• State Resolved
  • Locked Locked
  • Replies 5 replies
  • Subscribers 31 subscribers
  • Views 354 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.

MSP-EXP432P401R: noise

Laszlo Garab
Intellectual 965 points

Part Number: MSP-EXP432P401R

Hello,

I am following up slaa821 application note and testing the AD. The results are terrible I need help to identified noise sources.

Simple made a resistor divider 1kOhm-330Ohm and short the differential inputs on that. The divider is powered from batteries to avoid noise altougth I have nosie 166kHz and 10kHz as on the diagram.

Please help me with any idea how to avoid this problem.

Regards,

LAszlo

  • 0
    TI__Mastermind 49692 points

    Hello,

    How are you connecting the inputs to the LaunchPad? Are you are using a bread board and jumper wires? If so, you could easily implement some filters using resistors and capacitors to attenuate this noise. Depending on your ADC sampling rate and your signal frequency, you could be seeing aliasing too.

    Be mindful of any digital signals being used near the ADC inputs. You could also try moving your setup to a different room/location to see if it's related to any environmental noise.

  • 0
    Genius 3945 points

    -- in general, ADC inputs should be driven from a low-impedance source. Your voltage divider input is not suitable.

  • 0 in reply to James Evans
    Intellectual 965 points

    I have tried filtering, but 100uF with the 1kOhm have needed that the signal was acceptable (3-5 LSB noise). I also tried to drive the inputs with OPA350 which should fit for the purpose but the problem exists.

    What I have noticed the first 2-3 samples has very high variance (18-20 LSB) which conclude sampling periode should be changed, maybe. Unfortunately, I am still not comfortable the setup of the AD parameters, later I will come back to study the AD and the area of testing.

    Thank you for your valuable notes.

  • +1 in reply to Laszlo Garab
    TI__Mastermind 49692 points

    I would start with a simple ADC code example found in TI Resource Explorer. Then, you can come back to SLAA821. High external impedance on the ADC inputs can violate the ADC settling time and can cause noise. However, if that was the issue, you shouldn't have observed the noise when using the OPA350, since it should have a low output impedance. Perhaps it is an ADC configuration issue. Let us know what you find.

  • 0 in reply to James Evans
    Intellectual 965 points

    I am very sorry for being so negative with the MSP432 ADC. I have to change my mind!

    I have been working on MSP430G2ET and thought I should have increased resolution from 10 bit to 14 bit with oversampling and would have made sense to put some filtering in it, also. I started collecting samples summarizing them shifting to earn resolution and filter in the same time.

    I have spent a night tuning the system with analog filter, low noise power supply and so on, without success. After a time I started from zero and changed sampling time even shorter (which is illogical at first look I think) but got much better results. That was the time I understand my system is oscillating just like with MSP432 probably and it is more complicated then just converting analog signal.

    The system contains capacitance on the input and virtual poles on my digital filter (additionally more than one), so the system easily becomes unstable or simple gives no improvement in resolution. Playing over the night obtained worse and better results, finally I got what I have wanted.

    Conclusion, I had to understand a capacitive input AD must be handled carefully and digital filtering is not just summing and shifting. Poles and frequencies, fine tuning of the total system is crucial for success. It makes sense to read some more literature and practice it on Launchpads.