After I took my sensor out yesterday I thought it might be quite interesting to take it apart and see if I could work our how it works and why it has a 14 day limited life.
What I determined is that the sensor comes in two parts. The "electronics" and the "electrochemical transducer", which is the real sensor.
Starting with the electrochemical transducer", what I think is happening is that the filament is detecting the changes to the glucose level in the interstitial fluid and via the combination of enzymes in the filament unit, constantly generating an electrical current. This is passed through a set of contacts on the sensor to the electronics.
The "electronics" are then storing this data. The electronics consists of an NFC chip, which we know about, a small battery (which is more than enough to power the device for longer than 14 days) and some circuitry. I'd guess that there is a controller and memory chip on there, but I wonder whether that is integrated into the NFC chip. I make this supposition because we know that there is limited storage capacity on the device itself. Plus of course the NFC antenna.
The simplest way for it to work would be that the variance in current or voltage caused by the variation in glucose level would be recorded in the memory of the device, and when the reader pulls the data off the NFC chip it interprets this to glucose level units.
Looking at the make up of the device, I suspect that the 14 day limit is imposed to try and maintain fidelity of the signals from the filament component of the sensor rather than anything to do with the electronics. I guess that as the filament is used continuously, the enzymes become degraded by constant exposure and that rather than risk getting variability in readings post 14 days, they have elected to cut off while it is known to be good. Or it could just be in order to recoup costs.
Interestingly, the design of the sensor unit with the removable electrochemical component does suggest that you could use the NFC component with a wider range of "sensor" clip in units to test for other things in interstitial fluid, or replace the existing filament with something with greater longevity.
Of course, this is mostly speculation based on what I'm looking at and has no basis in any info from Abbott, so I could be totally wrong!
Another thing I noticed was that I clearly caught a capillary on the way in as there was blood on the base of the unit. I wonder whether this affects the accuracy or otherwise of the sensor. It will be interesting to track how the new sensor works and check the base when that is expired.