ZTP 135 SR circuit

did you ever solve your questions on ZTP-135?

Hi, Mark. I see your question dates back more than two years, hope you could solve your doubts in the meanwhile.

Anyway you can try this: http://portal.fke.utm.my/fkelibrary/files/dhiyafarhanabintiramli/2012/404_DHIYAFARHANABINTIRAMLI2012.pdf

The circuit used is exactly the same you are referring to.

I have been working with the ZTP 135 SR thermopile for the past couple months and I am having difficulty with the calibration of the thermopile component of the sensor. Resources on circuit design for thermopiles are abundant but the calibration is not and is usually specific to the instrument. This is the only forum I have found speaking about the ZTP 135 SR. @michelangelo coca  I have read that paper and have duplicated its construction. The rest of the reading is broken english and not really helpful for the calibration. I have been able to calibrate the thermistor component using the datasheet resistance measurements at specific temperatures and using the relationship of resistance to temperature. 1/T = A + B*log(R)+C*log(R)^3 where T is the temperature in kelvin and R is the resistance of the thermistor.  The constants A B and C I solved for using simultaneous equations three unknowns on Matlab. The thermopile calibration is more difficult. I cannot seem to get the constants for the equation Vo= k * ( To ^(4-g) - Tamb ^ (4-g)) where Vo is the voltage output of the sensor, To is the object temperature, Tamb is the thermistor reading temperature for the cold junction, k is the proportionality constant and g is a correction to the power law.

     I have used one rail to rail OP amp but have seen that using an additional amplifier for summing is used to add the voltage from the thermistor to the output voltage of the thermopile .  Running this on arduino yun microcontroller thought I could just do this in the program adding the output of thermistor circuit to the output voltage of the thermopile. Any ideas of how to get proper output voltage for the thermopile calibration equation so I can solve for the constants k and g? I can attache circuit diagram and script for arduino if needed.

peteroakes or DAB can you help?

Thanks,

Clem

So the math illudes me a bit, it has been a very long time since I had to do that kind of math

but examining the circuit and assuming it works (It looks like it should), the opamp has a gain of atleast 100 or possible way more in the configuration shown, the thermopile will generate only a small voltage (mV) so the opamp is needed no matter what and the ouput of this should be sent to the micro, in the same way as the origional project did, unless you have a micro that has a gain stage that can amplify over 100* then you must use the op amp as shown int he sample schematic, your micro will nevery have enough resolution to resolve the temperaure without it. And if your doing the op amp, you may as well hook up the thermister as shown too, you can still read the thermistor output but it will automatically compensate for ambient without the micro

the thermistor output range is significantly reduced by limiting its supplied voltage by strapping it across R2 which is in the middle of a resistor divider chain 1K, (R2) 200ohm, 200ohm,  and this configuration is designed to limit the compensation to a similar range as the thermocould will generate I would think.

based on this code shown and the result graphs, the input to the micro is about 1V ish, so you could increase the value of R6 to maybe 200K to increase the gain and improve the overall dynamic range, could even go to perhaps a 470K if your feeding into a 5V controller and the reference is set to the VCC (Default on an UNO), remember the PIC will have been using a different VREF that equates to the 1023 reading so you will have to adjust the value when running on the Yun

the code clearly shows the math involved and would require very little change to get it working from the PIC it was written for, there bot C and to get the LCD working you would just replace it with the approprite library

The circuit works fine and I get output to the microcontroller. The problem is that the output is not linear. The data that the thesis above states is a linear correlation between the output voltage and temperature but this cant be for this kind of sensor which needs to be some form of power function. Do you know of anyone that calibrates thermopile sensors? The constant g or gamma from my above equation is an adjustment to the power 4 so must be in a range from 0 to 4 and the k constant I have seen on the order of 10^-4. I large error in the voltage will be significant to the calculation of the Temperature which I would like to be within +- 1 degree Celsius. 

I know this post is more than three years ago, but I have a similar issue. The hyper-link above are broken. Could you kindly show the schematics that you are talking about? Thank you very much!

Me too, I have the same problem, I can't open them and I don't have enough information about this sensor