This forum has moved to forum.jeelabs.net - this read-only archive is for reference only!
Just got my RF modules from Hope, I'm going to try and solder them on tonight. I'm excited to start building and helping out.
Anyway, anyone have any ideas on a WISP system with one of these radios? Potentially using RSI to triangulate a node that could be carried? I dunno how long a battery would last powering the radio (would it need a microcontroller too?). Just throwing the idea out there.
WISP as in http://seattle.intel-research.net/wisp/ I suppose?
You can't do much with an RFM12B without MCU. And as far as power is concerned - it consumes 10+ mA when turned on, so it's unlikely you can run it off harvested energy - or maybe very briefly, by accumulating energy in a supercap for a while.
I guess referring to WISP was a red herring, I was simply talking about finding the location of people. My thought is that I have a bunch of jeenodes in my house and I can locate where everyone is by triangulating a fob in their pockets. I didn't even think about running it off harvested energy, that would be amazingly cool, I was simply thinking a small battery. How would one go about harvesting energy? I assume the energy would be "caught" from radio, micro, etc. waves? Could enough power be stored in a cap to keep an MCU in some sort of sleep mode until it got enough power to send out a pulse? It would be truly amazing to get this stuff to work; it's a building designers dream to know where everyone in a building is.
Just got all 10 of my jeenodes soldered up, I'm going to test the radio portion out next. I can't say how much I appreciate what you've done on this project, I'll try to help in any way I can.
Ah, so everyone has a JN-running-on-a-battery in their pocket, and you've got a bunch of them located around the house?
It's not easy to get at the signal strength with the RFM12B - the module does have an analog signal you could hook up to an AIO port. The digital readout only says above-or-below a settable threshold, so I suppose it could be used by trying out various thresholds - but that requires multiple packets to try out various settings.
Since the one carried in a pocket presumably has less power to spare, you could also have the fixed ones send out packets at various xmit signal strengths, and have the mobile one report which ones it picks up once in a while. The weakest packets could then be used to indicate proximity.
Getting power levels way down is technically feasible, but hard. May require tricks such as TDMA, i.e. time slots so both sides communicate at synchronized moments in time and can turn off the rest of the time. Getting power down to the levels where energy harvesting becomes possible is probably extremely hard. I'd love to go there one day, since right now the #1 inconvenience with deploying JeeNodes is probably to figure out how to power them for long periods of time.
For now, I suppose LiPo batteries with simple re-charging is the best way to go.
RFID might be another way to do this. Either carry the RFID badge, or turn it around: place RFID's around the house, and report which ones you're close to - assuming you intend to carry a JeeNode in your pocket for various reasons anyway. Power use could be low, since the mobile JN only needs to become active and transmit a packet once in a while.
Well, it sounds like maybe a pipe dream considering my current skill level (though I'm happy to say I've built and tested 10 jee nodes, yay!).
Regarding RFID, my understanding is that readers that read from a distance of anything more than 10s of centimeters or so are power hungry, costly and big unless you're using active or BAP RFID which then require battery power. BAP could be something to explore though.
I think I'm getting ahead of myself, once you get those home automation plugs done, and I've got them installed, I'll think about this stuff :)