For the driver part (not antenna) would it be possible to utilize an pn532 or rs522 board, replacing the antenna with a custom one?
Almost none of the stuff I ordered a month ago has arrived, but I’d like to give this a go if someone could consult on what to remove or where to cut🙂
The RS522 can be quite small if the antenna is removed
So my LCR meter finally showed up…
I tested my coils, and unsurprisingly the inductance was well off (like an order of magnitude), especially the one above that I malformed around my wrist. So that bodes well me thinks. This weekend I have been doing some spring cleaning but over the week I plan on making a coil formers based on my wrist shape, much like the last attempt with the 3d printer then I’ll really try to dial in the resonance with the LCR meter, hopefully it goes well. I’ll also try to make the model parametric so anyone else with an interest and access to a 3d printer can have a go.
Sooo, is there anything I can do to help out? I’d be really interested in putting this project forward in whichever way I can.
Well, first things first.
Nice to meetcha. Welcome to the forum.
I guess I’d say it depends on your skills, which I know nothing about, but you got the right attitude for sure. You might try knocking up a circuit similar to some of the ones above that work, then seeing if you can improve it a little here and there.
So, I’m thinking there are two places we can go with this.
HF “Dumb” bracelet
Basically that circuit I posted at the beginning, driving a mosfet at higher voltage and current so we can get greater distance. @NiamhAstra is gonna try this out with some parts he has lying around. Here’s an example schematic:
LF “Smart” Device
When operating at 125kHz, we’ll naturally get longer range. We’ll also be below the clock speed of most microcontrollers. This means after we chip our teeth on the HF dumb version, we can make a smart version that reaches farther and responds to commands from your smartphone over bluetooth for notifications and such. Downsides are that it will require more development and the inductor coil will need many more turns.
We could also have a HF smart device. Where the smarts is just turning the coil on or off. Or use a chip that has a high enough PWM for HF.
I quite like the HF option purely because of
I was thinking something like a RN4872
alright, thats it, i’m interested. I’m ordering an an led implant and beginning development, give me a few weeks to get a prototype together.
Building an oscillator from discrete components is unnecessary really. 13.56MHz is pretty standard and packaged oscillators such as this are readily available. DC in 13.56MHz RF out 
https://eu.mouser.com/ProductDetail/Microchip-Technology/DSC1001BI5-0135600?qs=Cv1v43EOJpqpoFx%2F8gxmMw%3D%3D
I’d use that (might need buffering) to drive something like a HF class E amplifier, as these are reasonably efficient. A few SOT-23 packaged BS170s’ in parallel can easily produce 5 watts. House it in some cheap large faced watch casing with a small lipo or two. Probably would try ribbon cable as the strap, offset by one conductor to form a muti-turn inductor for the antenna.
Now we’re talkin’. Or maybe a flex PCB tucked inside a real watch strap that has exposed traces that mate with pogo pins inside the watch body.
That’s definitely the plan. We just wanted to do some proof of concept circuits first to see what kind of performance we can expect. I can definitely miniaturize everything on a PCB once we decide if it’s worth it, but I haven’t had the bandwidth to thoroughly test this because of other dev work. Doing it with discrete components makes it easier for anyone to try out, plus it will incidentally leave a DIY/open-source version of the project for others down the line.
What’s the rationale in using that?
This is just for HF, because cheap MCUs often don’t have 32MHz clock speeds. We’re just trying to blast out a signal to light the xLEDs
We take a crystal oscillator set up to output a super weak 13.56MHz waveform. We pre-amplify that mV waveform to a few volts using the 2n3904 transistor. Then we take the output from that and put it through a larger amplifier to boost the output even further. It might not work great, but it was an inexpensive solution so I’m willing to try. Our existing plan to source an RF capable mosfet was hitting some snags, so this is an alternative.
Being a practical man, I’ll be repurposing a tabletop reader for my own “bracelet” - which will sit under my arm, to light up my doNExT on the other side. And since it’ll be battery-powered and I plan on making the field remote-controllable, ideally I’ll reuse a BT reader.
Of course, the obvious choice for easily repurposable, easily programmable hardware for my project would be a PM3 + Blueshark module. But… Holy expensive bracelet just to light up 3 LEDs 
Still, it’s an option, and I might justify the purchase on the grounds that not only will I have a nifty device to turn the doNExT into a morse code display, it’ll double as a handy portable NFC reader I’ll carry on me all the time.
I don’t want to piss on your fire but that device is a MSA0886 MMIC. It’s used mainly as a low level amplifier such as a receiver front-end. Although its Noise Figure isn’t great. If you look at the datasheet its 1dB compression point is about 12dBm, that’s 16 milliwatts! You’ve already got way more than that from a properly designed 2N3904 stage. Also that device is usable up to about 5GHz, you wouldn’t typically use something like that at HF, very good chance of it being unstable.
Yep should be relatively straightforward to do. I guess some might want just the power feature to light LEDs etc rather than a full on reader option. Doing something custom there is a chance of improving on its supply efficiency. I guess the reader designers aren’t that bothered about optimising the supply requirements, as long as it’s reasonable and doesn’t smoke your USB port then that’ll do.
I’m not too sure about that. I mean most of those readers are usually nothing more than a frontend and an enclosure for an integrated NFC chip like the PN533, and those aren’t terribly thirsty for power to begin with, since they may be used in battery-powered devices.
Also, in my case, I’m actually interested in something that can programmatically turn the field on and off. If I were to roll my own, I’d have to add some logic to it anyway to do that. So I might as well go for something that already exists.
Anyhow, for the money and for the lack of headache, I’m happy to charge up a bit more often
Thanks for the heads up. I was concerned that it wouldn’t provide substantive amplification, but it was ~$12 so whatever. I’ll keep it for another project. I don’t have as much experience with far field equipment as I do with near field, if you couldn’t tell
I caved and bought an expensive RF Mosfet in a TO-220 package (I want this first revision to be DIY friendly). The model I bought was the MRF101AN from NXP. I made a nice little protoboard version using the schematic I posted earlier:
Here’s the updated schematic with values for the components on the right side:
I verified that the Pierce Oscillator outputs a 13.56MHz waveform. It has a pk-pk amplitude around 1.3V. That’s not enough, though. The RF mosfet has a minimum Vgs of 1.7V and a nominal Vgs of 2.2V. I stole the Pierce oscillator design because I haven’t biased a BJT in a while, and I certainly never did it with series diodes 
. If anybody has any insight how I can reconfigure the Pierce Oscillator circuit to allow a greater Vgs across the mosfet, I’m all ears.
The whole diode thing dosnt sit right with me, personally I would use a colpitts but thats not the point. I think there’s a few issues with the biasing and if im correct the circuit below should fix it and do away with the diodes aswell.
NOT MY CIRCUIT ITS STOLEN.
It is very interesting to see what you are all doing here. It my opinion it would be nicer to just keep it simple. I’d aim for a royer oscillator tuned to roughly 13,5MHz. For transistors i’s aim for something like the ss8050 or bc547c because they are really cheap and go far beyond 13MHz. For a very compact version i’d look at KI2302DS. It’s a common smd mosfet thats small and powerfull. I didn’t try it yet but i can try out how high it goes in therms of frequency. When you need power (and i mean lots of power) you can look at the Irf510 mosfets. When you drive it hard enough you can bring it up to 50MHz at 40w per mosfet. I have all the parts here so i could throw some circuits together if you like. I have no xled here to try but i once built a field tester from a bicolour led, some wire and a capacitor.
Perfect timing @spyfoxls!
leumas95 has just started experimenting with the high power prototype I sent him and he was planning to test out the IR510 as a cheaper alternative.
I am working on the light version (on and off). I have everything planned out and parts selected except for the surface mount fets. I’ll look into the ones you suggested.
This is a community effort so if you want to try out a slightly different version yourself that would be most welcome. Having different types will help us settle on the optimal design down the line. Let us know what you figure out or if you want to bounce some ideas around.
