Hi everyone, new to the forum and the whole biohacking world (thank the modern rogue for that lol).
So, I’ve been doing some research into the kind of implants that dangerous things has available, and started to wonder if anyone has tried putting a microcontroller into an implant with the RFID coils.
This might be a bit insane, but I’m almost certain an ATTiny would fit into the glass capsule, specially if one were to use the chip with no housing.
I’m not to sure how programming would work, but I’m certain there’s a way to do it wirelessly (maybe even pack a BLE chip in there or something? Idk)
A bit impractical due to size but as things get smaller they would become more practical. As for implantable without a package (not dip/soic) just the die then most would fit in a implantable size. Being able to communicate would be the trick but not impossible what are you wanting to accomplish?
I’m interested in emulating almost any RFID/NFC protocol there is out there while using a single implant. While it is possible to do it with what we have now (xNT, xEM, NExT, etc), I thought it’d be nice to simply plop a microcontroller in there to cover for anything that those can’t emulate, as well as any new protocols that show up
Gotta admit, this made me laugh,
That’s honestly the idea. No need to worry about all the different protocols out there if you can simply program each one into a microcontroller.
And it’d allow you to (ideally) store multiple IDs that can be presented until one matches (which may not work because, while the standard makes it possible, programers are lazy and most devices out there don’t implement this…)
As far as circuitry goes, it’s actually possible to power the ATTiny with no extra stuff (but it’s probably not optimal, xD). Here’s a post I found when doing some digging into it: https://trmm.net/AVR_RFID
I guess the biggest pain would be reprogramming the thing while it’s already inside you
This is totally doable. I’ve tried some stuff in this vein, but there are challenges. Here’s a few:
Most MCUs require a few mA of current to operate when they’re not in sleep mode. Most of the RFID energy harvesting schemes I’ve seen can only muster 15-20mW (20mW / 3.3V = 6mA). You’re also lucky if you can get 3.3V, some of them cap out at 1.8V. The best way around that is to use batteries, but we don’t do that here. The Grindhouse Wetware folks are making good headway on that front, but they’re more interested in using it for sensor platforms than for card emulation. You can use caps to store energy from the field over a few seconds to provide a larger burst of power for energy intensive operations like writing to the EEPROM and changing the session registers.
Most MCUs only come in QFP packages which take up a lot of real estate. You could do what devilclarke suggested and use a bare die (or flip chip package) to save space, but that comes with it’s own hurdles. In addition, the RFID analog front end (AFE) is often a separate component, so that’s two huge ICs. There are very application specific NFC+MCU chips that you could use for this type of thing, but they’re not made for our use case so there’s always logistical hurdles to overcome. I wouldn’t plan to use something like an ATTiny, unless you’re just trying things out. Check out MCUs like this one from TI that have everything you need in one chip [Edit: actually I’m not sure that chip can operate standalone, but standalone tag emulators are out there]. Bottom line is, if you’re going to experiment with something like this, don’t plan on it fitting in glass right away.
Hmm, from reading online I thought NFC would typically give off power in the couple hundred mW range, though this might indeed be on a case to case basis.
As for needing and AFE, I’m quite certain it can be skipped for transmitting. Maybe it’s necessary for receiving, not sure though.
I’ll keep looking into options to use. As long as I can find a package that’s around 2.5mm in width, that could work
Oh, I’d be interested to see where you read that. From my experience they usually cap out at less than 20mW. Here’s an STMicro tag that claims 100mW, but you can see in the datasheet that the peak draw is 6mA which also droops the voltage to 1.7V. I don’t know how they’re contorting the data to get 100mW.
I found this really detailed paper which gives a rundown of the current state of NFC Energy Harvesting tech. Might be worth a read.
I also found this field powered MCU: LPC8N04
It looks pretty capable, and you can rewrite the firmware over the NFC interface via a second-stage bootloader (which I don’t fully understand). The downsides are that you’d probably have to write most of the tag emulation code yourself (NXP’s not going to hold your hand) and it’s 4mm x 4mm x 1mm (a bit big for glass).
You may be right about that. As long as the clock rate is high enough (some of these low power MCUs cap out at 8MHz, you’re going to need at least 16MHz) you should be able to bit bang some kind of FIFO buffer like the AFEs have using the interrupt pins, but that’s going to require some very careful timing.
I’m more of a hardware guy myself. I write microcontroller firmware sometimes, but I definitely don’t enjoy it. I say the best embedded developer is one who does as little embedded development as possible.
Just to chime in, I’m currently in the middle of a project involving the design of a field powered implant, and the configuration I considered was an RF430 powering a TI MSP430F2013-EP and communicating with it over I2C. The regulated output of the RF430 can apparently supply 20mA at 3.3V.