So i got lucky and found a rack of 10 TL071IP linear opamps laying around. I need to use one eventually to amplify the voltage off of a piezoresistive sensor. I attached the setup.

The sensor varies from about 20k to something my multimeter cant measure (>40M).

So as the topic title suggest the problem is in the opamp output. putting the oscilliscope on + pin on the opamp gives me a nice .2v to 4.16v reading, which is expected. However the output of the opamp does this weird clipping thing... whenever the sensor is completely depressed (in the >40Mohm state) the opamp output is 4.56v, but then when i touch the sensor the slightest bit (reducing the resistance to the 1-2Mohm range) the opamp outputs as expected.

So my question to you is:
is this a busted opamp or am i simply not delivering enough current to to the + port for the amp to make a useful reading?

Please help, ive been stumped on this for a while

~

 

putting the oscilliscope on + pin on the opamp gives me a nice .2v to 4.16v reading, which is expected.

that's your problem #1.

with the sensor at its lower resistance (200k), you should read 2.21k/(200k+2.21k)*5v=50mv.

with the sensor at its highest resistance (40meg), you should read practically 0v.

 

that's your problem #1.

with the sensor at its lower resistance (200k), you should read 2.21k/(200k+2.21k)*5v=50mv.

my bad, its 20k (though it can go all the way down to 6k or lower if its really stressed). i should have measured it again before posting

\(\frac{22100}{20000+22100} \cdot 5v = 2.62v\), which is really the maximum the sensor is rated for (the really stressed 6k comes closer to the max voltage i was reading)

~

 

The common mode input range of the opamp does not contain the -Vcc. That is, the opamp does not take inputs down to ground. The opamp is working as specified. You need to use a rail-to-rail input amplifier for this configuration, one that accepts inputs to both the positive and negative rails, or add a negative supply to the amplifier if possible.

The TL071 is specified for +15v positive supply and -15v negative supply on the data sheet I saw. It is only guaranteed to operate correctly with inputs between +/- 11v, although it typically works better on the positive side.

 

The common mode input range of the opamp does not contain the -Vcc. That is, the opamp does not take inputs down to ground. The opamp is working as specified. You need to use a rail-to-rail input amplifier for this configuration, one that accepts inputs to both the positive and negative rails, or add a negative supply to the amplifier if possible.

The TL071 is specified for +15v positive supply and -15v negative supply on the data sheet I saw. It is only guaranteed to operate correctly with inputs between +/- 11v, although it typically works better on the positive side.

Oh, that makes perfect sense.


With the voltages im getting from the sensor unamplified its questionable whether i really need the opamp or not, but thank you for clarifying the supply voltage issue.

~

 

Oh, that makes perfect sense.


With the voltages im getting from the sensor unamplified its questionable whether i really need the opamp or not, but thank you for clarifying the supply voltage issue.

~

the opamp here is functioning like a buffer, or impedance converter.

whether you need it or not depends on the rest of your circuit.

As to common mode, there is no common mode signal precent so I am not sure why that would make a difference.

when the sensor is not pressed, what's the voltage at the non-inverting pin, the inverting and the output pin?

 

You may want some sort of amplifier just to lower the impedance at the point of the sensor, otherwise, anything you hook the sensor and resistor divider to, e.g. an a-to-d convertor, will affect the voltage in the divider.

You may try swapping the sensor and the resistor. Your opamp works better on the positive side, even typically including the positive rail.

 

For an ideal opamp, the V+ and V- terminals of the opamp are the same voltage. In reality they are slightly different as determined by the gain of the opamp itself, typically 200 for the TL071, and any offset voltages and bias currents ...

So, even though the original signal is not common to both terminals, the operation of the opamp make the two terminals nearly equal, so the amplifier sees common mode voltage. The TL071 does not operate correctly when the voltage of the V+ and V- are close to the negative rail.

 

The IC description of the TL071 from Texas Instruments says its minimum supply is 7V.
the datasheet shows that its inputs stop working properly when they are within 3V or 4V from the negative supply which is ground in this circuit.

The TL07x opamps have the Opamp Phase Inversion problem when the output suddenly goes high (+3.5V in this circuit). I don't know how its output went as high as +4.56V unless the supply is +6.06V.

 

For an ideal opamp, the V+ and V- terminals of the opamp are the same voltage. In reality they are slightly different as determined by the gain of the opamp itself, typically 200 for the TL071, and any offset voltages and bias currents ...

the opamp loop gain for a typical opamp will likely in the 100db or more territory.

for this particular part, it is 200 v/mv, or 106db.

that means for the output to swing 1v, the input differential needs to be typically 1v/(200v/mv)=1/200mv, or 5uv.

 

I apologize. I inadvertently left the units off of the number.

 

You may want some sort of amplifier just to lower the impedance at the point of the sensor, otherwise, anything you hook the sensor and resistor divider to, e.g. an a-to-d convertor, will affect the voltage in the divider.

Thats what i was thinking originally with the voltage follower. I also didnt know and A/D drew enough current to affect the voltage, so i guess im going to have to look into this again.

You may try swapping the sensor and the resistor. Your opamp works better on the positive side, even typically including the positive rail.

The reason the sensor is where it is is because i want to keep the signal as linear as possible. The resistance of the sensor is inversely proportional to the force applied to it, so the voltage from the divider is:
22.1k/(n/f + 22.1k) * 5v

where f is the force and n is the coefficient. having the n/f in the denominator makes the output voltage more linear with force than it would be if it were in the numerator.

**The real question i have is how do i generate the voltage required to power this thing? Otherwise should i just think about getting different opamps that would better operate in the rather limited voltage region I have (well not limited but it has to be one voltage, no negatives)?

I apologize. I inadvertently left the units off of the number.

Its ok, i wouldnt have bitten you for it

 

I would definitely get better opamps for a limited voltage range unless you are going to change the input voltages entirely. What you probably need is rail-to-rail input and output on the amplifier if you are going to use 5v.

As Audioguru correctly pointed out the output of the TL071 opamp is also not guaranteed to reach 5V with a 5V supply.

I don't have any part numbers off of the top of my head but I am sure that Analog Devices or Linear Technology or others have parts to fit your needs much better than the TL071.

 

I would definitely get better opamps for a limited voltage range unless you are going to change the input voltages entirely. What you probably need is rail-to-rail input and output on the amplifier if you are going to use 5v.

As Audioguru correctly pointed out the output of the TL071 opamp is also not guaranteed to reach 5V with a 5V supply.

I don't have any part numbers off of the top of my head but I am sure that Analog Devices or Linear Technology or others have parts to fit your needs much better than the TL071.

Thanks so much for the guidance. Like i said, i found the TL071's just floating around, and hoped i could use them. guess i should have read the data sheet more carefully.

Is there a more specialized IC or set of discrete components i could use for an analog buffer? it just seems like a waste of an opamp to only do voltage following. (im relatively new to opamps, so i wouldnt know if this is a very common and accepted use for them)

 

It is actually a very common use for an opamp.

It provides isolation of the resistor network so it is not disturbed by the impedance of the next component, maybe an A/D converter or filter. It also provides drive capability to charge elements down the line. If you are charging an A/D converter or filter, charging through a 40 MegaOhm resistor can take a long time. Accuracy and stability are nice benefits. I'm sure there are more benefits but those are at least a few... and opamps are not very expensive in the end.