Most LDC devices measure shifts in inductance by measuring the resonant frequency of an LC tank circuit. For this reason, it is most useful to calculate LDC resolution in units of Hz. However, many users will spec their desired resolution in terms of the distance between a target and the sensor coil, and also want to determine the SNR of the frequency shift associated with a target movement. For these reasons, this section will give a set of steps to use FEMM to simulate the sensor inductance, and use the LDC EVM and software GUI to estimate the SNR.
- Calculate the Frequency:
- A real inductive sensor will consist of the sensor
coil and fixed, physical capacitor. FEMM accurately calculates the
magnetic fields of the coil - and therefore its inductance - but it will
not include the effects of the capacitor.
To work with this
limitation, we describe two sets of steps below. The first will
assume the sensor coil geometry and the capacitor value are known,
but the coil inductance and tank circuit resonant frequency are
unknown. The second set of steps will be be based on a known coil
geometry and tank circuit resonant frequency, but the coil
inductance and the capacitance are unknown. This will allow you to
simulate the coil in FEMM to determine its inductance, and then
calculate the capacitance.
Neither approach
considers the effect of the capacitor into the FEMM simulations.
Instead, a few calculate/simulate steps will be used to to quickly
iterate to the final inductance value and resonant frequency.
(1) - Known Coil
Geometry and Capacitor Value
-
For the first
iteration, pick a frequency and a maximum target distance
and complete Part 1. Use the FEMM .ans file to calculate the
inductance.
- Use the known
capacitance and the newly calculated inductance to calculate the
corresponding resonant frequency as shown in Equation 2:
Equation 2.
- Use the new
frequency and the original target distance in FEMM to calculate
a new inductance.
- Repeat steps ii and
iii until the inductance does not change significantly between
iterations.
- Move the target by
the minimum distance for the desired resolution and repeat steps
i-iv. Determine the difference in calculated frequencies between
the two target positions.
(2) - Known Coil
Geometry and Tank Circuit Resonant Frequency
- Complete Part 1 using the
resonant frequency and target distance. Use the FEMM .ans file to
calculate the inductance.
- Calculate the
capacitance using:
Equation 3.
- Move the target by
the minimum distance for the desired resolution and perform the
steps i - iv in (1) just above. Determine the difference in
calculated frequencies between the two target positions.
- Calculate Resolution Using Simulation Results and
Noise Floor Measurements:
- Measure the Noise Floor:
For inductive sensing applications, the first step is to measure the
noise floor for the system. Basic noise floor measurements can be done
using the LDC evaluation modules, which can be purchased on ti.com.
Accuracy can be improved - with respect to the end application - by
making measurements over the expected temperature range, and using a
sensor designed for the end application. The default evaluation module,
however, will still provide a reasonable noise floor estimate.
To measure the noise
floor using an evaluation module, first download and open the Sensing
Solutions GUI. For more detailed installation and use instructions, see the
user’s guide associated with the EVM.
- Open the GUI and
navigate to the Configuration tab.
- Configure the RCOUNT setting so that the LDC device
will sample the sensor frequency at the desired rate.
- For non-default sensors, the user must use
an oscilloscope now to verify that the sensor oscillation
amplitude and the sensor frequency are within the acceptable
range for the specific device, as specified in the
datasheet.
- If the sensor
oscillation amplitude is outside of the recommended range,
the sensor drive current will need to be modified. Please
see Setting LDC1312/4, LDC1612/4, and
LDC1101 Sensor Drive Configuration for
instructions.
- If the sensor
oscillation frequency is outside of the frequency range
supported by the device, either the inductor or the
capacitor in the LC tank will need to be modified.
- Next, navigate to the
Data Streaming tab.
- Click the Start Streaming button to begin the
noise floor measurement.
- Ensure that the graph displays Detected Sensor
Frequency (MHz), which is controlled by the drop-down menu
in the upper left corner.
- Click on the Show Statistics button to
display the average and standard deviation. The data's standard
deviation will be used to estimate the system noise. Increase the
number of decimals of the standard deviation until the last digit is
rapidly changing.
Log the data in a .csv file and use your favorite analysis tool (e.g.
Excel, MATLAB, etc.) to compute the standard deviation over all variations,
including temperature. In general, the noise floor measurements will be more
accurate over longer measurement periods.
- Calculate the Final
Resolution via Signal-to-Noise Ratio (SNR):
To calculate the SNR, use Equation 4, using three times the measured standard deviation as the total
noise.
- Remember that the SNR should be at least 10 to
achieve the desired resolution in a real-world application.
- Note that
resolution significantly increases when the target is closer to
the sensor. The FEMM simulations should test the resolution at
both the minimum and maximum target distances using one of the
two steps in (1) just above.
- Also note that very
fine resolutions may not be achievable at high sample rates. For
more information about how sample rate affects resolution, see
the Optimizing L Measurement Resolution for
the LDC161x and LDC1101 application
report.
- If the FEMM
simulations show that the desired resolution is not viable, then
the sensor design or the target design must be modified. LDC Sensor Design and LDC Target Design are good
resources.
Equation 4.
- SNR is the
signal-to-noise ratio of the inductive sensing system
- Δf is the
calculated frequency shift from 1
- σN is
the measured standard deviation of the noise from 2.a