Hi Red,
The subject of rail-to-rail input/output vs. non rail-to-rail is vast. Rail-to rail op amps primarily use CMOS or a combination of CMOS/bipolar processes because their transistor operating voltage conditions are more suitable for rail-to-rail designs than bipolar or JFET. Specialized design techniques are employed for the rail-to-rail designs to give them that ability.
The TI Precision Labs - Op amps series in the Input/Output Limitations section has a slide labeled "Input Stage" and then "Output Stage" that depicts a simple non rail-to-rail input stage design. That slide shows how the VSAT and VCE limits of the input transistors placed in the design prevents the it from being able to handle rail-to-rail input voltages.
The input stages of rail-to-rail typically use either a complementary NMOS/CMOS dual differential amplifier input pair, or an on chip charge pump to pump up the internal supply rail voltage to keep the input pair out of saturation (VSAT). Although not complete explanations here are two slide from the PRAMPS Op amp Technology Tech Day session that show the concepts.
If you Google the rail-to-rail op amps subject on line, or obtain a current text on op amp design more information should be available.
Regards, Thomas
Precision Amplifiers Applications Engineering
