers just to propagate the half supply offset set up in the first stage input bias network after the input ac coupling/blocking capacitor. Where you might have used a resistor here connected to "virtual ground", you use two resistors each double the original value, as a divider, in order to develop the half supply offset (needless to say, these resistors are not bypassed.) Elsewhere in the circuit, connections ordinarily grounded directly (to virtual ground), are connected to the low supply (the offset ground) through capacitors. I vaguely remember TI  or Burr-Brown may have had an app note on using ordinary amplifiers in situations where wide common mode range amplifiers (ones that allow the negative rail as common mode limit, for example)  would otherwise need to be used. This is one situation you look at the common mode range on the spec sheet. The technique  of using a divider instead of a single resistor may be used in other circuits as well, for example, a full wave precision rectifier may benefit from some doodling on the back of an envelope. 
Reply
  • 0
    In some circumstances, careful design can eliminate most need for a virtual ground. This is especially useful for AC coupled circuits. I have used dc coupling in a sequence of amplifiers just to propagate the half supply offset set up in the first stage input bias network after the input ac coupling/blocking capacitor. Where you might have used a resistor here connected to "virtual ground", you use two resistors each double the original value, as a divider, in order to develop the half supply offset (needless to say, these resistors are not bypassed.) Elsewhere in the circuit, connections ordinarily grounded directly (to virtual ground), are connected to the low supply (the offset ground) through capacitors. I vaguely remember TI  or Burr-Brown may have had an app note on using ordinary amplifiers in situations where wide common mode range amplifiers (ones that allow the negative rail as common mode limit, for example)  would otherwise need to be used. This is one situation you look at the common mode range on the spec sheet. The technique  of using a divider instead of a single resistor may be used in other circuits as well, for example, a full wave precision rectifier may benefit from some doodling on the back of an envelope. 
Children
No Data