NFC field generator

[grymoire]

I am working on an art project, and I want to build a thin panel that can generate an NFC field which can power some NFC LEDs, such as  https://www.sparkfun.com/products/14888.

All I need is to light them up. I would appreciate any help. A thin design would be great, but I also am interested in increasing the XMIT range, so I can light them up from a distance. I don't think I need to read any information.

Thanks

[Berni]

You basically just need to power a coil of wire with the right frequency AC signal. NFC tends to use around 13MHz so those LED tags are probably optimized for that.

How you generate the alternating current for the coil is completely up to you. But i suspect that getting it to work over a long range would include large coils and a lot of power put into it.

[grymoire]

Yes, NFC is 13.56Mhz. I found some NFC coils that are used for smartphones. And I can get an NFC board with a build in antenna. I don't want to use a large board if I can avoid it. It's the

How you generate the alternating current for the coil is completely up to you

part is where I need the help.

[Berni]

Well you need to make an oscillator that generates the desired freqency, Any oscillator can be used or even a microcontroler can be used to toggle a pin at that rate. Whatever you are most comfortable using.

That needs to be put trough some sort of power stage. This means using a transistor to switch the power to the coil at that freqency or using some sort of driver chip to do it.

You also might want to have a capacitor wired across your coil and tune it to a value that makes the capacitor and coil oscillate as a LC circuit at the desired frequency. Take it step by step and play with it until you get it working.

[judge]

I'm just trying to do the same thing. I found a circuit for an oscillator here that uses three NAND gates and a 13.56MHz crystal (and a few discretes). The oscillator basically works but the voltage is kind of low (around 2V p2p), plus the waveform looks very noisy. I'm not sure if the noise is real or an artifact of me hooking up my ancient oscilloscope.

I figured I could maybe feed the signal into the gate of a MOSFET to switch the full 5V, but the MOSFETs I have to hand are too slow (rise time of around 200ns!). So I was wondering if anyone could point me at a suitable part, advise about how to clean up the waveform, tell me that this will never work and I should do something else instead?

Also curious if the OP every got this to work?

[Benta]

Do a search for "wireless charging" or "contactless charging". Several suppliers have solutions (TI, Infineon ec.)

[judge]

Wireless charging works at about 100KHz. NFC is at 13.46MHz. I have tried these LED tags on wireless charging stations. They don’t work. They only work with NFC.

[PA0PBZ]

This should get you going: https://www.aliexpress.com/item/32959673490.html
Looks like they use an IRF530N

[judge]

This should get you going: https://www.aliexpress.com/item/32959673490.html
Looks like they use an IRF530N

Your searching skills are way better than mine! That is one helluva amplifier! Interestingly the datasheet for that MOSFET seems to show that the total rise/fall time is 91ns, which is greater than the period of a 13.56MHz waveform (73ns). I guess it would be less at a lower voltage/current? Also the Vgs(th) is 2V which is right at the limit of what my circuit produces. Or maybe not. 2V is the p2p voltage, I think that is offset from 0V some. If it isn't, maybe I can offset it.

I can also start searching for 13.56MHz amplifiers now. Many thanks.

[Ammar]

If you just care about lighting them up get one of these: https://www.amazon.com/M24LR-DISCOVERY-Discovery-transceiver-M24LR04E-R-powered/dp/B01EJ933VK

And follow these instructions to get it delivering power continuously:

Go here: https://www.st.com/en/evaluation-tools/m24lr-discovery.html#tools-software

Download "PC software for executable demo..."

Connect the daughter board to your PC via USB

Open CR95HF development software

User commands → CR95HF Command List → 02: Protocol Select → Choose Reader protocol code 01:ISO15693 → Click send frame

That should do it.

[Ammar]

Only 10W??

This would really get things going: https://au.mouser.com/pdfdocs/IXYS_PRF-1150.pdf

Really depends what the op is after. By distance do they mean a cm or like 10 m?

[Ammar]

This should get you going: https://www.aliexpress.com/item/32959673490.html
Looks like they use an IRF530N

[T3sl4co1l]

Switching times are limited by device capacitances alone -- RF amplifiers' bandwidth is subject to the same limit, but the center frequency is subject to more detailed characteristics (in this case, limited by the dropping gain / rising losses of the MOSFET with increasing frequency, and stray inductance affecting stability).

An oscillator with something like IRF530, at 13MHz, is pretty typical.   Get out some wire, some smaller capacitors, and you're good to go.  (Would recommend a Colpitts type.)

Tim

[t1d]

Only 10W??

This would really get things going: https://au.mouser.com/pdfdocs/IXYS_PRF-1150.pdf

Really depends what the op is after. By distance do they mean a cm or like 10 m?

Obsolete item. No stock. No price. Doh!

[t1d]

I'm asking, here... Could a directional antenna be built?

If the airport scanner would make them blink, that could be fun...

[T3sl4co1l]

Taking NFC literally, no.  You can make an asymmetrical field, but you can't make a field stop arbitrarily (no matter how many times the textbooks draw diagrams as if you could).  To be exact, the field satisfies Laplace's equation, \$\nabla^2 \phi = 0\$.

A very large antenna (in human terms; a dozen meters or so) can have directional gain (i.e., better than dipole) in a propagating field, but that's not near-field.

The usual approach is to place a coil over a ferrite plate, which acts to short out the magnetic field lines on that side.  This reduces the volume of field (reducing the magnetic path length more than the effective area is reduced, hence the inductance goes up), giving less range on the open side, and higher mutual inductance up close.

The shield is as much practical as anything: it improves coupling between coil and target, and shields the circuitry underneath, allowing a compact build.

Tim

[t1d]

That's a good explanation, Tim; thanks!

[Berni]

Only 10W??

This would really get things going: https://au.mouser.com/pdfdocs/IXYS_PRF-1150.pdf

Really depends what the op is after. By distance do they mean a cm or like 10 m?

Wow that is quite the powerful little RF amp.

I feel like the datasheet for that should include at least a few big red warning boxes around this monstrosity.

[judge]

Switching times are limited by device capacitances alone -- RF amplifiers' bandwidth is subject to the same limit, but the center frequency is subject to more detailed characteristics (in this case, limited by the dropping gain / rising losses of the MOSFET with increasing frequency, and stray inductance affecting stability).

An oscillator with something like IRF530, at 13MHz, is pretty typical.   Get out some wire, some smaller capacitors, and you're good to go.  (Would recommend a Colpitts type.)

Tim

Fortunately bandwidth isn't something I need to be worried about. I assume. I wish I knew more about how these LED tags work. There's a coil and an LED of course, but there are some discretes on there too, and not just capacitors! It would be nice to know what sort of voltage a typical NFC reader shoves through its coil so I know what I am aiming for. And I am assuming voltage is important in this application because although LEDs are current-limited devices, they also need a minimum voltage to get them going. And, of course, I am assuming here that it is the 13.56MHz carrier that is lighting these things up.

Also, I read a lot about 'antenna tuning' for NFC readers, but my understanding is that these coils aren't antennas at all, just inductors, and I never read about 'inductor tuning', so I am curious as to what the deal is here. There is clearly more about getting the efficiency of these coils up, like apparently wanting the self-resonant frequency to be above the driving frequency, balancing range v. efficiency, high 'Q' v. bandwidth. It is hard to know which of these are relevant to my application, which has no requirement for bandwidth. Plenty to learn, and experiment with.

[judge]

Only 10W??

This would really get things going: https://au.mouser.com/pdfdocs/IXYS_PRF-1150.pdf

Really depends what the op is after. By distance do they mean a cm or like 10 m?

Wow that is quite the powerful little RF amp.

I feel like the datasheet for that should include at least a few big red warning boxes around this monstrosity.

I read elsewhere that 13.56MHz is used for more than just NFC. Something about welding and induction coils?

[T3sl4co1l]

Yes, 13.56 is the free ISM band in that range.  It's rather narrow though, which severely limits applications: induction and dielectric heating needs an active tuner (mmm, servo driven vacuum variable caps!) or an amp with a huge tolerance for SWR; NFC and other transmissions must be low bitrate; etc.

NFC at least being near field might permit more data rate up close (it could also be driven with QAM rather than ASK, giving more bits per symbol), while still meeting the limits in the adjacent shortwave bands; although I don't think anything can do that.  If nothing else, a subtle modulation scheme wouldn't be easy to manage with self-powered hardware.

Tim

[judge]

I'm not getting anywhere making an amplifier using a single IR530, either in LTSpice or in reality. Reality matches what I get in LTSpice though (which is at most 1:1), which is always nice. On the other hand, I did manage to light up the tags by adding capacitance in series with the coil to form a resonant circuit:

[David Hess]

You also might want to have a capacitor wired across your coil and tune it to a value that makes the capacitor and coil oscillate as a LC circuit at the desired frequency. Take it step by step and play with it until you get it working.

Making the coil a tuned circuit allows for much higher circulating current between the coil and capacitor while lowering the current through the amplifier so a less powerful amplifier can be used.

[Berni]

I'm not getting anywhere making an amplifier using a single IR530, either in LTSpice or in reality. Reality matches what I get in LTSpice though (which is at most 1:1), which is always nice. On the other hand, I did manage to light up the tags by adding capacitance in series with the coil to form a resonant circuit:
(Attachment Link)

Yes resonant LC circuits are the way to go for wireless power transfer.

The resonant nature of it causes the amplitude of the signal to rise significantly despite a small input power. This is because such circuits can 'recycle' energy really well. The inductor and capacitor exchange energy back and forth without you providing any input power. The amplitude of the oscillation only dropping slowly according to resistive looses and as much energy is transferred out to something else. So you only need enough input power to cover the system looses + transferred power.

Same trick tends to be used on the receiving side too. Being a resonant LC circuit it can develop a large amplitude from a tiny input power, so very little magnetic coupling is needed to the transmitting coil in order to develop enough voltage for powering a LED. However this only works well if both the transmitting and receiving circuit are tuned to the same frequency.

The lower losses you can get in your transmitting tank circuit (Low resistance coil, high Q caps etc) and the closer you can match the transmitter and receiver frequency the less input power you will need to get the same range/performance.

[judge]

Further messing around with LTSpice shows me what kind of circuit I need to make the amplifier work, but with the frequency, peak-to-peak voltage and possible bias voltage I have available, the IRF530 isn't much use. The peak-to-peak input signal is around 2V, which just doesn't seem to be enough to completely turn the mosfet on/off.

A mosfet I've used for other projects - the AO4294 - is much faster and has a lower gate threshold voltage. I might see if I can get that to work in LTSpice, but does anyone have ideas for a plan B? I want the circuitry to be small and unobtrusive - this is for a wearable project.