LM358 Output saturation at 3.707V.

[debrajdeb]

Hi,

I am working on a temperature controller based on LM35 (Temperature sensor) and interfaced to PIC microcontroller ADC, with 5V as reference for ADC.

To increase the resolution of my signal from LM35, I have added an amplifier -- LM358, with a gain of 11 and supply of +5V and 0V.

My problem is that the output of LM358 (op-amp) saturates at 3.707V. I know that LM358 is not a rail-to-rail Op-amp, but I never expected the output to saturate at 3.707V (when supply is +5V and 0V). I decided to reduce the gain of amplifier so that the output never reaches 3.707V for my application need.

My query --

Where do I get this figure 0f 3.707V output saturation in the data sheet of LM358 (that will help me to figure out such error during design of any other project)?

 

[Nigel Goodwin]

I suggest you consult my PIC tutorials, where you can see how I designed around opamp limitations, and improved performance at the same time.

 

[Papabravo]

There is a parameter called "Output Voltage Swing" It has two items in the LM358 datatsheet. Voh is the output high voltage which is always lower than the supply voltage and Vol which is a couple of tens of millivolts.

Note also the "Common Mode Input Range" It is usually specified as V+ minus 1.5 to 2 Volts. Inputs at the rail caus problems also.

 

[picbits]

Use a seperate 7805 with two diodes in the GND leg which will bump up the supply to the opamp by another 1.4 (ish) volts. Feeding the LM358 with 6.4 volts intead of 5v will get you to that magic 5 volts output (well 5.1 volts).

 

[Roff]

Here's the relevant spec.

 

[audioguru]

The graph in the datasheet for the LM358 shows that its output gets 1.2V from the positive supply when it sources no current. Its loss is worse when it sources current.

 

[Speakerguy]

Just get a R-R op amp from TI. TLV2462 is a good part.

 

[Hero999]

Just power the op-amp from a higher voltage supply.

The chances are, you're running the regulator from a >9V so it won't be a problem.

 

[picbits]

Just power the op-amp from a higher voltage supply.

The chances are, you're running the regulator from a >9V so it won't be a problem.

Hence why I suggested running from a seperate 7805 with a couple of diodes in the gnd leg.

This would bump up the 5v output to approx 6.4 volts but more importantly should give the opamp a swing from around 300mv (assuming he doesn't have a negative power supply feeding it) to around 5.1 volts which is within the spec of the PIC input.

I've done a fair bit of research on the LM358 as I have around 4000 of them sitting on reels in the workshop

 

[Hero999]

The LM358 can nearly go down to 0V and doesn't require a regulated power supply so another regulator would be pointless.

 

[picbits]

I was thinking more of limiting the maximum output to a 5.1 volts max.

That way you won't damage the microcontroller if the opamp was to output its maximum.

In simulations the LM358 was notoriously unreliable at lower voltage inputs unless used with a negative supply.

I found a workaround by using a 7805 with three diodes in the ground pin and a diode on the output pin before the feedback resistor. This would give me a reasonably reliable 0-5v swing.

 

[Hero999]

That's a fair point but I thought microcontrollers had diodes to protect them from that sort of thing.

 

[audioguru]

The output of an opamp with its supply voltage too high might have enough current to destroy the input proctection diode of a PIC.

 

[Hero999]

Not if you use a series resistor to limit it to a safe level.