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.

MSP430 A/D input noise / instability

Other Parts Discussed in Thread: MSP430F2012

We see noise / unstable values when we A/D convert one of the analog inputs on an MSP430f2012.  This input is being directly fed a clean DC voltage through a meg-ohm resistive divider.  Do we need to add a low output impedance buffer between the resistive divider and the A/D analog input to reduce this noise?

  • Robin Padden said:
    Do we need to add a low output impedance buffer between the resistive divider and the A/D analog input to reduce this noise?

    I tend to say 'yes'. Why? The ADC input has an input impedance of several kOhms, but compared to a meg-ohm devider it is quite low. Also, the impedance changes periodically, when teh samplign input is opened and closed. Then there is some inrush current. into the sample&hold capacitor.

    If your soure is changing very slowly, you can add a capacitor to the ADC input which will compensate for mos tof the sample inrush current. You can also increase samplign time to allow the signal to settle.

    The users guide (at least the 54xx, or was it the device datasheet of the 54xx) has some calculations about input switch resistance, smaple&hold capacitor and signal settling times.

    Also, on such fragile sources, it is recommended to put the CPU into sleep mode during conversation, so the intermittent current of the CPU does not disturb the conversion. (the CPU will wakeup if the ADC is programmed to trigger an interrupt after conversion)

  • Hi Robin,

    High-impedance sources are tricky.  They are more susceptible to EMI, and they have almost no tolerance for even the tiniest bias currents coming from the ADC inputs.

    But if you haven't already double checked your sampling time (an ADC setting), you should.  You should make sure you are sampling the signal for long enough.

    If you are sampling long enough, then you could also consider adding a capacitor to the signal you are measuring.

    Good luck!

    Jeff

  • Like any SAR ADC, the input impedance of the ADC10 has a rather odd characteristic. As the sampling gate opens there is a sudden in-rush of current to the sampling capacitor, which can load a megaohm impedance quite a lot. If you leave the gate open for a long time you should eventually get a fairly accurate reading, but you may see a lot of noise pickup from such a high impedance source.  A better approach is to put a capacitor between the ADC input and ground. Make sure you choose a nice quiet ground point, or the capacitor might just couple more noise into the ADC. If you are only measuring a DC level you can make the capacitor fairly large. This will allow you to use a much smaller sampling time, and the large capacitor will smooth out most of the noise you may be picking up.

    Steve

     

**Attention** This is a public forum