• State
  • Locked Locked
  • Replies 3 replies
  • Subscribers 47 subscribers
  • Views 2592 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.

50 Hz noise

Vaidyanathan Balakrishnan
Intellectual 705 points
Other Parts Discussed in Thread: INA326

Hi All,

I am working on recovering spectral response in the 5 Hz to 250 Hz band on signals which consist of a dc component from 0.5V to 2V and an ac component from 5 mV to 30 mV.


The system is powered from a 24V SMPS followed by isolated DC-DC converters to generate system 5V and 3.3V. The front end is a INA326 device followed by OPA188.

To recover the AC component, I have a DC blocking capacitor followed by a sallen key low pass filter cornered at 2 Khz. and then the signal is amplified to be 0 to 3V (centered at 1.5V) before it is fed to the microcontroller (TI, Hercules family).

The problem I am having is that there seems to be a 50 Hz and 100 Hz component constantly present in the FFT results.  When the AC amplitude is above 15 mV, this component is low and I am able to recover the actual frequency. When the AC amplitude falls between this, the 50 Hz and 100 Hz signals dominate.

I am not able to figure out where these signals are coming from and how I can get rid of these. The problem is that these two frequencies fall in the signal range also!


I am desperate for any suggestions!

  • 0
    TI__Genius 12325 points

    Hello,

    50Hz and its associated harmonics are the result of mains interference (the electrical grid in your country most likely runs at 50Hz). Eliminating this low frequency interference from sensitive measurements requires a combination of careful circuit design, PCB layout, good cabling, and potentially enclosure shielding. If you could include a schematic of your circuit I can make recommendations on how to improve its noise immunity. To start, the INA326 should be the first amplifier in the signal chain and its input should be a differential signal, shielded twisted pair cabling would be ideal for the input cable. 

  • 0 in reply to John Caldwell
    Intellectual 705 points

    Hello John,

    Thanks for the response.

    1. Yes, mains are at 50 Hz frequency in my country (India).

    2. Yesterday I tested by giving power from a 24V battery, and it worked like a breeze, so the issue is coming from the mains.

    3. I tested 4 of my cards: 1 card, I do not know what we did right, but it works like a dream. The 50/100 Hz component in the signal is very low and all our measurements come out right. 2 cards perform averagely and in 1 card, the 50/100 Hz peaks dominate the spectrum. This test was done with a linear 24V power supply.

    4. I am including the analog part of the circuit . The signal I am using is AIN_AC and we have to get the component of the input down to 5muA (that is 5 mV across the input 1K resistor).

    5. In this circuit, we have found that we get better noise response if we do not use the INA326 and that the INA326 keep0755.50Hz Problem Circuit.pdfs blowing out once in a while.

    Regards,

    Vaidi.

  • 0 in reply to Vaidyanathan Balakrishnan
    TI__Genius 12325 points

    Vaidi,

    The INA326 is most likely blowing because the inputs are not protected from transients of both polarities. For example, diode D9 does not protect against positive transients on the negative input. A positive transient on this input would forward bias the internal ESD diode of the INA326 (and potentially destroy it) long before the combination of R29 and D1 clamped the input voltage. This assumes that you are in fact using BAS16H diodes. According to the datasheet I found for them ( http://www.nxp.com/documents/data_sheet/BAS16_SER.pdf ) they are not zener diodes as your schematic symbol for them would suggest. 

    The fact that one of your boards performs well, while others do not suggests to me a manufacturing defect. Perhaps there is a cold solder joint which is reducing the effect of shielding or grounding. Is it possible to re-flow the defective boards and check if the performance improves?