1
\$\begingroup\$

I am a bit confused related to the concept of common node in a circuit. I know that all voltages are measured with reference to this node. But here's the problem :

I have some big circuit, containing a few IC, common emitter amplifer to amplify their output, followed by op amp filters. Now I want to know that in my circuit, do I need a single common node, which has the negative terminals of all the batteries included( for ic, bjt and op amp) and the ground terminals of ic's, op amp and bjt ?

Or do i need separate common nodes each for ic's, bjt and op amps ? If so, then will I just connect the output of each component to the input of next via a single wire ?

The circuit will be wired to generate a signal from ic's, amplify the signal using bjt and then filter the output.

\$\endgroup\$
2
\$\begingroup\$

Short answer: Use a single common node for the whole circuit.

Long answer:

When one part of the circuit sends a signal to another part of the circuit (for example one IC sending a digital signal to another IC, or an op-amp sending an analog signal to a filter) there has to be a complete circuit for the signal to be transmitted. If the signaling scheme is notionally "single-ended" (for digital) or unbalanced (for analog), then the current associated with the signal must return by some path. Typically this is the "ground" node, or some other node that is, at least for ac analysis, shorted to the ground node.

Also, consider if two circuits have their ground isolated from each other, and one delivers a "5 V" signal to the other. If somehow their individual grounds have drifted so that there is 50 V potential between them, it's very unlikely that a signal that is "+5 V" for one circuit will be within the safe operating limits for the other.

Of course other options are available. You can use differential signalling, which tolerates small variations in ground between the parts of the circuit. This is common when the two subcircuits are several feet apart from each other.

You can also use differential signalling with galvanic isolation if you need to accommodate 10's or 100's of volts difference between the grounds of the two circuits. But this adds cost and typically limits the performance of the system (for example, transformer isolation won't pass signals that include a dc component, and opto-isolators have speed accuracy limits)

If you have no reason to do otherwise, use a common ground between all the parts of your circuit.

\$\endgroup\$
  • \$\begingroup\$ .Thanks. Just one more doubt. Common node even if power req are different ? My ic uses 5v while others 10 and 15v. I was concerned that using a common node in this situation might damage the ic. \$\endgroup\$ – Plutonium smuggler Sep 2 '14 at 4:36
  • \$\begingroup\$ If the different parts of the circuit damage each other, it won't be because you used a common ground for all of them. But without knowing more about your circuit I can't say you can connect a 10 V subcircuit to a 5 V subcircuit without damaging something. For example if the 10 V subcircuit produces a 10 V output and you connect that to an input of the 5 V subcircuit, that's likely to be a problem if you don't specifically account for it in your design. \$\endgroup\$ – The Photon Sep 2 '14 at 5:24
  • \$\begingroup\$ Ok. I take it that as far as my output of one subckt is within the input limits of the following one( which it is), there will be no damage solely because of common. Thanks a lot. \$\endgroup\$ – Plutonium smuggler Sep 2 '14 at 5:51
  • 1
    \$\begingroup\$ +1 for a well formulated, easy to understand answer with some bonus information. \$\endgroup\$ – Rev1.0 Sep 2 '14 at 7:03

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.