# How is it possible to pull up an op amp's output to its negative or positive rail when its input is floating?

If signal source is varying between zero and 5V; and if we buffer this signal with an opamp which works with +10V supply or */-10V dual supply, is it then possible to pull up the op amp's output to its around -10V negative or around 10V positive rail when its input is floating(when the signal is decoupled from the opamp's input)? Can you show a topology example?

The following behaviour I mean:

input ---> opamp buffer --> output

10mV ------> opamp buffer ------> 10mV

5V ------> opamp buffer ------> 5V

Floating ------> opamp buffer ------> -opamp's negative rail(-10V) or positive rail +10V (but not zero V)

Edit:

To be more specific the following opamp buffer does not go to positive or negative rail when its input is floating:

So in simulation output is no behaving how I need. how can we modify this to obtain -15V or +15V output (around one of the rail voltages) when the input is floating?

Edit for an answer:

• This website has a schematic editor. Mar 11, 2020 at 23:37
• your design specs are rather floppy. Mar 11, 2020 at 23:55

is it then possible to pull up the op amp's output to its around -10V negative or around 10V positive rail when its input is floating?

simulate this circuit – Schematic created using CircuitLab

Figure 1. A buffer amplifier with pull-down resistors. With SW1 closed the output follows the input. With SW1 open the output swings to -8 V.

Here R1 pulls the input down to -8 V set by divider R2 and R3. I've chosen -8 V rather than -10 V to prevent latch-up which can happen with some op-amps if driven too close to the supply voltages. You can modify this to suit your device specifications.

For a positive default just connect R3 to +10 V instead.

Now that you've revealed what your application is - a wind vane using a 330° pot with a dead-zone I suggest the following:

simulate this circuit

Figure 2. Adding C1 to the buffer input will hold the reading at the most recent value when the pot is disconnected.

The advantage now is that the reading won't jump between V+ and GND every time you run off the top end of the pot and reconnect.

For this to work well you will need a very high impedance input op-amp with low bias currents and a low leakage capacitor.

• Thanks a lot! Based on your circuit I will add a new schematics in LTspice and ask you couple of things about it. Mar 12, 2020 at 0:46
• Now I updated your circuit added to the question and it can be found here i.stack.imgur.com/AEQeL.png . So the input is coming from a 1k wind vane modelled in dashed rectangle in my drawing. For better precision I first needed to use larger resistors for R1 R2 and R3. I hope it is fine? Mar 12, 2020 at 1:48
• Besides that I have two questions: 1) I can only get very low offset with this LT1007 with dual supply. But when I use single supply precision opamps(such as zero drift) I never get such small offset as in dual supply, is that expected. I ask because it is easier to with single supply. 2) Do I need a feedback resistor with any value regarding this kind of load? Mar 12, 2020 at 1:48
• (0) Larger resistors should be fine. Your circuit will bias to 560 / (560 + 820) of Vss. (1) I'm not familiar with the LT1007. I don't understand what you are saying about small offset. Is your op-amp able to swing its output to GND? (2) OA1 is configured as a unity gain buffer. It has 100% negative feedback. Mar 12, 2020 at 7:13
• Thanks. But I think I will go for the first one since I and some other friends want to capture when the vane passing through dead band and omit those points in post analysis. I increased R1 to 100Meg and added C4 for wiper noise(suggested here electronics.stackexchange.com/questions/21297/…) and used a low offset opamp for single supply operation. In simulation it works very good. Here is my final one: i.stack.imgur.com/9QAYW.png What do you think? Mar 12, 2020 at 10:27

Yes.

You can even ground the output (choose the output =0Vdc not PE Gnd) when operating from a battery +/- to make a bipolar or split supply.

simulate this circuit – Schematic created using CircuitLab

A battery generally has a lower ESR for motors for surge currents than a Darlington buffer or even a SMPS. So be careful how you choose gnd and loads.