Measure temperature without a microcontroller

[vukhanh97na]

Hello everyone . I'm having a project about : Use thermorcouple to measure temperature without microcontroller(0 - 200 oC) . I researched and i think : " use opamp and icl7107 for display " . But i have 2 problem :
- I can't amplify exactly signal from thermorcouple (because it's very small ~uV) .
- I don't know how to display icl7107 with single ended input . ( i think i designed as datasheet) .

So i hope tobe given some specific guidance from you.

Sorry , 'cause my writing isn't good enough.

 

[sparkie]

Can you give more information about your application?

Off the top of my head, you could use a pid controller. It is a packaged unit that can display the temperature and give various outputs based on temperature setpoints.

 

[Averagesupernova]

I used to work for a company that built gauges to measure exhaust gas temperature using a thermocouple. A pair of op amps were used which fed the A/D input of a microcontroller. You obviously will have to find a way without using a microcontroller, but you can certainly use google to find an appropriate op amp to use. I am sorry I can't recall which op amp was used but I can tell you that amplifying a signal from a thermocouple was in the application notes. It was taylor made for it.

 

[jrmichler]

Try search terms thermocouple signal conditioner or thermocouple signal transmitter. Two companies that come up are Omega (www.omega.com) and Automationdirect (www.automationdirect.com). I have had good experience with both of those companies.

Omega also has several panel meters that connect directly to thermocouples. Their DP20 (https://www.omega.com/pptst/DP20.html) will work with ten different types of thermocouples, and costs about the same as a thermocouple signal conditioner.

On the other hand, if your project is to design and build a circuit to read a thermocouple, I hope that you have been studying about isothermal junction blocks, cold junction compensation, electrical interference, impedance, and several other topics related to making a thermocouple actually work.

 

[Baluncore]

You will need a reference temperature or a compensator, an amplifier and display.
Here is an example low-power circuit, with a kit of parts.
http://www.herdware.com/shop/analog/thermocouple-amplifier/
Follow that with any analogue meter or DMM.

 

[Svein]

(http://www.next.gr/meter-counter/th...it-diagram-of-Temperature-Monitor-l59900.html)

 

[vukhanh97na]

You will need a reference temperature or a compensator, an amplifier and display.
Here is an example low-power circuit, with a kit of parts.
http://www.herdware.com/shop/analog/thermocouple-amplifier/
Follow that with any analogue meter or DMM.

thank for your help . i think i know how to ampify . I used INA122 . But i can't display icl7107 . I connected single ended input to Vin+ (of icl7107 ) , GND to Vin- . But it didn't , The simulation didn't run . But when i use Battery ( replace for thermoucouple ) it work . i don't know where i fail .

this is the first time i do it . sorry if my question make you feel uncomfortable.

 

[Baluncore]

A thermocouple produces a low voltage. There are also thermal voltages produced where you connect it to the amplifier. Unless you compensate for those other voltages, your temperature measurements will drift. The simplest way to elliminate drift is to use two thermocouples back to back which eliminates the connector thermal differential voltages. You then accurately measure the temperature difference between the two thermocouples. Only when one is in melting ice will you get an accurate reading.

The other way is to use only one thermocouple, but with a thermal compensation voltage, so you don't need the bucket of ice. That is done in the circuit and kit I referenced.

The ICL7107 is a differential converter designed to run on a single 9V battery. The common voltage is set internally by the chip, so you must reference your input to the common, (pin 32). The amplifier will need to amplify relative to that common reference provided by the chip.

If I remember from 40 years ago, connecting the reference inputs to the signal inputs of the ICL7107 chip should give a reading of 1.000; Check that on your simulator. The display goes from 0.000 to 1.999 which will become 000.0°C to 199.9°C.

What type of thermocouple will you use ?
A common type K thermocouple, = chromel–alumel, produces 41 µV/°C, which for a 200°C range produces a total of only 8.2 mV. Without amplification, the differential reference voltage would need to be half that = 4.1 mV.

You could wind the converter reference voltage down to only 4.1mV and see how it goes. Conversion noise can be reduced by using a higher reference voltage, say about 250 mV. That suggests you need a gain of about (250mV/4.1mV) = 61 times. Your example circuit shows a gain of only 3.

I will not do all your work for you. You need to read and understand the way the 7107 generates it's common voltage and how the quad-slope converters work by switching between inputs. Get a copy of the full data sheet and read up on reference voltage range etc. Maybe it is time I had another look at that datasheet.

 

[jim hardy]

I really like that analog kit that @Baluncore found.

In the really old days we used a hand built d'Arsonval meter movement .
You had to use leadwires with correct resistance.
Cold junction compensation was by a temperature sensitive zero adjust spring in the meter movement itself.

Holy Cow ! It looks like somebody still makes one !
http://ptc1.com/pdf/326.pdf

old jim

 

[Chris Riccard]

If your looking at set temp use a comparitor and a threshold setting resistor.

 

[Baluncore]

This is probably the lowest cost solution for type K compensation and amplification. Approx US$5.00 inc postage. (see also post #6, above).
https://www.ebay.com.au/itm/AD597-K-type-Thermocouple-Temperature-Control-Board-Module-Ultimaker/192177271302