Physical connection to RFID/NFC chip with wires?

[casper.bang]

I suspect most RFID/NFC variants rely on the two-wire to coil, as the only connection into the IC much like Dave showed in a recent video. I was just wondering, in an effort to minimize latency and speeding up communication, wouldn't it be fairly trivial to bypass the antennas and connect host chip (reader) with the client chip (card)? Perhaps a 1K resistor in parallel might be all that's needed, any thoughts on this?

[Eviltech]

You know how the RFID chips together with the coils are designed to comunicate via modification of the host magnetic field? This also results in special designs to demodify the magnetic field. Also the host device is monitoring the current to it's sender coil verry carefuly. Removing the coil will confuse the host device, and connecting the chip to the host device either parralel to the coil or instead of it will probably burn it or not recognise it. Just use a chip card and a chip reader.

[minime72706]

Yeah I don't see this working. Near-field communications has been named as such for a reason. The device being read literally modulates how much it loads down the transmitter in the exact same way putting a heavier load on the secondary of a transformer affects how much energy is drawn from the power source connected to the primary. The technology relies on this type of interaction, which is of a magnetic nature.

[nctnico]

There is the answer: use a transformer. I think a transformer for telephone lines (like you'll find in a modem) is a good start.

[minime72706]

There is the answer: use a transformer. I think a transformer for telephone lines (like you'll find in a modem) is a good start.

My understanding of magnetics is lacking heavily, but I would expect that there is no guarantee that the inductance seen by the RFID reader IC will make it "Happy" if you use any old transformer.

[casper.bang]

There is the answer: use a transformer. I think a transformer for telephone lines (like you'll find in a modem) is a good start.

My understanding of magnetics is lacking heavily, but I would expect that there is no guarantee that the inductance seen by the RFID reader IC will make it "Happy" if you use any old transformer.

That then brings me to the next question. What kind of inductance (H) can I expect from an RFID/NFC coil? I have no way to measure.

[nctnico]

That depends on the type of RFID. Since you mention NFC: NFC is basically Mifare. If you look into Mifare specs you should be able to get a ball park figure.

[codeboy2k]

google image search for "RFID inductor" and you will see there is an immense variety of shapes and sizes of the antenna. 

Using approximate sizes, the RFID stickers are circular, about 25mm diameter, with a 20mm coil diameter, and about 20-30 turns.  That works out to about 90uH or so.  The bigger square ones are harder to calculate. There is an empirical approximation for a square planar coil spiraling inward:



N is the number of turns, and D is the length of the largest (outer side) in mm. This works for squares, not rectangles. So lets assume a square of 5cm x 5cm , with 6 turns, then L would be about 850nH.  So probably any high Q coil 1:1 coupling of about 1-100uH would be fine. You could probably put two small inductors side by side on the same board and they will couple, or you can actually use a transformer with primary and secondaries of a few uH. And, in fact, with such close coupling, you could probably get away with a very low inductance, perhaps just 100nH or so on a chip scale inductor.