FDC2114EVM: Single ended sensor results with FDC2114 EVM

Hello,

Let me comment on the default configuration and then I will explain how to get your value correctly calculated.

The FDC2114EVM default configuration has 33pF in parallel with 18µH.  There is also about 10pF of board parasitic capacitance and pin input capacitance for a total of 43pF || 18µH, which produces a sensor frequency of around 5.7MHz. You can compute the expected DATA value from equation 8 of the FDC2114 datasheet. Using the following inputs:

• CHx_FIN_SEL = 1
• fSENSORx = 5.7MHz
• Frefx=40MHz (40MHz external oscillator selected with CHx_FREF_DIVIDER = 1)
• OUTPUT_GAIN = 4 (4 bit shift)
• CHxOFFSET = 6000

This produces a nominal DATAx value of 3339 and decreases in the presence of a proximity event. Note the max value that can be shown before saturation is 4096 (12 bits). If you are saturating at 4096, you need to increase the CHxOFFSET value to pull it within range.

For the example you provided, you added an extra 470pF in parallel, so you have effectively 513pF || 18µH (assuming you left the 33pF onboard). This produces a sensor frequency of 1.65MHz. Additionally, I would recommend to set the divider setting to 0x1001 so that both dividers are set to 1. There's no benefit to increasing the FIN divider unless you need to maintain FREF / FIN > 4, which isn't the case here. Then with the following settings you can compute the expected DATA value.

• CHx_FIN_SEL = 1
• fSENSORx = 1.65MHz
• Frefx=40MHz (40MHz external oscillator selected with CHx_FREF_DIVIDER = 1)
• OUTPUT_GAIN = 4 (4 bit shift)
• CHxOFFSET = 0

This produces a nominal DATAx value of 2703, which should be in range to be detected. Now if this is for a proximity sensing application then you want to use as small of a sensor cap as you can (33pF is typically recommended) and adjust your settings accordingly. This will allow you to get the most sensitivity out of the sensor because it minimizes the capacitive divider from the fixed capacitor.

Once you have valid settings you should be able to convert between DATA and your sensor frequency using equation 8 and then back to capacitance using equation 7.

Also note that you should be maintaining a sensor oscillation amplitude of between 1.2Vpk and 1.8Vpk on the IN3A and IN3B pins. You can adjust the amplitude by using the IDRIVE register which controls the amount of current that drives the sensor.

Hope this helps!

Regards,

Luke LaPointe

Hello,

In the single ended configuration your additional capacitor to ground is only connected half of the time. Assuming you previous example, the nominal sensor is 43pF || 18µH, which produces a sensor frequency of approximately 5.7MHz. A better way to state this is total sensor period is 0.175µs. If you probe the INxA and INxB lines you should see a half rectified sine wave on each line with a positive pulse width of 0.0877µs. Now if you connect a 470pF between INxA to ground then you essentially have 43pF || 18µH = positive pulse width of 0.0877µs for the INxB phase and 513pF || 18µH = positive pulse width of 0.301µs. The total sensor period will be 0.0877µs + 0.301µs = 0.389µs or 2.57MHz equivalent sensor frequency as reported by the FDC2114. If you use the GUI to back-calculate to capacitance it will not give the correct value because it assumes a differential sensor. So if you enter 18µH and 33pF for the values along with a frequency of 2.56MHz it will assume the total sensor capactiance is 213pF and that your sensor capacitance is 180pF which sounds like it's matching what you are seeing. 

Additionally, if you connect the sensor to ground in a single ended configuration the Rp value of each phase will be significantly different so the amplitudes could be significantly different. We recommend to operate both phases of the sensor between 1.2Vpk - 1.8Vpk. If one sensor is within range and the other is outside this, then you could get excess noise or poor performance. For this reason we typically recommend to connect the sensor in a balanced configuration which is between the INxA and INxB pins.

Hopefully this clarifies things!

Regards,

Luke LaPointe

Hello Andrew,

The FDC2114 can indeed be used for single ended applications. The key is that the imbalance caused by the single ended sensor should not be too great as to cause one of the phases to be below the recommended amplitude setting. Single ended sensors work well for proximity sensing applications where the single ended sensor electrode has a low capacitance and only changes ±20pF or so with the presence of a capacitive target. If you have significantly larger mismatches like 470pF then this could potentially cause issues and we'd recommend either a differential configuration or to add a fixed capacitor to the unconnected input to pseudo-balance the sensor. However in your case it sounds like adding 150pF of imbalance still produces a valid amplitude for both phases so you should be ok. Just ensure that you are computing the frequency based on the individual phases as mentioned in the previous post.

Regards,

Luke