Non-linear response from an instrumental amplifier

[r.youden]

I am building an instrumental amplifier using a LM6154 and operating in single supply mode (this can't be changed).

Tested the circuit and everything looks good. I have now built the circuit but the gain is well down on what I expected (by a factor of 4) and the response in not linear. There is nothing connected to the output which would cause any issues and it is really starting to bug me now! Has anyone else ever seen this issue. I will attach a circuit diagram and output plot shortly.

Thanks.

 

[ericgibbs]

I am building an instrumental amplifier using a LM6154 and operating in single supply mode (this can't be changed).

Tested the circuit and everything looks good. I have now built the circuit but the gain is well down on what I expected (by a factor of 4) and the response in not linear. There is nothing connected to the output which would cause any issues and it is really starting to bug me now! Has anyone else ever seen this issue. I will attach a circuit diagram and output plot shortly.

Thanks.

hi,
When you say instrumental amplifier do you mean Instrumentation Amplifier.?, because the LM6154 is a standard OPA not a IA.?

Please post the circuit when ready.

 

[r.youden]

OK, probably should have been as a differential amplifier, as below:

 

[ericgibbs]

OK, probably should have been as a differential amplifier, as below:

hi,
That configuration looks OK to me.???

Was your diff gain aimed at ~65.??
Whats the supply voltage.?

 

[r.youden]

Yeah the gain is around 65ish. The supply voltage is 24V which is the upper limit of the LM6154 but I wouldn't have thought that would have a negative effect on things. I have taken some data and I will post that on here in a minute.

Thanks.

 

[r.youden]

OK, I have got some real world data.

This circuit is designed to provide a 0 - 2.5V output from a PT100 temperature sensor with 100Ω = 0V and 138Ω = 2.5V (it is nearer 2.6V but the gain can easily be adjusted).

The data and plot below shows what happens when I implement this circuit in anger. I have the reference input (TVREF) and the PT100 input (TVPT100). PTDIFF(2) is the output of the amplifier stage and Theoretical PTDIFF is the calculated value for what PTDIFF(2) should be with the component values selected. Unfortunately this plot is not large enough to show how non-linear the response of the circuit clearly.

This is really giving me a headache.

Thanks.

 

[ericgibbs]

OK, I have got some real world data.

This circuit is designed to provide a 0 - 2.5V output from a PT100 temperature sensor with 100Ω = 0V and 138Ω = 2.5V (it is nearer 2.6V but the gain can easily be adjusted).

The data and plot below shows what happens when I implement this circuit in anger. I have the reference input (TVREF) and the PT100 input (TVPT100). PTDIFF(2) is the output of the amplifier stage and Theoretical PTDIFF is the calculated value for what PTDIFF(2) should be with the component values selected. Unfortunately this plot is not large enough to show how non-linear the response of the circuit clearly.

This is really giving me a headache.

Thanks.

hi,
Have downloaded your data, will look thru it.

How do you have the PT100 and TVREF connected to the inputs of the IA.?

 

[ericgibbs]

hi,
Have downloaded your data, will look thru it.

How do you have the PT100 and TVREF connected to the inputs of the IA.?

hi,
Do you have it connected like this.??

This is using LTspice.

 

[r.youden]

Yeah that is pretty much it. I have two 1mA constant current sources driving the PT100 and a 100Ohm reference resistor (with a pot to increase accuracy).

 

[MikeMl]

With R7 grounded (see OP's circuit), you must either run this configuration on split supplies, or you have to create an artificial offset between the differential inputs (on purpose), in which case the output will be linear but not down to zero volts.

In a single supply configuration, this type of IA usually has the right end of R7 connected to a positive well-regulated reference voltage, in which case the output is biased to be Gain*(Vin+-Vin-) centered on the reference voltage. The reference voltage can only be ZERO if the IA is operated on split supplies.

 

[r.youden]

That has been my fear. I can easily get a voltage onto R7 but it is getting rid of that offset again which is an issue. I need the output to be in the range 0-something. And there is no way I can get a negative supply running in there. Damn. Ah-well....

 

[ericgibbs]

Yeah that is pretty much it. I have two 1mA constant current sources driving the PT100 and a 100Ohm reference resistor (with a pot to increase accuracy).

hi,
The problem is that due to the IA configuration the top OPA output is trying drive below 0V.!
I have added a -12V supply to show this effect, you will have to bias the PT100 and Vref above ground/0V.

I would consider a bridge config of the PT100 and Vref

 

[MikeMl]

Why not just use a single-supply opamp as a Differential Bridge Amplifier. The response around zero will still be flaky. Just unbalance the input bridge a bit.

Here are previous threads on the PT100.

One


Two


Three

Just redefine your temperature range to be slightly wider. For example, instead of making 0degC = 0V, make -10degC = 0V