Single supply op amp with negative input

I used simple lm358 as inverting amplifier with single supply. The input voltage to the inverting terninal is negative 0.5 volt .i think that i lost the op amp .how to fix this problem?

mshh said:

I used simple lm358 as inverting amplifier with single supply. The input voltage to the inverting terninal is negative 0.5 volt .i think that i lost the op amp .how to fix this problem?

Click to expand...

What problem?
It should work with a single positive supply as an inverting amp with a negative input (it won't work with a positive input).

Post your circuit.

it was working fine, after a period of time there was no output. IS it allowed to use negative input with single supply?


- - - Updated - - -

I think that i should tell you that the input is ranging from +0.5 v to -0.5 volt so i used inverting and non inverting on the input to get positive output in every case, here the complete circuit

A negative input is okay as long as it doesn't exceed -0.5V.

If want both the positive and input input voltages to appear both positive (rectified) at one output you could use an ideal full-wave rectifier circuit.

i think that there is no ideal full wave as the diodes won't turn on till the input voltage exceeds 0.4 volt which means the input below this value won't be noticed. I need from 0 to 0.5 v.

mshh said:

i think that there is no ideal full wave as the diodes won't turn on till the input voltage exceeds 0.4 volt which means the input below this value won't be noticed. I need from 0 to 0.5 v.

Click to expand...

Au contraire.
You put the diode inside the feedback loop so its forward drop is reduced by the open-loop gain of the op amp, meaning the diode offset drops from about 0.7V to a few microvolts.

Here are a couple of such single-supply op amp circuits:

Hi,

If the signal goes to an ADC and then to a microcontroller, then you could do the "rectifying" in software..
Just add a DC voltage to your signal.

Klaus

A junction-isolated technology op amp could be driven into
latchup if (say) you pull the base of a lateral PNP input
(Nwell) below ground (Psub) waking up the parasitic NPNs
all around it. There is often not only a negative voltage
limit but a pin current limit, and many older parts were not
so good about latchup protection (challenged as they were
to fit the circuitry into a 50x50 mil die, in turn constrained
by the acceptable package options and coarse feature
sizes of the day).

If it says "don't do that", don't do that. They may not say
why, but they said, so it's on you.

The initial question confuses the problem. In normal operation, there's no negative OP input, feedback pulls the non-inverting input to ground +/- a few millivolts which is well within specified LM358 operation. Problem is possible negative overload and latch-up, as discussed by dick_freebird.

In case of LM358, behavior with negative input current is detailed discussed in the datasheet. The device has even a maximum current rating for Vin < -0.3V of 50 mA, which is quite a lot. Most modern OPs have a clamp diode rating of 5 or 10 mA. In other words, there shouldn't occur destructive latch up if the voltage across the 1K resistor doesn't exceed 50V.

It's further discussed that the output voltage can behave irregularly during clamping action, under circumstances gain reversal may occur. If you want to avoid it, place an additional schottky clamp.

I'm referring to original National Semiconductor data sheet, don't know if there are cheap replicas that don't manage to avoid latch up.