# Is it a good idea to make one instrumentation amplifier with three operational amplifiers for thermistor sensing circuits?

Is it a good idea to have those three amplifiers with LM741 Operational amplifier? Has anyone done this so far? I thought of this as this was a cheapest idea. The instrumentation amplifier costs about $6 but the quad LM741 costs like$2. I know that buying instrumentation amplifier would be most effcient. But does anyone see any huge disadvantage in this making of instrumentation amplifier with three LM741 op-amps?

-
Please don't use '741s. – Ignacio Vazquez-Abrams Mar 2 '14 at 1:03
For thermocouple use? Not sure which type thermocouple you plan to use, but in general: no, absolutely not, don't waste your time, forget about it. For a lower gain application, sure, fine and dandy learning excercise. – Matt Young Mar 2 '14 at 1:14
sorry, not thermocouple but I am using a thermistor for accurately measuring the voltage across and inputing to the ADC of FPGA. I am still researching about the kind of buffer they have in Spartan 3E. If the buffer has really high input impedance and voltage requirement, I may not even have to use any amplifiers. – dr3patel Mar 2 '14 at 1:43
A thermistor is not a thermocouple, that changes things and you should edit the question. – Matt Young Mar 2 '14 at 1:46

It is possible, but the CMRR (common mode rejection ratio) will be worse than a monolithic InAmp. In a monolithic InAmp, the resistors R1-R3 are matched. Also, they are on the same die at the same temperature.

By the way, the above holds independent of the OpAmp model.

A Designer's Guide to Instrumentation Amplifiers has more on inner workings and rationale behind the design of InAmps.

Building an InAmp with 3 separate OpAmps could be a neat learning exercise**. You can plug in different OpAmps and compare results. You can compare CMRR with a factory-made monolithic InAmp.

** As Matt wrote in his comment. He had beat me to it.

-
I saw an app note from ADI or TI that compared a monolithic and homebrew INA. The results were shocking, just wish I could come up with a search term that would find it. – Matt Young Mar 2 '14 at 1:31
@MattYoung [This is more of a reply to your comment under the O.P.] It also depends on the temperature range, which the O.P. is interested in, and on the desired accuracy. TC measurement in a kiln around 700°C with accuracy of ±10°C are easier to pull off, compared to a medical thermometer at 37°C with accuracy of ±0.1°C. Most of the standard thermocouple types output between 20 and 70 μV/°C. – Nick Alexeev Mar 2 '14 at 1:36
@MattYoung: Sorry, What you mean by shocking! were the results of monolithic and homebrew INA were same or totally different? – dr3patel Mar 2 '14 at 1:45
@dr3patel The homebrew INA was much, much worse. – Matt Young Mar 2 '14 at 1:45
@yogece Found it, the comparison was actually in the AD620 datasheet. analog.com/static/imported-files/data_sheets/AD620.pdf starting at the bottom of PDF page 13. – Matt Young Mar 2 '14 at 20:51