I want maximum 5V output from an Op-Amp. Is it a good idea to use a supply voltage for the Op-Amp of more than 5V?

Question

I have the following circuit. An Op-Amp (currently TLC2272CP) as a buffer (voltage follower). The non-inverted input of the op-amp comes from a DAC (MCP4922). I measure with an ADC (MCP3202) the input voltage of the Op-Amp and I measure with the same ADC the output voltage. Currently I use 5.0V supply voltage. For testing I set the DAC to a fixed voltage for a few seconds. My goal is to have exactly the output voltage at the Op-Amp output which I set with the DAC independent of the load (max a few mA). The desired output voltage will change slowly (less than 1Hz)

My testing shows: If there is no load on the Op-Amp output then the output voltage is equal to the input voltage almost for the full range from 0V up to 5V. This is what I expected.

But with a load (initial test with a 2.2kOhm resistor to V+ or GND) the output of the Op-Amp goes only to a maximum of about 4.1V.

I looked at the data sheet of the Op-Amp and it shows a High-Level output of minimum 4.25V with a load of 1mA.

I want to be able to get nearer to 5V. Probably something like 4.8V is good enough.

The ADC and DAC allow max 5.5V supply voltage. That Op-Amp allows up to 8V supply voltage.

I guess I won't be able to do this with these parts and conditions. My idea is to use 6V or 7V supply voltage for the Op-Amp and continue with 5V supply voltage for the ADC and DAC. I guess with the higher supply voltage for the Op-Amp it should be no problem to reach 5V output, even if the output current is a few mA.

My question is: Is this the proper way to do this? Or what should I do to have output voltage, with load, of at least 4.8V?

And when I use a higher supply voltage for the Op-Amp I have to make sure that the output voltage is never higher than 5V. Because I don't want anything higher and a higher output might destroy the ADC which measures the output voltage. Currently my idea is to use a Zener diode to make sure the output is never higher than 5V (also not when the circuit is just switched on). Is that a good way to do this?

Answer 1

You have two problems. Firstly, you are quite correct, the output voltage achievable from a 5 volt rail is limited to circa 4.25 volts with a 1 mA load and, to improve this you can raise the positive supply rail a volt or so.

The 2nd problem is that the input common mode range is limited to only 4 volts (page 6 of data sheet) and, given that you are operating the op-amp as a unity gain buffer, this is likely to be the problem actually seen when powering at +5 volts DC. However, this will improve if you run at a higher supply.

And when I use a higher supply voltage for the Op-Amp I have to make sure that the output voltage is never higher than 5V. Because I don't want anything higher and a higher output might destroy the ADC which measures the output voltage.

You might be able to use a series resistor in your ADC input to limit the current to an acceptable value. Read the ADC data sheet to check. Else, I would use a schottky diode from the op-amp output to the local 5 volt rail of the ADC. A zener could be used but they tend to be a little leaky below their clamping point and may not quite be good enough for what you want.

Also, the op-amp you have chosen is not that great on input offset voltage but, if you are content with a standing error of up to typically +/- 1 mV then that's OK.

Answer 2

Even with a so-called "rail-to-rail" input/output op-amp you will not be able to get quite to the rails while supplying current, so your plan is a good one. If you need to sink current with 0V out you should also consider a bipolar supply for the op-amp, and even without sinking current it's possible the op-amp output won't quite get to zero without a negative supply.

Even with a somewhat higher supply voltage you should pay attention both to the output swing specification (under load) and the input common mode range of the prospective op-amp. Since you have a voltage follower, the input common mode range has to be 0-5V.

---

Clamping the output voltage to the allowable range of the ADC is another issue. Typically an ADC or similar part will be okay with -0.3 to +5.3 with a 5V supply, but read the datasheet carefully. Using a zener diode is usually not a very good way since the Zener has tolerance and does not turn on precisely at a given voltage. One approach is to use a rail-to-rail op-amp and use a supply that is slightly beyond the rails, such as -0.2 and +5.2V. A series resistor and Schottky clamps to appropriate voltages can work too (take care the 5V-ish voltage rail can actually sink enough current), but you need to consider the leakage of the Schottky diodes.

Sometimes it's best to just work with some reasonable range like 0.5 to 4.5V and give up 20% of the resolution in exchange for simplicity.