Using NFC to measure a frequency shift

I'm working on an idea for a wireless sensor. My idea is to have a smartphone and a passive NFC tag. The tag operates as a sensor and is slightly detuned from the 13.56MHz resonance when it measures a large signal. I would like to be able to quantitatively measure this frequency shift on the transponder (smartphone) side.

From my knowledge of coupled inductive systems, I know it should be a simple manner to detect this frequency shift using just one port in a network analyzer plugged into the transponder resonator. S11 would be a function of the L, C, and Q on both antennas, in addition to the mutual inductance between them, so if the L on the sensor side is the only variable changing, I can measure it. But lots of things are easy on network analyzers.

Perhaps this is a problem better suited for the Digital Communication Forum, but how could I use NFC to detect this frequency shift? Can a smartphone using NFC measure anything close to Insertion Loss?

A NFC poller has no features to tune the transmitter frequency. In so far this won't work.

You should also consider that the tag resonance frequency is shifted by metallic objects in the vicinity (and much more by additional NFC or other 13.56 MHz tags).

Hi,

I ask myself: Do you really see a frequency shift or a phase shift?

I′d expect that a "mistuned receiver" can′t move the transmitter frequency, nor sends out a "mistuned frequency". I also expect a mistuned receiver has decreased sensitivity.


Klaus

Thanks, that is pretty much the heart of my question. After reading more on how NFC occurs via ASK, this makes sense.

I suppose in my thought experiment of a one port system plugged into an LC resonator which is coupled to another LC resonator, I have not investigated much how S11 changes as a function of L2.

So in the above circuit both resonators start with the same resonant frequency, but then the second inductor is variable. I suspect with all other parameters fixed (something I know is much easier in simulation than in practice) S11 would change in phase and amplitude, like you suggested. Unless the coupling is very strong, you are correct that a frequency shift would not be measured directly. It is worth mentioning in my application coupling will be relatively weak (k < 0.1).

I see possible approach to make the primary circuit non-resonant (omit C1) so that the secondary resonance shows more clearly in S11. For low coupling factors, de-tuning by L1+Z0 is small but still present.

That is a clever suggestion, but I'm not sure the design is NFC compatible, which is my primary goal. Or, I should say the primary goal is for transponder to be a smartphone, I've just decided NFC would be more useful for my application than wifi or bluetooth.

Attachment 135505

Here is my current idea for making the tunable inductor something NFC can conveniently sense. I'll leave the Tx and Rx NFC antennas alone, since as FvM mentioned relying on the comm antenna for sensing will lead to all sorts of interference headaches.

I could rectify the NFC into a small DC voltage. I've made up some block that would use this low power to generate some noise. That noise passes through the resonator made with the variable inductor, which acts as a BPF. The filtered signal would then turn a switch on and off, thus signaling back to the transponder using OOK (or LSK? Sorry I'm not great at describing comms). For this application R1 would be relatively small to the switch resistance in its off state. The frequency of the OOK would correspond to the detuning of the variable inductor, which I can now isolate better.

I'm sure I'm going to find all sorts of pitfalls with power management and minimum voltages, but that's the jist of my thoughts.

I didn't see post#4 related to NFC, S11 hardware can't measure S11, as previously discussed. Don't mind if someone could correct me.

If you want to utilize NFC, you need to speak NFC protocol, decode the commands, generate the correctly modulated subcarrier to answer it.

Hmm, looks like my circuit didn't load, here is what I was trying to show.

The circuit can be expected to generate load modulation. Problem is that you are using a variable frequency while the NFC subcarrier is fixed to integer fractions of 13.56 MHz (/32, /64, /128). Even if a particular NFC chip could decode other signals, there's no documented interface how to access this additional functionality in a smart phone.

Ahhh, it that sort of subcarrier detail that is exactly the particulars of NFC I am so ignorant of. Thank you for the tip.

Reading up on the RFID handbook, they describe an example of NFC wherein the load resistor is clocked at 848 kHz (13.56 MHz / 16, just as you suggested FvM).

The 848 kHz subcarrier is then ASK modulated, apparently through Manchester encoding. I would think that modulating the the subcarrier using my variable resonator would be doable as long as my variable f0 is a good deal less than 848 kHz. Obviously there would be a resolution limit, but if I modify my circuit to include an 848 kHz oscillator, would that do the trick?
Source: http://rfid-handbook.de/about-rfid/a...owall=&start=2