They would be suitable for testing, but not for the finished product (at least I am) envisioning. Taking several rigid PCBs and smashing them together won’t be much better than the pegleg. I’m hoping we can get this on a custom flex PCB that is around 45mm x 35mm x 3mm
I got one for testing with a lithium coin cell based led implant that I never truly finished and the receivers for these can be truly tiny!
Look at this… This is a 5V receiver o.O
I have an idea on how to recover the vocore2 but it requires some hardware and testing. Parts are on order and I’ll make a post if it returns anything worth while.
The idea I had was to use a set of infrared LEDs and photodiodes to facilitate a serial console on the vocore. That would allow more indepth recovery abilities along with being able to write a new OS image to the device even after it had been implanted.
So far my testing has been with an optointerrupter connected to a Teensy2.0 as a loopback serial device. It works at every speed up to 57600 baud. Beyond that communication begins to degrade and typed characters get dropped or misread. It may be worth pursuing further if we can get the serial console on the Vocore2 to operate by default at 9600 or 14400
So, I am building an image, and trying to limit the functionality to what we need.
Do we want IPv6 support? I don’t really see a need for it if we are only going to be using a private unconnected wifi network.
Equally the default would be for luci to be running on port 80. This is the openwrt web configuration tool. If we put the file service on port 80 and remove luci would that upset anyone?
The only configuration you might normally want to do is to change the wireless network options and upload new firmware. Both of those are available through the command line.
I agree on not needing ipv6. How much space is saved by removing the luci interface? I think as long as we have a good amount of documentation for doing what people want using the cli won’t be too much of an issue.
For me, i require some means to transfer files whether that is ftp, scp, or something else. Are you able to find a way to reduce the default serial port baud rate to 9600? if you can that would make it at least reasonable for us to integrate an optical serial link into the device for some semblance of recovery.
I’m going to order it after Chinese new year is over on a 0.8mm thick 4-layer rigid PCB. It will be able to source 750mA of current at 5V with the configuration I have selected. StringFellow determined that the VoCore dev board draws 600mA peak at power on for ~100ms, then drops to 300mA continuous.
Here’s a reference of the form factor in aiming for:
Starting the layout now. I already have the antenna geometry mathed out. A four layer stack each with seven 1mm thick traces with 0.2mm spaces that has an outer dimension of 38mm x 28mm should give us between 9uH and 14uH which aligns with the Qi specification. After I receive it I can check the inductance and determine the resonant capacitor values. Using such thick traces will also keep the temperature rise low (aiming for less than 1.4°C rise at maximum draw).
To achieve this form factor the SD card might go on the bottom of the board with some passives, although we may be able to position the SD card on the top but with the rear over the coil without affecting performance. Here’s a purchase link for the MT7628AN chip from the VoCore that we’re planning to use.
I’m hoping that we can encapsulate with the biopolymer. If we need to go to PMMA then the entire assembly could go from 2.5mm thickness to 5mm very easily.
I didn’t just fall into this skillset. I figured out what I wanted, and targeted the skills that would get me there. I’ve been at this a decade now though. A part of it is definitely natural affinity, but I feel like anyone can target the skills that will make their dreams a reality.
I hope the trajectory of our society makes that option easier for the next generation. “Innovation” is all that the US has left. We’ll likely never be a manufacturing power house or lead the way in the production of anything but poisons ever again. The least we could do is use the remainder of our colonial exploitation wealth to educate and enable a generation of thinkers and dreamers that can fix the problems we’ve caused.
For software, I think it would be great if it could act as a host for communication using Element.
Hosting Element is explained here, and leaving out Jitsi would probably be best since video conferencing wouldnt work (or would work terribly & make everything else awful). It explains install for use on Debian.
That would cover secure communications.
And some type of routing service would be good for if there are multiple people with this implant, in the same area, so people can connect from one to the next, instead of needing to disco/reco from one to the other. But Idk what kind of handshake protocol would work for that, or to block malicious servers, because that would definitely become a problem.