Casio Smart Pay Watch

I've wanted to get a smart watch to easily pay for things with a tap, but also don't want yet another thing in my life to charge every night. I thought how cool it'd be to just repurpose my casio watch as a tap-to-pay watch if it contained the NFC chip from my credit card! First step is to melt down the credit card in acetone. I left it in a sealed mason jar until the plastic card got soggy enough for the chip to easily come off of the card. You can see it on the left side if you look carefully.

Once the chip was off the credit card, I 3D printed a mold to wrap magnet wire around that represented the antenna size/layout of how it would be under the watch face.

I got some very strange looks at checkout while paying with this:

This is the custom NFC antenna printed onto a flexible PCB, with the NFC chip from a credit card soldered to it. The wiring is 30AWG insulated magnet wire. Online calculators estimate it to be around 3 microHenry but haven't verified actual inductance. It works when paying at card readers in this configuration.

Here is the antenna fitted in front of the watch face. There is this 'retaining' plastic that separates the watch unit w/LCD screen from the front of the watch face. The pcb is in front of that dividing material, and will sit behind a sheet of acrylic like the ones seen with lettering on the front of cheap casio digital watches.

Pic of front of the watch. Not shown is the piece of acrylic that sits in front of the pcb. it's ~1mm thick and painted black on the back of it everywhere except the view port.

Finally, the watch unit is inserted and the NFC chip fits in a small cavity next to the battery. It's wrapped in kapton tape to prevent shorting. In this configuration, it does not work.

I tried in vain to wrap the card in metallic tape to act as a shield. The goal was to reduce the parasitic capacitance I was getting that seemed to throw off the resonant frequency antenna and the LC tank circuit it forms.

I used an arbitrary waveform generator to send a sweeping square wave over the antenna + chip assembly, and an oscilloscope with Fast-Fourier Transform feature to see what frequencies the system (antenna + SIM chip) would resonate at, which in turn would help me figure out how large of a capacitor I'd need to add on to target the right frequency of tap to pay chip readers

This was the result of that, and where this project currently is left off at :/

I'd love to resume this project, it was so close to being complete! My plan is to try adding thin ferrite sheeting behind the flex PCB and chip, but if that doesn't work I'll try and find access to another arbitrary wave function generator to see how it's being detuned when placed inside the watch. If you have any advice on next debugging steps, let me know.