Hi guys,

Sorry for the long post.

I have made a few general observations related to the usage of single supply and dual supply op-amps and wanted to know if I am moving in the right direction. Please correct me where wrong.

- If input signal is positive and has no -ve part then single and dual supply op amps are analogous?
- If signal is single ended (having both -ve and +ve part) and we have a single supply op amp - just bias the op amp at half the supply?
- If signal is differential (having both -ve and +ve part) and we have a single supply op amp what do we do now?
Could we add an offset to the input signal before giving it into the op-amp?
- If we have a dual supply op-amp it doesn't matter what type the input is?

Thanks in advance.

 

All single supply opamps have the same number of power pins as dual supply opamps, and behave the same most of the time. For example, the input signal never should be greater than the positive supply or less than the negative supply, whether the negative supply pin is tied to GND or -12V or whatever. The main difference between the two types of opamps is how the part behaves when the input signal is *close* to the negative supply voltage.

A "normal" opamp can operate just fine with a single positive power supply, with the opamp Vee pin connected to GND, IF the input signal range never gets very close to that. On the datasheet, look for the input voltage range or input common mode voltage range. This tells you how close the input can get to either Vcc or Vee. Closer than this and the opamp will not function correctly. So for an older part like an LM741 with its Vee connected to GND, if the input gets down to +2V or lower the part will not amplify correctly.

A "single supply" opamp has a different type of input stage, usually using PNP transistors, that allows the stage to amplify correctly even if the input is at Vee *or even below it*. This is a very cool trick. So if for example you have an input signal from a sensor or something, and the signal range is from 0V to 3V, you cannot amplify or buffer this with an LM741 unless it has a Vee supply of at least -2V and preferably -5V. But an LM358 can handle this signal with its Vee pin connected to GND. Of course you can connect the LM358 Vee pin to -2V or -5V and it will work just fine, but it is not required the way it is for "normal" parts.

A single-supply part is an older term. Today many opamps are called "rail-to-rail". These parts are even better than older single supply parts because the input voltage range can get very close to both the Vcc and Vee voltage levels.

ak

 

@AnalogKid suppose the input signal is differential like na ECG signal having -2.5mV to 2.5mV signal range and I only have a single supply op amp then that would clip the -ve part of the signal while amplifying. What could be the way around it?

 

You just use a split supply or Virtual Ground using an opamp,

 

@Dodgydave yes that could be used but won't that be for a dual supply op amp. I want to use a single supply op amp because even if I split the power supply I would need 10V for dual supply whereas I would need only 5V for single supply.

 

All op amps are single supply, you just use the ground or half of the supply as a bias point for the signal input.

 

@Dodgydave would something like this work


R3(1) and U1 are 5V while the opamp inputs are sine waves 180 out of phase.

 

R2 needs to go to the junction of R3,R4, instead of the negative supply, then the op amp will sit at half supply.

Maybe a Differential amplifier would be better

 

There is more to an amplifier circuit than just its input voltage range. How much gain do you need, and what is the output voltage range? If you want a gain of 1000, so the output is +/-2.5 V, then a single supply opamp sill not work even though its input range is ok.

ak

 

Guys I am actually trying to get the ECG signal using a dual supply instrumentational amplifier (INA128) giving -5V and 5V at supply and things are working good but now I want to move to the single supply Instrumentational amplifier INA122 so that I don't have to use 2 batteries. But since the ECG signal is differential so using INA122 would actually clip the -ve part, so I somehow need to give an offset to the ECG signal.

 

You are looking for a beyond-the-rail amplifier, amplifiers whose common mode range extends beyond their rails. They exist but are generally quite rare.

A typical to get around is to introduce a common mode bias to your output signal so that is above ground. That diff signal can be differentially amplifies to get rid of that common mode bias.

 

Your amp design is close. The main difference between split rail and single rail amps is the lack of a ground reference to drive your signal around.

Obviously with only +5V your signal can't go to -5V, but you can make it go +2.5 and -2.5 around 2.5V. (Huh?) so when there is no input the amp output is 2.5V, a positive input drives it towards +5V and a negative input drives it to 0V.

Here is how a single ended input amp would look:

Make sure that R3 in parallel with R4 is much less than R2 (low impedance source). Adding a cap across R4 is a good thing.

You can use the same R3/R4 divider for multiple stages as long as it is "stiff" (solid and non-varying for all the amps connected to it.

 

You need to be concerned not only with the value of the gain but also the sign of the gain.

Think of every input signal as containing DC and AC.

Signal = DC + AC

After amplification, we get

Output = Gain x ( DC + AC)

Is the Output voltage within the range of the supply voltages?