• TI Thinks Resolved

LDC1612EVM: Drift while reading data

I am trying to read data from LDC1612 EVM board and I tried with internal and external oscillator configuration. In internal oscillator configuration the noise is more than drift and whereas in external oscillator configuration the drift is more than noise. We are getting drift around 20 Hz to 1 KHz and as per my knowledge I think its due to crystal as I tried heating the crystal using hot air gun, the drift was more. I will attach the screenshot below. Can we zero the drift using differential coil configuration? because I am seeing similar drift in both the channels but not exactly same. I am stuck at this point. Please suggest me a way.

I am using 50 turns, SWG 22 copper wire as sensor and 47 pf capacitor in parallel with it. Have set the current code to 9 which is 60 uA so that the voltage peak is above 1.2 Vp and below 1.8 Vp. Sensor resonance frequency is 1.3 MHz ans inductance is 308 uH. 

First picture shows noise from internal oscillator configuration and second picture shows drift from external oscillator configuration.

  • Hi Mahesh,
    For best results, I would try external clock with multi-channel configuration. Please refer to Section 8.1.3 of the datasheet: www.ti.com/.../ldc1614.pdf.

    Best Regards,
    Bala Ravi
  • In reply to Bala Ravi:

    Yes and why do you think there is so much of drift while using external oscillator?, Because I am trying with all possible way but still I have not been able to suppress the drift.

    We are using it for extreme sensitivity and this drift is a big problem for that as we are not able to see the reading variation as this drift encountered is overtaking it.

    Thank you 

  • In reply to Mahesh J Shetty:

    Hi Mahesh,
    First off, you're measuring two different parameters. With internal clock, you're measuring frequency and with external clock you're measuring inductance. Does the temperature change drastically in your application? If so, you need to use multi-channel configuration.

    Best Regards,
    Bala Ravi
  • In reply to Bala Ravi:

    Thank you Bala Ravi for your response.

    And for the note, we are measuring with external oscillator configuration and not with internal oscillator configuration. I tried with internal oscillator just to see the difference from external oscillator configuration, So for just now don't consider anything regards internal oscillator configuration and lets talk about only external oscillator.

    Have you ever worked with sensing GUI provided by Texas instruments?, if not I am going to tell you briefly that , we are not measuring two different properties, it is that the sensor is made of copper coil with capacitor connected in parallel with it so that it forms a LC tank and this will be oscillating at resonating frequency so that if any metal is brought near or away from it, the resonating frequency changes and from it the value changes in the reading so in GUI, we have option to see Inductance value of coil, frequency and raw data. I can say that the frequency graph and inductance waveform exactly inverted to each other. 

    So here my main issue is Whenever we are taking readings, even though there is no external influence to the sensor, the value is either increasing or decreasing with time instead of staying straight. Here we are using two-channel configuration, one for taking reading and other as reference coil.

    And there is no temperature fluctuations as it is tested out in AC room with temperature 26 degree Celsius with plus or minus less than 1 degree fluctuations.

    do you think the problem might be due to the oscillator that is fixed on the board by Texas?

    and drift is increasing with time. Please help me to solve this problem.

    Best regards,

    Mahesh Shetty

  • In reply to Mahesh J Shetty:

    Hi Mahesh,
    I am familiar with the GUI. In the GUI have you tried the configuration page and set the test to multi-channel measurement? If you did this make sure you also set the correct external clock frequency. Please let me know after trying this and what issue you're facing. Please also share a screen capture of the settings used and the data measured or plotted.

    Best Regards,
    Bala Ravi
  • In reply to Bala Ravi:

    Yes I have set the test to multi-channel which is sequential measurement and have selected the external oscillator. But the issue I am facing is even though the coil is left free without any external influence or interface instead of getting constant reading with time, I am getting increasing or decreasing reading with time which I don't want.

  • In reply to Bala Ravi:

    We are getting graph like this instead of horizontal straight line. This drift we want to resolve.

    Best regards, 

    Mahesh Shetty

  • In reply to Mahesh J Shetty:

    Hi Mahesh,

                        Your setting seems right. If you look at the plot, the inductance is drifting between 33.115 uH to 33.119 uH. That's a percent change of 0.01%, which is completely acceptable. 

    Best Regards,

    Bala Ravi

  • In reply to Bala Ravi:

    Thank you Bala Ravi.

    Yeah but that data I attached previously is taken only for a few seconds which is less than 10 seconds of data I can say and if we consider it with time then, the drift is more for minutes or hourly basis. As per my test, for 10 minutes of data the drift was almost 200 Hz in frequency and the drift is increasing with time. So I need help regarding cancellation of this drift or figuring out the reason for this drift.

    Best regards,
    Mahesh Shetty

  • In reply to Mahesh J Shetty:

    Hi Mahesh,
    Can you please share a picture of your setup? Also can you please collect data for a longer period of time? I would like to see if the drift levels off at some point. Is there any significant temperature change in your application?

    Best Regards,
    Bala Ravi