Hi Tom,
Can you give some more detail? What seems to be working, and what not?
Are you sure the cap is necessary? Can you check the system clock? Did you also check with a scope plot?
When reconfiguring pin 7 as output, all current will flow to the (huge!) capacitor, resulting in a voltage drop to near 0. The code will need to charge the cap slowly.
Via NFC, only 1.8V can be supplied to the IC. Think of the NF controller as a power supply of 1.85V, with a resistance of about 200 ohm in series. The code must make sure the load is not too high, to prevent a too sever voltage drop.
An external capacitor can be charged in two ways:
- configure pin 7 as input, with a pull-up. This will slowly charge the cap.
When the cap is nearly full (at least more than half-way) reconfigure pin 7 as output, to get rid of the resistor.
This works, but with a pull-up resistor of 75 k ohm, the charging will be very slow. A faster way is: - [correction] Connect a second pin to the capacitor, with a resistor in series of, say, 4700 ohm. First configure that second pin as input with pull-up. When the cap is sufficiently charged, reconfigure both pins as output.
- [correction] Connect a second pin to the capacitor, with a resistor in series of, say, 4700 ohm. Configure that second pin as output.
- [correction] When the cap is sufficiently charged, reconfigure pin 7 as output too.
Also, best is to use a smaller cap, to reduce the charging time. 2 µF already seems plenty.
[addition] With 4.7 ohm and 2 uF, waiting 20 ms seems OK (time constant = 4.7 kOhm * 2.2uF = 10.34 ms).
Below a scope plot which shows the steady voltage supply after charging a cap. The small jump on the green graph is when the pins are reconfigured to output. After that, the code is performing several I2D and ADC conversions. 
The NHS3152's NFC controller is type 2 compliant. There is no way to pause the NFC communication, or to delay the tag reader's readout of the tag's NFC memory. This means you must prepare an NDEF message as soon as possible.
This means that your passive program cannot make the measurements in time: it must revert to displaying the result of the previous tap.
So the order would be, on the <n>-th tap:
- power-up due to NFC
- start charging of the cap
- create an NDEF message with the result of the measurements performed at tap <n-1>
- wait until the cap is charged
- start measurements
- store the result in EEPROM
- wait until the NFC field is removed and the chip dies.
Kind regards,
Dries.