Post Go back to editing

MAX31865 resistance measurement error

Category: Hardware
Product Number: MAX31865

Hi!

I'm working with the MAX31865. I built a prototype with the recommended circuit shown in the datasheet for 3-wires and I can communicate wit the IC ok.

The matter is that it is measuring with a big error, like a 15%. The error is either obtain with a PT100 or if I connect a fixed 100 ohm resistor.

For example, in the case of the 100 ohm resistor, it measures 115 ohm. The reference resistor is a high-precision 400 ohm.

Can anyone help me?

Thanks!

  • Hi  , it looks like this part will eventually be supported in the  Precision ADCs forum, but there's some replies to the effect that it isn't yet.

    In the meantime, can you share a schematic, layout if you can, and ideally, a high resolution photograph? I don't support this product but let's look for something obvious:

    • Is the supply voltage within the allowable range, not noisy, and well-bypassed?
    • Is the reference resistor within the allowable range, and have you measured it as well?

    -Mark

  • Hi Mark!

    Thank you very much for your contact.

    The schematic is the same as shown in page 24 of the datasheet: https://www.analog.com/media/en/technical-documentation/data-sheets/max31865.pdf

    I verified the supply voltage and the reference resistor and they are ok.

    What I've found is that current in reference resistor is different from the current in the RTD. From what is explaiend in the datasheet, this is not ok! I'm right? And if this is the case, what can be happening?

    Thanks!

  • Hi  ,

    While your schematic may be based on page 24, we can't really use that to debug your circuit - for example, Page 24 shows the minimum recommended local bypass capacitors (0.1 uF on both Vdd and DVdd), but nearly all systems will have additional capacitance. Most voltage regulators (nearly all) require considerably more capacitance, or they will oscillate. If your supply is oscillating, then that could be the cause of your problem. What is supplying the Vdd, and have you looked at it with an oscilloscope to make sure it's not oscillating and does not have excessive noise?

    Also while a long shot, what is your controller, and how far away is it physically? Excessive undershoot / overshoot on SPI signals can affect analog performance - if digital signals overshoot beyond the abs max ratings (300 mV below ground or 300 mV above DVdd), ESD protection diodes can conduct, after which, all sorts of unpredictable things can happen.

    How are you measuring the current in the reference vs. measured resistor? With a micorammeter in series, or with a voltmeter measuring the voltage across the resistor? Note that if the former ( microammerter), the meter's resistance will appear in series with the reference/measured resistor. Have you set the part for continuous conversion mode? I'm not deeply familiar with the operation of this part, but refer to the description of BIAS(D7) on page 13 - in single-shot mode, you can disable the bias between conversions, which will affect your external current measurements.

    Again, a photo would be a big help, circuit layout directly affects performance.

    -Mark

  • Hi!
    Yes, I will post a photo as soon as I'm with the board with me!

    About the current, I measured with a voltmeter to obtain the voltage across both resistors. The IC was configured in continuous conversion with bias always on.

    Thanks!