• State
  • Locked Locked
  • Replies 2 replies
  • Subscribers 47 subscribers
  • Views 661 views
  • Users 0 members are here
Support feedback
Options
Options
Related

This thread has been locked.

If you have a related question, please click the "Ask a related question" button in the top right corner. The newly created question will be automatically linked to this question.

ADS1148 thermocouple grounding problem

Aung Thet
Prodigy 195 points

Other Parts Discussed in Thread: ADS1148

There has been a problem with our ADS1148 based temperature measurement circuit (see above) while practically

using it in our product. 

We have been using it for RTD measurement (Ratio-metric conversion based on 330 ohms reference resistor) without any problem. For Thermocouple measurement, we simply remove the jumpers joining the AIN1,AIN3,AIN5 and AIN7 so that four thermocouples could be independently measured by differential configuration. (AIN0-AIN1 for TC0, AIN2-AIN3 for TC1, AIN4-AIN5 for TC2, AIN6-AIN7 for TC3).

We have pumped in a few milli volts using external power supply into the differential channels to test our register configurations and linearization calculations. The values are all correct.

However, when we really put in the Grounded thermocouples on the heating system where the thermocouple probe is connected to the Earth wire of the power, the readings stays low around zero even at the higher temperature. 

When I removed the Earth wire connection to the thermocouple probe, the readings increase proportionally as the system is heated up.

What really puzzled me here is that the ADS1148 circuit ground has no connection to the Earth wire of the AC mains yet connecting Earth wire to the thermocouple probe would still affect the measurements somehow.

Please suggest me ways to make the ADS1148 circuit configuration right with the grounded thermocouple. I can't use the un-grounded thermcouples in our product. We have to try to find a way to make ADS1148 work with the grounded thermcouple connected to Earth wire. (while the DC supply ground for ADS1148 is isolated from Earth)

  • 0
    TI__Guru** 113765 points

    Hi Aung,

    The problem here is that we do not know where earth ground is relative to DC supply ground.  The ADS1148 has a common mode input restriction where the input must be within the following:

    In your desired configuration you really can't isolate the supply from earth ground as the ADC must have a common point relative to both analog supply and earth ground.  One way of achieving this is to use bipolar supplies (+/-2.5V) where analog ground and earth ground are at the same potential. One other option, but may not be viable for your situation is to connect the thermocouple shield to a biasing resistor where the other end connects to earth ground.  This requires VBIAS to be used to connect the thermocouple into the common mode input range, and also earth ground and analog ground to be at the same potential.

    Best regards,

    Bob B

  • 0 in reply to Bob Benjamin
    Prodigy 195 points

    Hi Bob,

    Thanks for the reply.

    I still do not really understand what you mean by .."The problem here is that we do not know where earth ground is relative to DC supply ground.  The ADS1148 has a common mode input restriction"

    Can you please explain a little bit more on how exactly this common mode input range restriction is affecting the measurement? I had tried traditional differential amplifier and successive approximation ADC before. And the symptom of common mode noise as per my experience was that it showed random values shooting higher than expected value.

    For the case of my ADS1148 design, the symptom is totally different, it stays flat zero when the heated plate where the thermocopule is attached is connected to Earth Ground. 

    Is it possible to explain in simpler terms?

    Also, if possible, please explain more detail on the VBIAS to thermocouple shield method. Some circuits and suggestions on connections will do whether I should connect VBIAS to positive or negative terminal of the differential input,  I'd like to try that out too.

    Thanks in advance

    Aung