invert negative signal(-0.9 v Max) to possitive without -VCC ??

[settra]

hello forum. i have a negative signal, that ranges from 0 to -0.9 volt, and i want to make possitive (1-1 ratio).
i know that the theory suggets, that i use a buffer inverting amplifier, but the problem is, i do not have a negative VCC in my circuit (or dont plan to anyway..)

but since i dont want the op-amp to generate any negative voltage, is it really neccecery to have a -VCC??
i am talking for a circuit like this :

 

[alec_t]

What is the source impedance of the 0 to -0.9V signal? You could use summing resistors and a single-supply rail-to-rail opamp to add a 0.9V offset to the signal and invert the result.

 

[settra]

no idea. and i cannot messure it... but i run a simulation of the above circuit, and it seamed to work?

 

[alec_t]

Good.

 

[settra]

good? so what do you think? will it work ?

 

[crutschow]

The circuit you show is a non-inverting follower so it can not work with negative input voltage and no negative supply, regardless of what you simulation may show. Change the op amp circuit to a simple inverter (input resistor and feedback resistor to the (-) input with the (+) input grounded) and it will work, converting the negative input to a positive output. You can also add gain to the signal if you want by increasing the value of the feedback resistor.

Note: The op amp has to be a single-supply or rail-rail type.

 

[MikeMl]

Even the lowly LM358 (Guru's favorite opamp) will do it because the common-mode range of the inputs goes to Vss-0.5V.

 

[BobW]

This should be quite easy to do using a current differencing type (Norton) op amp such as an LM2900, or if you need something faster, an LM359. They were designed for this kind of thing. Have a look at the app note: https://www.ti.com/lit/an/snoa653/snoa653.pdf
IIRC it contains a circuit that does what you want, or something very close.

 

[cowboybob]

crutschow, MikeMI...

Do the values look right to you? I've never actually made this circuit before but it seems to SIM just fine.

Pretty cool...

 

[ronsimpson]

Cowboy, The circuit looks good. 270 ohms seems low. We don't know if that will load down the -0.9V. The resistors could be 1k or 10k. R12=R2 is right.

 

[settra]

crutschow: yes, i might have put inverting and no-inverting, the way you sed in the simulation
speed is not an issue, as the input will not change fast at all.
i might use some resistors, to do a gain of x2, but not sure if i need it!
do i need rail to rail? i dont think, since the outpout will not swing near 5v ??

what would be good op-amps to do this with?

 

[audioguru]

The inputs of a lousy old LM324 or LM358 work all the way down to the negative rail which can be 0V.
The inputs of much better MC3317x or MC3407x opamps also work all the way down to the negative rail which can be 0V. Their total supply is from 3V to 44V.

The MC33171 is a single, MC33172 is a dual and MC33174 is a quad opamp with low power like the LM324 or LM358 but they have full output swing up to 35kHz and have no crossover distortion. The MC34071 is a single, MC34072 is a dual and MC34074 is a quad opamp that have full output swing up to 100kHz and also have no crossover distortion.

 

[crutschow]

...............................
do i need rail to rail? i dont think, since the outpout will not swing near 5v ??
.........................................

You don't necessarily need rail-to-rail but you need one where the input and output will operate down to zero volts with a single supply, which standard op amps won't do. Those are called single-supply amps. Of course rail-to-rail types also do that.

 

[settra]

cool! thanks allot for the reply's everyone!!