Odd behaviour from opamp

[Aethelstan]

Hi, I am working on a circuit to provide a controlled voltage between 0-12V from a PWM pin on a microcontroller. I have set up a non inverting op amp (TI RC4558) with a gain of two, just to test my ideas. The op amp behaves as expected as long as the input signal to the non inverting input is above 0.7V, with the output showing the gain as expected. However, when the input signal drops below this threshold, the output suddenly becomes saturated. I have performed a number of searches, but my google-fu is clearly inadequate. The op amp is connected to a single rail supply of +12V and GND. I have attached some screen shots of the basic circuit I am using and the output on my USB scope. I am totally boggled as to what exactly is occurring.

[Zero999]

Hi, I am working on a circuit to provide a controlled voltage between 0-12V from a PWM pin on a microcontroller. I have set up a non inverting op amp (TI RC4558) with a gain of two, just to test my ideas. The op amp behaves as expected as long as the input signal to the non inverting input is above 0.7V, with the output showing the gain as expected. However, when the input signal drops below this threshold, the output suddenly becomes saturated. I have performed a number of searches, but my google-fu is clearly inadequate. The op amp is connected to a single rail supply of +12V and GND. I have attached some screen shots of the basic circuit I am using and the output on my USB scope. I am totally boggled as to what exactly is occurring.

You're exceeding the common mode range of the RC4558, which isn't specified to work with its inputs that close to the negative rail. Swap it for the LM358 and the problem will go away.

Refer to page 5 of the data sheet:
V_ICR    Common-mode input voltage range (+/-15V supply at 25°C):     ±12V (worst case) ±14V (typical)

Your circuit is exceeding expectations by working down to 0.7V, rather than 1V, as the data sheet would suggest.
http://www.ti.com/lit/ds/symlink/rc4558.pdf

[danadak]

A calculator, useful, but as Hero999 pointed out you need to consult
data sheet for allowed CM range.


https://masteringelectronicsdesign.com/differential-amplifier-output-common-mode-voltage-calculator/


Regards, Dana.

[IanMacdonald]

The reason this arises is that most opamps rely on a 'long tailed pair' differential input stage, and if both inputs approach the supply rail then neither of the two transistors will be conducting, so the output will be indeterminate. 7.2 in the data sheet shows how this works, with two PNP's as the input pair.

[Aethelstan]

Thanks, some nice useful links. The silly thing is the final circuit will be using  +— 12V supply, I was running ahead of myself in testing this part of the circuit on a bread board before I received my lab supply. Hopefully I will remember this well for the future.