• State TI Thinks Resolved
  • Locked Locked
  • Replies 9 replies
  • Answers 2 answers
  • Subscribers 30 subscribers
  • Views 1443 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.

Original question:

LDC0851: LDC0851

LDC0851: LDC0851 through metal face

User7311
Prodigy 185 points
Part Number: LDC0851
Other Parts Discussed in Thread: LMP91300, LDC1612, LDC1312, LDC2112

Hi, I am looking to use LDC0851 to sense a metal target through a metal membrane.  The coils will be inside a metal enclosure pressed up against a 0.006" thin metal face as part of the enclosure.  I am using 2 separate PCB coils (identical design from the same manufacturing lot), one for the sense and one for the reference.  To try and keep the coils matched due to one having a constant metal "bias", I have placed the same thin metal piece against the reference coil.  I cannot get the outputs to switch when a metal target approaches the sense coil. 

Theoretically the inductances should be matched since the sense and the reference see the same bias.  I have confirmed that the LC tank frequency with this configuration is within the datasheet operating conditions.  I am unsure if there is enough inductance change when trying to sense a metal target through this metal face that would cause the output to switch.  Has this chip been used in this type of configuration before?  I am interested in this chip because it is purely analog with little temperature dependence due to the dual coil design.  The metal face I am referring to is something like the below.  

  • 0
    TI__Mastermind 23810 points
    You need to calculate the skin depth of your thin metal sheet. There are a few factors that will help "see through the metal":

    - less conductivity of metal barrier
    - thinner metal barrier
    - lower excitation frequency

    You can use the "skin depth calculator" tab in our LDC tools spreadsheet to quickly run through these calculations. You want the "Percentage of current" to be as low as possible.

    www.ti.com/.../slyc137
  • 0 in reply to Clancy Soehren
    Prodigy 185 points
    The first two bullet points are fixed by the design parameters and cannot be changed.

    Regarding excitation frequency, a frequency of 1.5MHz will give me the desired ratio of "less than 0.5 skin depths" with respect to my metal barrier thickness of 0.006". This frequency, however, is out of range with respect to the design space of Figs. 9 and 10 on the datasheet. My inductance would have to approach > 20 uH to still be in the valid range, which may be a limitation due to space constraints of my design.
  • 0 in reply to User7311
    TI__Mastermind 23810 points
    Robert,

    You seem to have a good grasp on the constraints of your application. You can increase the inductance of your PCB sensor by increasing the number of PCB layers, decreasing the spacing between traces, decreasing the trace widths, and increasing the number of turns.
  • 0 in reply to Clancy Soehren
    Prodigy 185 points
    What is the maximum inductance and minimum frequency that the chip can support? All I see in the datasheet is Fig. 9 and 10 which don't go above 20 uH, and the maximum frequency values from Table 7.5.
  • 0 in reply to User7311
    TI__Mastermind 23810 points
    The minimum frequency supported is 300kHz. There is no restriction on the maximum inductance, although you will see practical limitations due to:

    - space
    - frequency + C + L combination
    - higher L can lead to higher Rs depending on sensor design, which will increase the current required to drive the sensor. The maximum drive current is 6mA for VDD = 3.3V.
  • 0 in reply to Clancy Soehren
    Prodigy 185 points

    Thank you I will try at these limits to reduce skin effect.  I also have LMP91300.  In general, does this device have advantages over LDC0851 in the application that I am trying to design?  If it can do better sensing through metal I can switch to this device.  However if the same performance can be achieved by both devices, and just strictly depends on coil design, I will stick with LDC0851 due to simplicity.   

  • 0 in reply to User7311
    TI__Mastermind 23810 points
    Robert,

    I do not recommend using the LMP91300. You can look at the LDC1312, LDC1612, or LDC2112 for more advanced LDC devices. However, these are all much more complicated than the LDC0851. You will also have the same problem of looking through a metal plate with all of them.
  • 0 in reply to Clancy Soehren
    Prodigy 185 points

    Hi, what are the reasons that you do not recommend LMP91300, aside from it being more complex than LDC0851?  LDC0851 does not operate below 300kHz, which I think is my problem, my inductance is currently around 300 uH. 

    I have not been able to see through the metal sensor face.  From what I have been modeling I need to be somewhere less than 50kHz, unless this has been successfully accomplished with this device at 300 kHz.  The other devices are too complex only because I will not be using a microcontroller.

  • 0 in reply to User7311
    TI__Mastermind 23810 points
    The LMP91300 only supports Rp measurements, which we have found to be difficult to use. For instance, Rp is highly sensitive to temperature changes. You'll notice that we completely moved away from Rp with all of our newer devices.