# Op-amp output clipped

I've been trying to get into using opamps... I have a bit of experience with digital electronics, but this is definitely new ground for me. I tried to set up an inverting amplifier like this:

simulate this circuit – Schematic created using CircuitLab

When I measured the output, it seemed to be "clipping" in that it suddenly levels off at a consistent level: (these two graphs have equal scale)

I expected the output to go very close to 0 and 5V, perhaps with a slight margin, but in the range of millivolts.

To test this I removed the feedback resistor so the opamp would essentially attempt infinite gain, but the same problem occurs:

Notice how the output is still constrained.

Am I doing something wrong? Is there something I don't know about opamps?

• Ain't no way something as old as the ua741 is going to be rail-to-rail... – Ignacio Vazquez-Abrams Nov 16 '13 at 11:39
• Can you tell the output range of an opamp from its datasheet? – Matthew Sainsbury Nov 16 '13 at 11:58
• You power a 741 from an asymmetric 5V? Does that leave any output range for that poor old chip? – Wouter van Ooijen Nov 16 '13 at 13:04

Welcome to the world of real opamps. Real opamps are not ideal. For this particular case - the output does not swing exactly to rails. Event for opamps that are advertised as "rail to rail" output will be several mV from the real rail.

Output swing is usually specified in the datasheet. It's called output swing, min/max output or something like that, look for output specifications. For 741 it's +-16V when powered from +-20V or +-13V when powered from +-15V. It varies with load and many different factors.

Usually when designing opamp circuits you try to opearate in the middle between the rails, and avoid going near the rail (except for comparator configurations). Note that in amlifier case opamp will start to behave funky when you swing close to a rail - you wil start getting distortion. YMMV, depends on how good your opamp is.

Another thing - you are doing inverting configuration wrong. You have to tie noninverting input to a midpoint between the rails. 2.5V in this particular case. It does not necessarily have to be the exact mid point - it all depends on your input signal.

Think about it intuitivelly - in your case for the opamp to swing high, your input signal has to go below the ground, so you set up "virtual ground" somewhere between rails so that your opamp has some reference agains which to invert the input signal.

You are designing a linear circuit and the first rule with opamps is that the voltage at both inputs is identical in dc level and ac level. Negative feedback ensures this but, only if you set up things reasonably.

Your circuit will try and adjust it's output to make both input pins exactly the same but it can't because the output cannot go sufficiently negative to achieve equilibrium. The problem is solved by ensuring that the non inverting input has a positive voltage on it wrt the most negative supply of the opamp. Then it stands a chance but beware, opamps are non ideal and can't perfectly swing up or down to the supply rails.

For your circuit you would need to put a more central voltage on the non inverting input, possibly by equal resistors up to and down to pos and neg supplies respectively.