Why would ther be a ground pin on opamps??
You have inputs, output, power supply. All these are referenced to ground.
Now I see you changed the question - you should also change the thread title to not mislead people!
I don't really understand your new question. Ground is normally the reference for all signals as well as the supply.
Jan
You have inputs, output, power supply. All these are referenced to ground.
Now I see you changed the question - you should also change the thread title to not mislead people!
I don't really understand your new question. Ground is normally the reference for all signals as well as the supply.
Jan
Opamps don't have a ground pin and the use of ground on dual rail devices is that that's where the signal is referenced to.
A comparator has a ground pin, at least a pin that can be grounded, to interface with logic circuits that are referenced to ground.
A comparator has a reference pin that can be grounded if you want ground to be the reference...
Jan
Jan
Remember that a multimeter always has two probes. The meter displays the voltage between the two points.
You can't have 'a' voltage on a pin, it's always between two points.
So there is always a common reference point which we normally call 'ground' but that's just a convention.
We could have called it pupukaka and nothing would change in the circuit operation.
Jan
You can't have 'a' voltage on a pin, it's always between two points.
So there is always a common reference point which we normally call 'ground' but that's just a convention.
We could have called it pupukaka and nothing would change in the circuit operation.
Jan
Last edited:
The input currents typically flow to one of the supply rails, but its nanoamps anyway, all that matters in an opamp is the difference in input voltages. In a negative feedback circuit the opamp acts to keep that input difference (very nearly) zero. There is no need for a ground reference for that.
No idea whether this answers the question, but...
What's a bit odd about the classical triangular op-amp symbol with two inputs and only one output is that it violates Kirchhoff's current law, because a current that flows into the output has nowhere to go (no return path). In reality, the current gets split in some non-linear fashion between the positive and negative supply pins. The split supply brings those parts together again and the combined return current then flows through the ground wire, assuming the load is connected between the output and ground.
What's a bit odd about the classical triangular op-amp symbol with two inputs and only one output is that it violates Kirchhoff's current law, because a current that flows into the output has nowhere to go (no return path). In reality, the current gets split in some non-linear fashion between the positive and negative supply pins. The split supply brings those parts together again and the combined return current then flows through the ground wire, assuming the load is connected between the output and ground.
Signal processing is all about the signal path, details of power supply, decoupling or return paths are usually omitted (or placed elsewhere in the schematic) for clarity and less clutter, be it RF or lower frequency circuitry.
The end-run around - is to read and appreciate Analog Devices application note AN202 by Paul Brokaw.
It's a classic.
It's a classic.
Hi,
The opamp, although it is an integrated circuit, it doesn't normally integrated everthing you need for an amplifier.
It is more like a component on a circuit. So you have to understand the final circuit as a whole.
I'll try to ilustrate it.
Firstly, how a transistor works (BJT NPN for a basic amplification example).
You inject a current in the base and the transistor drains a larger current from the collector.
Below it's the simplest opamp circuit - a unity gain buffer using symmetric power supply.
Observe the positive and negative cycle inside the opamp - I've created a very basic BJT opamp just for illustration.
The 2 circles in the circuit indicate current sources to simplify the circuit (it will be composed by other transistors, resistors, zeners etc).
The input current flows through the input, internal transistors, power supply and returns through the ground to the source.
Output current flows from the power supply, the internal transistors, the load and returns through the ground to the power supply.
The opamp in this case is not connected directly to the ground.
Ground is used to connect the source, the load and the power supplies.
Of course, you can have an IC that has everything inside and then it has a ground connection - but this would not be an opamp generally speaking.
I'm simplifying things just to illustrate. There are many other types of opamp.
I'm using the very basic BJT class AB based opamp example.
Dig deeper ! See what happens with XLR output, using symmetric op amps.
You’ll see that electric ground becomes useless.
Then you can ditch the dual rail -15v 0 +15v for your favorite 0 +30v eventualy adding a few mA ground splitter in case you persist in the need of a middle electrical ground.
You’ll see that electric ground becomes useless.
Then you can ditch the dual rail -15v 0 +15v for your favorite 0 +30v eventualy adding a few mA ground splitter in case you persist in the need of a middle electrical ground.
The circuit described above refers to an unbalanced input and output, which have ground reference.
That answers your first question: "Where is the ground pin on operational amplifiers ?"
The answer for the question regarding dual rail devices is Shielding.
In a balanced signal (XLR) the "ground" is used as a shield - there is no user signal flowing through it.
That answers your first question: "Where is the ground pin on operational amplifiers ?"
The answer for the question regarding dual rail devices is Shielding.
In a balanced signal (XLR) the "ground" is used as a shield - there is no user signal flowing through it.
Last edited:
You must make sure that the lines on the XLR stay within the max and min supply voltages.
So whatever the supply voltages, bipolar or single polar, you need to somehow make a reference for the signal.
With a bipolar supply a convenient reference for the input (and output) is mid supply, or the supply ground. In an unipolar supply you can make a half-supply point as reference.
Jan
No, shielding has no relevance in balanced signals.
The purpose of a balanced connection is that it is not sensitive to hum, noise etc. So shielding has no use.
And it doesn't make a difference what sort of supply you have, as long as the signal levels stay within the supply range to keep the circuit working.
Jan