Post Go back to editing

AD7147: Using an External Sensor With the Evaluation Board

Thread Summary

The user is experiencing noise and fluctuations when reading capacitance values from an external AD7147 chip on a PCB. The solution involves adding decoupling capacitors to Vdrive and Vdd, and ensuring unused CIN pins are not externally connected to BIAS but are configured internally. The user should also read the CDC_Result_Sx register for accurate measurements.
AI Generated Content

Hello,

I have the AD7147 evaluation board, and I'm trying to use it to read capacitance values from an external AD7147 chip on a PCB. I'll include below the schematic I am using for my external sensor. The external sensor is designed to measure force, a  copper pad on a PCB connects to CIN0 of the AD7147. This copper pad is covered in a layer of dielectric (silicon) and then a piece of aluminum rests on top. By applying force to the aluminum (and squishing the dielectric) I expect a change in capacitance. If there is a problem in my schematic design please let me know.

I have been able to connect the external sensor PCB to the evaluation board, and they are communicating, however there is a lot of noise and fluctuations on the output. Without any load the values jump between 32500 and 35500 randomly. Applying force to the sensor also does not really seem to be changing things. 

The external PCB has two copper pads, one connected to CIN0 and one to CIN2, (the blank space of CIN1 was just to make soldering easier). The rest of the pins are connected to BIAS as suggested in the data sheet, with the exception of CIN1. CIN1 is connected to nothing, this again was just to make soldering easier. Is it possible this is causing a problem?

Right now I have set up the evaluation software as follows:

Register Configuration> CIN Configuration>Stage 0, set CIN0 to +input path, all others to Bias. Then clicked Get Offset.

Register Configuration> CIN Configuration>Stage 1, set CIN2 to +input path, all others to Bias Then clicked Get Offset.

All other stages have everything set to BIAS, but I am only using stage 0 and stage 1, correct?

Then I am going to Register Configuration>CDC converter to read the CDC. I was looking for a way to directly read the CDC_Result_Sx register, but couldn't find it. Is this the same thing though?

Thanks,

Dave

attachments.zip
  • Hello Dave,

    A few comments on your schematic.

    I do not see any power supply decoupling capacitors on Vdrive or Vdd, do you have capacitors on these supply pins ? Not having these capacitors could cause large amounts of noise in the measurements.

    I think you may have misinterpreted the datasheet where it reads "unused Cins connected to Bias".

    For Cins that are not used, the pins should be left unconnected and the stage configuration should be such that those Cins are connected to bias internally on the chip. You should remove the external connection to bias of the unused CIN pins. This is likely the cause of your problems.

    The result register for a single stage can be read as shown below.

    Regards,

    Maurice.

  • Hi Dave, Ideally, decoupling should be placed as close to the chip as possible, however, if decoupling is present on another board then there will be some decoupling, enough to ensure functionality of the chip but not perhaps optimum performance. I think the external connections from CINs to BIAS is the root of your issues.

    Maurice.

  • Hey Maurice,

    My understanding was that I could use the decoupling capacitors on the evaluation board, however I see I must have misinterpreted that as well. Thank you very much for your assistance.

    -Dave

  • Hi Maurice,

    One last question. The example circuit shown in the AD7147 datasheet and the circuit shown for the AD7147 evaluation board data sheetdon't totally match up. Should I add decoupling capacitors to both V_drive and V_ad7147 (Vcc)?

  • Hi Dave, The recommendation is to place a 0.1uF decoupling capacitor on both Vdd and Vdrive as shown on page 26 of the evaluation board document.

    Maurice.