• State Resolved
  • Locked Locked
  • Replies 5 replies
  • Answers 2 answers
  • Subscribers 31 subscribers
  • Views 296 views
  • Users 0 members are here
Support feedback
Options
Options
Similar topics

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.

LDC1614: LDC1614 measurement of coil temperature

Jo
Intellectual 290 points
Part Number: LDC1614
Other Parts Discussed in Thread: LDC1612,

Hello there,

If one want to compensate for L or Rp changes due to inductor temperature change, the usual approach is usage of a
differential coil system, or a dedicated temperature sensor for that matter, 

I think this works well if one can ensure good thermal matching characteristic between target sensing coil and 
temperature sensor (or differential coil), which implies that those components have to be placed close together
but not too close in order to minimize interferences etc. So if there is a panel of 6 target sensing coils already
close together, it might be challenging to find a good place for differential coils.

My question, would be feasible, or even a good idea, to measure the temperature of the sensing coil more directly?
One approach might be to multiplex the target sensing coil between a LDC device and an independent resistance
measurement facility, by means of analog switches. Since most LDC devices need a MCU anyway, this should not
be too complex. Component count would increase of course, but my point is that those components don't need 
thermal matching with the target sensing coils, hence less space constrains to consider.

Just wondering about opinions on this approach.

Regards,
Jo

  • 0
    TI__Mastermind 36005 points

    Jo,

    I have never heard of anyone implementing this approach before.
    From what I recall, analog switches and MUXs have their own temp behavior to consider, and their RON behavior may be a factor as well.
    As an example, the curves below show the RON behavior over voltage, with temperature held as a parameter.
    They show RON is not constant or linear, and there are some relatively large shifts in RON over temperature; this will impact the sensor signal and the subsequent frequency and temperature estimates.
    Regards,
    John

  • 0 in reply to John Miller - Sensing
    Intellectual 290 points

    Hi John,

    RON would be a big concern. Initially I thought it might be possible to use an unused switch of the same IC dye
    as a reference switch, to obtain the resistance. But I am not sure how closely the values would match with the
    other switches.

    Overall, when considering increased BOM and more advanced firmware requirements, this approach seems
    a bit over the top. 16 bit ADC it likely required too, not available on many MCUs. Maybe this is still a good idea
    for systems requiring larger coils, lets say diameter > 50mm. Each reference coil would take considerable PCB
    real estate.


    Regards,

    Jo

  • 0 in reply to Jo
    Intellectual 290 points

    Hi John,

    What about a capacitive isolation of the LC tank? DC current applied through PCB coil for purpose of resistance measurement,
    and AC current as generated by LDC device. Then no need to use analog mux.

    Regards,

    Jo 

  • 0 in reply to Jo
    TI__Mastermind 36005 points

    Jo,

    I have never considered that approach, and am pretty sure you need that DC current path.

    In reference to my earlier reply, I found a reference design that uses analog MUXs to share a single LDC1612 among a 16-button keypad. 
    Granted, this is for a keypad, which is different than for a linear sensing application, but I thought it might be good to mention.

    Regards,
    John

  • 0 in reply to John Miller - Sensing
    Intellectual 290 points

    Hi John,

    Yes you are right, a DC path to LDC1614 is needed all the time, otherwise no oscillation. I was trying a configuration
    to bypass additional series capacitors with a switch to enable oscillator startup, but when switch is opened afterwards,
    oscillation is not maintained anymore.

    Thanks for your additional information of the mux application- have seen it before.

    Temperature dependency curve of Ron could also be more flattened by using switches in parallel,
    but I suggest we let it there for the time being. My actual project is only about button press detection,
    and the usual baseline tracking approach is working fine for me.

    Regards,
    Jo