# Opamp circuit help needed

Discussion in 'General Electronics Discussion' started by rkovach12, Aug 16, 2010.

1. ### rkovach12

2
0
Aug 16, 2010
I am trying to design a circuit to read pressure with a voltmeter. I have a
pressure transducer that has an output of .53 volts at 0 psi to 4.53 volts at
500 psi. I used Ltspice to test my circuit and it works fine with the
simulation. The problem is when I build the actual circuit, the output is not
correct. Any input between 1 and 4.53 volts will give me the correct
output,(within 1 to 2 psi which is close enough for my use). The problem is
in the range between .53 and 1 volt. This is the range I would like it to be
the most accurate. At .53 volts input, the output reads .14 volts and should
read 0 volts. As I increase the voltage to 1 volt, the error gradualy
decreases. I am useing a 741 opamp with a dual-polarity supply from two 9
volt batteries. Would a different opamp help? Please note that I am a novice.

File size:
24.1 KB
Views:
269
2. ### (*steve*)¡sǝpodᴉʇuɐ ǝɥʇ ɹɐǝɥdModerator

25,448
2,809
Jan 21, 2010
What do you reference your input voltage to? -9V or 0V?

3. ### NickS

367
0
Apr 6, 2010
My first instinct was that the amplifier Vss was tied to gnd and not a rail-to-rail device. But I think the 741 is rail-to-rail and you say Vss is at -9V. So my first thought is not the answer. Can you show us on the schematic exactly as you have wired it with the batteries included.

4. ### (*steve*)¡sǝpodᴉʇuɐ ǝɥʇ ɹɐǝɥdModerator

25,448
2,809
Jan 21, 2010
I'm pretty sure the 741 *isn't* rail-to-rail wrt the inputs or outputs.

No, the output voltage swing is quite restricted, and this is likely to cause all sorts of weird feedback problems (leading to inaccuracy) when the output is supposed to be approaching a supply rail.

Clearly only an issue if voltage is referenced to a supply rail other than 0V (when using a double ended PSU)

5. ### rkovach12

2
0
Aug 16, 2010
Here is the complete circuit. Sorry for the poor quality but I had to use MS paint because my version of LTspice does not have all the components used in the circuit. I am useing R5 to simulate the transducer input for testing because it is a lot easier than adjusting the pressure up and down with the pressure regulator, (and the results are the same).

File size:
35.3 KB
Views:
205
6. ### Laplace

1,252
184
Apr 4, 2010
It was not really obvious to me how the circuit worked so I tried to analyze it. For analysis purposes the circuit can be simplified. R1 can be replaced by a Thevenin equivalent bias voltage Vb and source resistance. The equivalent input resistance Ri is determined by the setting of R1, R2, and R3. The feedback resistance Rf is determined by the setting of R3 and R4. So writing the node equation for summing node B which will have the same input voltage Vi as the "+" reference point:

(Vi-Vb)/Ri + (Vi-Vo)/Rf = 0

(Rf/Ri)(Vi-Vb) + Vi = Vo

Vi(1+Rf/Ri) - Vb(Rf/Ri) = Vo

From this we know that the ratio Rf/Ri = 0.25 since a 4 volt change in Vi must provide a 5 volt change in Vo. Also, when Vi=0.53 then Vo=0 so we can find the value of Vb:

0.53(1.25) - Vb(0.25) = 0 or Vb = 0.53*1.25/0.25 = 2.65

So the drill is to first adjust R1 to provide 2.65 volts into R2, then adjust R3 and R4 so the ratio Rf/Ri equals 0.25. Which you have done. With the 741, the first thing I would try is using 18 volt versus 9 volt batteries to improve the linearity. No ideas otherwise.

7. ### NickS

367
0
Apr 6, 2010
Well that looks like you wired the power in correctly so you should actually be closer to distortion at 5V than at 0.5V(headroom |+9 - 5|=4V; |-9 - 0.5|=9.5V). Before you resort to higher voltages(which may work) it may be time to consider a better/different op amp. I would look for something that is specified for rail to rail operation. I am no expert on op amps but I have always had success with the ones I get from linear tech. If I were making this circuit myself I would fully expect to be able to achieve satisfactory results with a single 9V battery(and the right op amp).

As a starting point check out the LT6004 and see how it compares with your 741.

8. ### (*steve*)¡sǝpodᴉʇuɐ ǝɥʇ ɹɐǝɥdModerator

25,448
2,809
Jan 21, 2010
And as a quick check, I'd double (perhaps triple) check that your schematic matches what you've wired up.(on a breadboard?)