LMV612: LMV612

Hello Jixia,

You said gain is R7/R6 ; I assume there are matching resistors on the IN+ side of op amp too. 4 resistor difference amplifier.

To reduce variation, use a low Vos op amp and high accuracy resistors. Consider changing op amp to a instrumentation amplifier (built in resistors).

https://www.ti.com/amplifier-circuit/instrumentation/products.html

Which one to chose depend of the source impedance of V1 and V2. 

Hi,

    I don't know how to paste this schematic diagram.

When i use TLV2387 ,but still get the erro input.

Hi,

 V3 is the voltage of R4

请问对于我的方案IB和IOS带来得输出误差，有相关得计算公式吗？

Jixia,

I don't know what different rows in the table mean. I expect the value of 2.5V power source also varies a little. 

For the circuit, it lacks balance and has far more resistors than required. For example R3,R4,R5 doesn't balance with R7.

What is the desired VOUT and temperature range? Temperature can be in ohms. The PT100 I'm using for reference is [0C , 100C] [100 ohms, 138.51 ohms] 

Hi,

 The PT100 I'm using for reference is [-400C , 200C] ,then [84.6 ohms, 177 ohms] 

And as you say my circuit, it lacks balance and has far more resistors than required,then can give some advice on the  balance,suah as how  R3,R4,R5  balance with R7?

Thanks a lot!

Hi,

The PT100 I'm using for reference is [-40C , 200C] ,then [84.6 ohms, 177 ohms] 

And as you say my circuit, it lacks balance and has far more resistors than required,then can give some advice on the  balance,suah as how  R3,R4,R5  balance with R7?

Thanks a lot!

Hi,

   Use R3,R4,R5 to generate 1V bias voltage.

Thanks a lot!

Jixia,

To be sure that I suggest a circuit with good linearity at what temperature is resistance 100 ohms?  In other words, I need a third point in the middle

Hi,

The  third point in the middle can is 70℃.

Thanks a lot!

Jixia,

I setup a constant current though PT100.  This current is Vref*(G-1)/(G*R1) ; where G =11 (IOP1) gain)

This is an accurate R to V convertor. 

So [-40C to 200C] becomes [0.641V to 1.34V]

The gain is set by both R1 that sets the current and RF/RG+1 which amplifies the PT100 voltage.

Hi,

    can get precision of 0.1℃.

Thanks a lot!

Hi,

R1,R2,RGand RF must be 0.1%.

Thanks a lot!

Hi,

I want to get  [-40C to 200C] becomes [0.641V to 3.3V],and how to modify this circuit.

Jixia,

Resolution of 0.1C = 2400 discrete values; that is easy enough.

Accuracy of 0.1C is another matter. Four wire wire kelvin is needed for the PT100; so I'll have to make IOP1 a difference amplifier.  How much processing power is in this application? The curve from -40C to 200C has a significant 2nd order component so that will need to mathematically corrected. I also expect the the application will need a calibration table. 

The output voltage range is naturally 2.092:1 (177ohms/84.6ohms); converting that to 5.148:1 (3.3V/0.641V) requires adding an offset and gain which will be an additional potential source of error. Is it worth it, I'd need to know more about the ADC (it's bits, reference, and range).

Let me know the processing power and calibration table capabilities. Also what is the temperature range of the application not including the PT100?

Hi,

The processing power  in this application is 3.3V .12 bit ADC of  STM32F030F4P6 ,and VREF is 3.3V.

Thanks  a lot!

Hi,

I use Single power supply for  LMV612MMX/NOPB.

Thanks a lot!

Hi,

I use Single power supply 5V for  LMV612MMX/NOPB.

Thanks a lot!

Jixia,

Using the ADC Vref is a good choice for PT100 amplifier's VREF.  In schematic, each VCVS gain of 1 is an unity gain op amp stage. Rtop & Rbot simulate the cable IR losses. 

To support 4 wire PT100 sensor cable, I had to make the current source it's own loop, so it needs a buffer to keep output current (1mA) immune to its output voltage.  Switch 1 and switch 2, can select PT100 , 82 ohm, and 162 test resistors (low cost in 0.01%) . All the many linear minors errors can be corrected. The ADC reads the output values for the two test resistors which allows making a linear formula for resistance versus VOUT. The PT100 is read and that voltage is run through the math formula (just created) to get an accurate value for PT100 resistance. Each input of diff amp also has a buffer to reduce initial error (before correction).  

I made the amplifier source 1mA and translate [84.6ohms, 177 ohms] to [0.1V to 3.2V] 

I have a excel sheet that does all the math for me (so easy). The only problem is that the 'b' reference is negative voltage. If I swap the diff amps inputs the 'b' reference will be greater than 5V which is also a problem.

To fix the 'b' reference issue, I changed to diff amp to allow an extra inverting input. This greatly complicates the component value math equations. So I did this change empirically. I set B to zero then picked R4 to get the desired offset. I added R5 to match adding R4. This increase diff amp gain a little so I knocked R1 down a notch to balance the extra gain. Current still near 1mA

All resistor are 1% except test resistors that should be 0.01% ; these are the resistors that can cancel out the minor errors in the circuit and Vref.

I'll leave it to you on how to convert resistance to temperature. To get 0.1C accuracy significant math will be needed as well as calibration table for the PT100 sensor assembly used. 

My circuit is still just an R to V convertor (with kelvin support and self calibration option)

Hi,

 What's the   voltage of 'b' reference .

Thanks a lot!

And i emulate the follow fig, then Vref is 1.8V , 

and  translate[84.6ohms, 177 ohms] to [1.5214V to 3.1827V] .

/cfs-file/__key/communityserver-discussions-components-files/14/PT100-constant-current.png

And I use R3,R4 and R5 to get the bias voltage 1V, then to match R7,how  to change  R3,R4 and R5.

Thanks a lot!

Jixia, diff amps have three inputs, IN+, IN- , VREF , see red box

For your original schematic, R3=768 ohm ; R4=165k; R5=249k ; Thevenin solution for 2.5V in, Output is 0.996V at 100k resistance

Hi，

In my original schematic, R3=768 ohm ; R4=165k; R5=249k,R7 is 100K？

Thanks a lot!

Hi,

In my original schematic, R3=768 ohm ; R4=165k; R5=249k,R7 is 100K？

 How did you get it?

Thanks a lot!

Hi,

When i use AS1120(Low cost) direct acquisition PT100,is a good idea?

Thanks a lot!

Jixia,

This FAQ has an excel calculator. I used the Reverse Thevenin sheet to give an exact answer for two resistors changing 2.5V to 1V at 100k ohms

When selecting available values for 1% resistors the output resistance was a little under 100k so I used R3 to bump it up closer to 100k 

[FAQ] Voltage scaling for almost any application (single ended)

I have no advice on PT100 choice.