No announcement yet.

Thermocouple data wanted please

  • Filter
  • Time
  • Show
Clear All
new posts

  • Thermocouple data wanted please

    Has anyone plotted K thermocouple millivolts against temperature? I have a thermocuple but alas, my digital multimeter has no temperature readout. But I understand the voltage/temperature graph is a straight-line linear ratio which should make conversion a breeze.

    Data would be much appreciated.


  • #2
    Re: Thermocouple data wanted please

    Hi Robusto,

    I have some data for the K type , it has a temperature coefficient of about 40.28 microVolts per degree C , the reference data given at 100C is approx 4.095mV , at 400C 16.395mV and at 1000C 41.269mV ( this is with lead resistance of 6 ohms).
    Hope that helps



    • #3
      Re: Thermocouple data wanted please

      Hi Guys,

      Heres some very interesting and useful info about thermocouples. When you are using t/cs without a compensation device or a meter with one built in, you will need to either use your t/c with the addition of a "cold" junction, or adjust your measurements in accordance with a 0.0 degC reference.

      This site explains what you need to do pretty well though....
      Heres another one with a better explanation of whats required, etc...
      All the best,



      • #4
        Re: Thermocouple data wanted please

        Thanks for the useful data, Paul  and Mal.  I want to know whats happening in my Silvias boiler, but I know my data isnt accurate, which is why I enlisted your expert services.

        Ive done a very rough and ready experiment using an electric jug, k thermocouple, candy thermometer, and digital multimeter (DMM).

        NOTE: to avoid electric shock, clip the metallic thermometer to the side of the jug so it does not touch the heating element. Ditto for the thermocouple.

        Temperature readings from rising jug temperature were taken from the candy thermometer, with corresponding millivolts from the DMM, whose  probes touched the spade terminals of the k-t/c.

        The data range is:
        1.1mv at 40C
        1.5 mv 50
        1.9  at 60
        2.3      70
        2.7      80
        3.0      90
        3.3     100

        Now, placing the t/c on top of Silvias brass boiler, and using the corresponding readings from the jug experiment,  these are the results.

        The machine is allowed to stabilise for half an hour. Then the brew switch is pressed until the light comes on, indicating a drop in temperature and the beginning of the heating cycle. The brew switch is then released and readings taken every few seconds:

        Light on (heating) 83
        10"          77              
        20           77
        30           78
        40           83
        50           92 (light off)
        55           95
        60           100
        70           105

        Of course, the thermocouple is NOT in the boiler, but outside it, on top, so actual water temperature must vary from this.  

        From those readings, the best temperature surfing time is between 50 and 55 seconds from the light coming on.



        • #5
          Re: Thermocouple data wanted please

          Hi Robusto,

          Your method to correlate the t/c output to actual temperature measurements is very sound and so long as you keep the resulting correlation chart handy, it can be used any time you want to do a similar series of experiments in the same temperature range.

          If you wanted to use the same t/c to measure your Roasting Chamber temperature for example, you would probably have to chart another correlation curve through the temperature region that you would expect to see and then use this in the same way as you did for the Boiler. There probably wouldnt be a lot of error in just extrapolating your original chart out to the 150 to 250 deg C temperature range for this sort of thing but if accuracy is important then plotting a new correlation curve is the way to go.

          Great job Robusto and all credit to you [smiley=thumbsup.gif],



          • #6
            Re: Thermocouple data wanted please

            Thanks again, Mal.
            The volts v temperature graph is not quite linear -- otherwise it would be a breeze to calculate all sorts of readings.

            At the moment Ive stuck one of my several DMMs -- none of which has a temperature mode -- on the wall next to the machine. It would be great to have a digital thermometer, but most will only read to about 50 -70C.

            Im now getting to the stage where temperature surfing using stop watches is no longer necessary. Just go by the multimeter reading. If I sense the temperature will start to drop dramatically during an extraction, flicking the steam switch on should stop that.

            With enough practice it should be possible to get a very even and desirable temperature during extractions.



            • #7
              Re: Thermocouple data wanted please

              Just one word of caution. As thermocouples measure relative temperature only, if you dont know one of the junction temperatures (usually the cold junction temperature), your results will vary with the ambient temperature. If you make a calibration chart, note the ambient room temperature. When youre making your measurements on the Silvia, just make sure that you adjust the calibration chart either up or down depending on the ambient room temperature at the time. If you have climate stabilization, then you should be right.

              BTW: you can make a simple cold junction compensated thermocouple amplifier with a single op-amp and a few other bits and pieces. Ill be doing just that in a few months in the quest for a cheap PI(D) temperature controller.