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.

ADS124S06: Thermistor Conversion Solution

Part Number: ADS124S06
Other Parts Discussed in Thread: ADS1220,

Hello,

I'm looking for a solution to the following:

- Convert resistance from 4 thermistors (10k ohms at 25C, +/-0.1C accuracy, resistance can vary between 400 ohms to 40k ohms for the temperature range I'm interested in)

- Sample rate >80 sps / thermistor, so >320sps total

- Voltage resolution < 80uV

- Single supply (3.3V or 5V)

I'm planning to use a resistor divider for the R to V conversion (Vref - 0.1% resistor - thermistor - GND)

I looked at ADS124S06 and ADS1220 but the Vpp noise spec for ADS1220 seems kind of high for my application.

Also, it seems that I need to disable the PGA for these single ended conversions. Do I need to buffer the input signals in this case? 

If I need external buffers (opamps) it seems more cost effective to use an external mux, and put the buffer between the MUX and the ADC rather than using four buffers.  Is this correct? 

Is there another solution (ADC/MUX/buffers) which is better suited for this application?

Thank you.

  • Hi Tavi,

    There are tradeoffs in performance from one device to another. The ADS1220 will have lower power, but a higher noise result. The ADS124S06 is lower noise and has an internal reference output that can be used as your voltage reference. Either device has external reference inputs as well in case the supply voltage is used as the reference so that the measurement is ratiometric.

    It sounds like you will be making a single-ended measurement of the thermistor relative to AGND. If this the case, then the PGA must be bypassed. This will lower the input impedance and a small bias current will be drawn. I propose a slightly different approach where the measurement is made differentially and will allow use of the higher input impedance of the PGA. Instead of measuring the thermistor directly, make an indirect measurement by measuring the resistor instead of the thermistor and then calculating the value of the thermistor. This measurement will be within the input range of the PGA as the resistor is biased above AGND. You cannot use this method with the ADS1220, but you could with the ADS124S06 as you would need 4 inputs from the thermistor and 1 input to the reference voltage.

    Best regards,
    Bob B
  • Hi Bob,

    Thanks for the fast reply.

    I thought about your recommendation and I'm seeing the following problem:

    To meet the PGA input requirement the voltage across the thermistor must be > 150mV (for PGA gain = 1). For a reference of 2.5V and minimum thermistor resistance of 400 ohms this means the fixed resistor must be smaller that 6.26K. With these values, the power dissipated in the thermistor is high enough that it starts to affect the accuracy.

    Another alternative would be to connect the fixed resistor between AGND and ADC_IN, the thermistor from VREF to ADC_IN and then measure the voltage across the thermistor. I'm a little reluctant to do that since I'll have to route Vref to quite a few locations in my system because the thermistor connections are spread around quite a bit.

    Any thoughts?

  • Hi Tavi,

    You might be just fine with the PGA disabled. There will be some bias current with the PGA enabled or disabled. The difference is a few hundred pA. As far as using the PGA disabled mode, there is still a buffer between the input and the modulator. It should be relatively simple to test to compare PGA disabled to PGA enabled.

    One other possible solution is to change the PGA enable on the fly based on the previous reading. If the thermistor resistance lowers to a certain point you can switch to PGA disabled.

    Best regards,
    Bob B