You set the value of the U2 pot plus R3 equal to the thermistor resistance at the minimum temperature of -5°C (for simulation purposes you can eliminate U2 and just use R3).
That will give 0V out @ -5 degrees.
Then you select a bridge voltage, Vb, to give the maximum output voltage you want at the +35°C resistance (formula given in post #31).

Thanks a lot,

I will try the same and let you know results.
I got overwhelming support for this thread.

Thank you all !!!

 

You set the value of the U2 pot plus R3 equal to the thermistor resistance at the minimum temperature of -5°C (for simulation purposes you can eliminate U2 and just use R3).
That will give 0V out @ -5 degrees.
Then you select a bridge voltage, Vb, to give the maximum output voltage you want at the +35°C resistance (formula given in post #31).

Thanks,

I modified the schematic for thermistor 103AT but I got non linear response through out temperature range so could you modified the schematic if I have done mistake somewhere.

I want to let you know that I have 5V and 3.3V power supply on my board.



Regards,

 

hi,
Delete the 1m in the .tran, [ leave it blank] it will run faster and also give a smoother curve.

It will not fix the linearity problem, crutschow will deal with that.
E

 

Hi

I have done some excel calculation and output never goes zero for Vb=5V.


Regards,

 

hi,
Delete the 1m in the .tran, [ leave it blank] it will run faster and also give a smoother curve.

It will not fix the linearity problem, crutschow will deal with that.
E

Hi,

We are running .op not .trans .

May be Mr. crutschow fix the linearity issue.

Thanks

 

I modified the schematic for thermistor 103AT but I got non linear response through out temperature range so could you modified the schematic if I have done mistake somewhere.

You also changed to a different op amp that is not rail-rail.
You need a rail-rail type input and output op amp to operate from a single 5V supply, such as an LT1368.

 

The LT6275 used in the sim is rated for a dual-polarity supply of ±4.75V min. You are using a single-polarity 5V supply.
That opamp is not a rail-to-rail type.

Edit: Crutschow beat me to it!

 

You also changed to a different op amp that is not rail-rail.
You need a rail-rail type input and output op amp to operate from a single 5V supply, such as an LT1368.

Thank you so much Mr. crutschow.
It was my mistake , I did not check Opamp parameter and I blindly pick and placed into schematic.
I found linear results using LT1386.
Now could you please clear me my two doubts:

1. In x axis I don't see resistance starting from 33900 ohm. What could be the reason ?
2. Output never reaches zero volt so is it mandatory for ADC to read 0V at -5degC or we can assure 0.5V for -5degC.
What is your point of view.




Regards,
Shiv Mishra

 

I have done some excel calculation and output never goes zero for Vb=5V.

Showing me the results without the calculations tells me little about why it's incorrect.
I suspect an error in your calculations.
The output voltage should be zero for a thermistor resistance of 33.9kΩ.

What is your point of view.

Not sure what the problem is.
And I don't know why your plot goes from maximum to minimum resistance while mine does the opposite.

Below is my simulation using just the end point resistances and it goes to zero (3.3mV) at 33.9kΩ.
(Note your 27k value in the list is in error.)

 

Showing me the results without the calculations tells me little about why it's incorrect.
I suspect an error in your calculations.
The output voltage should be zero for a thermistor resistance of 33.9kΩ.

Calculation I posted in post #67.

Not sure what the problem is.
And I don't know why your plot goes from maximum to minimum resistance while mine does the opposite.

Below is my simulation using just the end point resistances and it goes to zero (3.3mV) at 33.9kΩ.
(Note your 27k value in the list is in error.)
Why it is in error. I took these data from datasheet ?

I also unable to see 33.9K value in plot.
may be someone LT spice expert can help me.

Mean time could your LT spice file so that I can simulate and have a look if there is some mistakes in my simulation file.

Regards

View attachment 182710

 

Hi ericgibbs,

May be you can help for simulation results.
As in my post #68 resistance value in x-axis start from higher to lower while Mr. crutschow simulation shows lower to higher.

Regards

 

May be you can help for simulation results

hi mishra,
Be happy too, please post your problem LTS asc file.
E

Look at this link posted by @Jony130 sometime ago.
https://mathscinotes.wordpress.com/2011/07/22/thermistor-mathematics/

 

hi mishra,
Be happy too, please post your problem LTS asc file.
E


Look at this link posted by @Jony130 sometime ago.
https://mathscinotes.wordpress.com/2011/07/22/thermistor-mathematics/

Thank you so much . Its much useful information.

Pl find attached file.
Try to fix as per Mr. crutschow results and let me the problem definition of my LT spice file.
Thanks again !!!

 

hi,
Look at this option.
Note the zero trim pot.
E

The OPA supply can be set back to 5V

 

Given the intended end use of the measurement (post #38) I'm surprised that accurate linearisation is being attempted. I would have thought a simple input circuit (post #11) plus piece-wise-linear adjustment in software would suffice?

 

hi mishra,
I guess you realise that the plot in the latest post is only showing as a linear response because of the way it has been plotted with that Xaxis step.???

E

Update:

 

Given the intended end use of the measurement (post #38) I'm surprised that accurate linearisation is being attempted. I would have thought a simple input circuit (post #11) plus piece-wise-linear adjustment in software would suffice?

Agreed. Thus the link I offered in post #14.

Some like to complicate things.

 

Also agree, I would use software.
His temperature span is only 40 degrees, with a required resolution of 0.5C, so a Table of only 80 values would be required.
ADC count value/s as a Table pointer.

 

I guess you realise that the plot in the latest post is only showing as a linear response because of the way it has been plotted with that Xaxis step.?

Yes, that circuit makes the output voltage linear with resistance, not necessarily with temperature, so you still would need to do a software correction for that.

 

hi,
Look at this option.
Note the zero trim pot.
E

The OPA supply can be set back to 5V

Thanks for your reply.

Is trim pot essential in opamp circuit to set opamp gain ? because it would be difficult set the value of Trim pot value in mass manufacturing.

Still i am not able to do simulation as same as your attached waveform.
Could you guide me about setting of LT spice.

Regards,