Suppose an opamp is set up as a differential amplifier. One of its input leads read 50v and another 51v; the voltage values are referenced to source circuit's ground does opamp take into account this ground? So the differential voltage is 1v. Is this setup safe for the opamp? Will I fry it up because individual voltage is too high? I basically can't guarantee individual voltage values...but the difference will never exceed 1.5v.
If there's a common ground between the op-amp and 50V you'll fry it. If the 50V is floating/isolated it doesn't "exist" without a reference (or a current path to ground) and the op-amp will only see the difference.
What voltage is too much for opamp inputs?
That is an easy answer as asked it is dependent on its power supply, generally never above that but you need to check the data sheet, it will give you all the details you need. You can also use a voltage divider to get the higher voltage down to a reasonable level. Use precession resistors and it will come very close.
Its risky to raise the inputs above the power supplies of the amplifier.
However since you said " set up as a differential amplifier" you can select Rs in the below circuit to allow any voltage you wish. Not considering safety for the moment, you could have inputs to 1000V.
However matching the two Rs and two Rf resistors becomes critical and you gain goes down.
So in differential mode I can basically get by without connecting the source circuit's ground and opamp ground? what about in comparator mode....
I can't use a voltage divider since the source can be bipolar.
The grounds always need to be connected.
Check the op amp data sheet for the "common mode input voltage range" specification, which states the voltage range that can be applied to the inputs.
I can't use a voltage divider since the source can be bipolar
Consider using a bipolar op amp power supply. That saves a lot of trouble.
2 Likes
Your reference point needs to be connected, generally as in this case it is the ground.
Thnx for the helpful responses, for now I will skip opamps since I still dont fully understand how they work....and find a workaround.
Not true. You can make a differential amplifier and connect a battery with no ground between the + & - inputs and it will amplify the difference "as expected".
...I recently built a differential amplifier with no ground to the input/source.
1 Like
You are relying on stray high impedance paths to provide a ground. It is very bad practice, that sometimes works, but is extremely susceptible to noise and damaged inputs.
Can you explain more about this part
I have one of those bipolar opamps like the lm324. Thing is they still have limits and there is no way around the voltage divider method. For positive input voltage i can drop the voltage with respect to ground, but what if the voltage is reversed or negaitive..that is ground, initially which was at the bottom, now should be at the top right and current going backwards?
Please describe exactly what it is that you want to measure, stating the full range of expected input voltages.
You seem to lack some very basic circuit concepts.
-
Ground defines zero volts, and all voltages in a circuit are measured with respect to ground.
-
Current flow from a voltage source that is positive with respect to ground is conventionally into ground.
-
Current flow from a voltage source that is negative with respect to ground is conventionally out of ground, into the negative source.
-
Current flow is continuous, as a loop in an unbroken circuit.
I want to subtract two voltage potentials with opamp of bipolar circuits...voltage can be anywhere from -40v to 40v
The possible issue is that the battery (for instance) could drift out our of the common mode of your amp. Isolated thermocouples can have this issue. Its usually a good practice to put a high value resistor (1+ meg) between the two grounds to keep them within an acceptable range.
However this usually is not an issue but could happen. And if it does it is difficult to troubleshoot as the "drift" is usually not consistent.
you need one of these
No those are two different experiments....I am basically noob in analog world....so doing bunch of experiments to verify my assumptions....
Yes