Op-amp buffer output voltage offset

Question

I am building a system which will measure and record the voltage of batteries. There are 8 channels in the system, each channel being capable of measuring 0-35 V. The accuracy we have decided upon is +/- 0.02 V.

The LMV324 op-amp was chosen as it has an input offset voltage of max. 6 mV specified. A 12-bit ADC is being used with a 1.8 V reference voltage.

I have built the circuit, and found that even with no input voltage applied, the output of the op-amp is 0.04 V. I have two of these op-amps on the same PCB, and all output voltages between 0.038 V and 0.04 V with no input.

At first I thought that noise/ripple from my DC-DC converter, which powers the op-amps, may have been the problem, but then I switched over to a linear regulator and the exact same thing happened.

Why is this happening?

Answer 1

Despite what some op-amps claim, none that I am aware of are able to get their outputs all the way to their positive or negative supply potentials. Furthermore, op-amps all have a limited range of permissible input potentials, often very similar to their output range. In some cases, taking an input outside of that range can cause the output to fly off to the opposite extreme, a phenomenon called phase reversal.

That means both of these situations are always problematic:

^{simulate this circuit – Schematic created using CircuitLab}

You have to take care not to exceed the limits imposed at the inputs, or expect an output swing extending to either supply. You applied 0V to the non-inverting input, and were surprised to find that the output was above zero, by some 40mV. Now you know why.

The obvious solution, but one that is quite inconvenient, is to provide the op-amp with a negative supply potential, for example \$V_{EE} = -5V\$. This solves two problems. Firstly, the inputs can now be taken all the way to 0V, and the output is able to follow, as expected.

Another inconvenient consequence of this solution is that the op-amp is now able to output negative potentials, which can be harmful to the subsequent ADC, and steps must be taken to protect it.

If you choose not to go that route, then have two issues to address. Firstly, an op-amp input of 0V may not be acceptable. Secondly, even it it were, the corresponding follower output of 0V is not even possible.

The simplest way I can think of to get past these hurdles, without having to obtain a negative supply somehow, is to raise the minimum op-amp input potential to 100mV or so. This would mean that 0V at your input would result in +100mV at the op-amp's non-inverting input, and an input of +35V would apply +1.8V.

Obviously, this means that you will lose some of the lower counts of the ADC, but that's still \$\frac{1.8V-0.1V}{1.8V} = 94\%\$ of full scale. This 100mV offset can be compensated for in software.

To introduce a +100mV offset will require some modification to the potential divider, and some algebra to figure out the resistor values. You will require a low impedance, low noise, stable positive reference potential. If you trust your op-amp's positive supply in these respects then you can use that:

^{simulate this circuit}

That circuit yields the relationship between input and output you are looking for, as shown in this sweep of input from 0V to +35V:

Your ADC has 12 bits of resolution, which is \$2^{12}=4096\$ distinct values, each representing \$\frac{1.8V}{4096}= 440\mu V\$. Power supply fluctuations will find their way to OUT, attenuated by a factor of \$\frac{R_2}{R_1+R_2}=\frac{1}{20}\$. That means 1 bit corresponds to \$20 \times 440\mu V = 9mV \$ of ripple in the +5V supply.

If your supply has more ripple or variation than this, and you wish to retain the full 12 bits of accuracy, then you'll have to use a precision reference, such as the TL431:

^{simulate this circuit}

Answer 2

Input offset means the op amp can see a difference of >6mV between it's inputs. The output can still be over 6mV. You need a rail to rail opamp which goes down to zero volts at its output. I don't know if there's anything cheap which can do that.

Here's a tutorial on input offset voltage https://www.analog.com/media/en/training-seminars/tutorials/MT-037.pdf

The usual way to get a zero output is to use a balanced dual supply. And, then some kind of protection circuit to make sure output is within uC input limits. This makes every single op amp ever made go down to zero volts.

Or, one can add the input with a bias voltage (65mV in this case) and then feed it to the op amp. But, this will likely introduce even more error.

If you can find an opamp that does have 0 V output, but has a high input offset, you can still use the LMV324. Set a gain on the LMV324 and feed it into the other opamp.

[The OP07 is not a rail to rail opamp. From the datasheet, with ±15V supply, output swing is ±12.5V max. That means, if power supply is single rail 5V, output will vary from +1.5V to +3.5V max, instead of 65mV to 4.99V for the LMV324 (typical), or even 5mV to 3.5V for the LM358 (typical)]

added At 3.3V single rail supply for the opamp, for measuring battery voltage, the LM358 is fine. I would try to increase the supply voltage a bit if possible to account for inter chip variations, as 3.3V leaves exactly 3.3-1.8=1.5V headroom (which is the limit for this opamp).

As Doodle and others say below, in reality a battery will never drop to zero volts, so rail to rail opamp is not really required.

Input null would be great for accuracy. However, if the OP is using a uC, the calibration can possibly also be done in software, if the input offset for the particular chip is stable (this is an assumption).

Answer 3

The accuracy is determined by the linearity and drift errors of the resistive divider, op-amp and the ADC. Such circuits are typically calibrated using an external voltage source on the production line or during test/commissioning, and you don’t need to use precision resistor networks nor low-offset op-amps.

If you can, it’s always an option to use some electronic switches on each input to allow for self-calibration on power-up and periodically thereafter if you want to completely cancel out thermal drift.

Don’t forget the EMI filtering on the input, as well as ESD protection!