Why does op amp saturation occur below supply voltage?

[mememe653]

I built an inverting op amp circuit on a circuit simulator and did a transient analysis on it. I set the gain such that I expected a clipped output, but to my surprise it got clipped at +9V and -9V instead of +10V and -10V, which were the voltages supplied to the op amp. So why would this be the case? What accounts for the 1V difference about the supply rails?

 

[carlgrace]

The output transistors need more than a volt in order to operate in their high-gain bias regions (forward active if this is a bipolar opamp, saturation if this is a CMOS opamp). The output devices are probably laid out in a common emitter / common source configuration. To understand it, draw a common-emitter NPN amp driving a resistor load. What is the minimum output voltage of this circuit? Why can't it go down to VEE?

You can buy rail-to-rail opamps which use a more complex output stage if you need this functionality.

 

[eq1]

Could also be the load was too strong. Was there an output resistor? What size feedback resistors did you use on the opamp? But most likely it is what Carlgrace said.