Chris - That (and a previous message) clears up a lot. This looks like an application begging to be run at 13.56 MHz. That particular frequency had been devoted to high-power transmitters (anti-theft sensors at the exit of retail stores) and other devices (rf excited flourescent light bulbs, for instance). It's also convenient because your pickup coil in the shoe could be a few turns (or maybe even one turn) with 1 to several square inches area, while your transmit coil from the PC side could be quite large. As soon as the shoe is within range of the antenna, it starts sending data. Here's how I'd do it: The shoe has a small loop antenna as described above, with a parallel capacitor to make a resonant LC circuit at 13.56 MHz. The cap value will be in the range of 50-250 pF, depending on your antenna. A simple diode rectifier (amplitude detector) going into a comparator would tell you when there's a field present. The whole thing could cost $0.50 to $1.00 US, and run on 10-20 uA. The PC has a large loop (single turn) antenna excited constantly or in short pulses that occur frequently. You have to be kind of careful with this because the antenna voltage is 100+ volts. Some commercial units use up to 500 volts. This unit would be more expensive. Depending on how you build it, the electronics could be anywhere from a few dollars to a few tens of dollars, and the mechanicals could be several times that. The transmit frequency has to be crystal controlled to meet FCC regulations for this band, which dictate precise frequency control. With this setup, I'd think you could get max ranges of 2-3 feet. If you want better range or discrimination, you modulate the transmit field either with a constant tone or with actual data. Your receiver then detects the tone (fairly simple) or the data (more complicated, but very doable in a PIC), and only when it detects the correct modulation does it turn on the RF transmitter. I wouldn't swear to it, but you may be able to get ranges of 5-10 feet this way. A technically clever person with basic electronic knowledge (calculation of RC filters, configuration of op-amp and comparator circuits) could get a system running without too much difficulty, but to optimize it well would require either extensive trial and error, or some more advanced skills. I would read Microchip's RFID literature, but I wouldn't bother with their developer's kit. The lit is available on the web, I think. One of the drawbacks of magnetic coupling like this is that the antenna coils are kind of directional. They want to be kind of near parallel planes and co-axial. Range drops off rapidly as you get off-axis, and as you get beyond +-30 degrees off parallel. That's why I think placing the transmit coil on the floor is good. When the shoe is within the loop (less than 1 radius from the center) and within 6-12 inches of co-planar, even a crude system should work well. As you get more optimized, you could operate within 2-3 radii of the center or even more. If you want to get the transmit antenna off the floor, you run into parallelism and directionality problems. Perhaps mounting the antenna at 45 degrees with the axis pointing at the location you expect the shoe to travel through (doorway?) might work, or mounting it in the ceiling may work if you get your range high enough. Hope this helps. Don >size limitations - that's the most important. It has to fit in a shoe, >so I'm aiming for about an inch square or so and very thin - about 3-4 >mm. It has to fit inside an insole. My circuit (PIC, EEPROM, telemetry >transmitter) fits this criteria at the moment, but as I say I am using a >switch to tell it when it is within range. I'd like this to be >automatic. >cost - < $100 >limitations - it has to be inside the shoe - well, I'm starting to think >about mounting it under the instep, which might permit infra-red >approvals required - let's not worry about that for the moment :-) >hobby or commercial - well, university research - somewhere in between >power available at each side - plenty at the receiver (it's a PC - the >reciver - a Linx LC - goes into the RS232), about 10 mA max at the >transmitter >reliability requirements - I'm flexible at the moment, but clearly this >would have to be good in the long run >Note that I am not trying to send the data this way - I already have a >conventional RF transmitter. I just need something to tell it that it's >in range of the receiver. >So, are you saying that an RFID tag wouldn't cover much distance? > >Chris -- http://www.piclist.com#nomail Going offline? Don't AutoReply us! use mailto:listserv@mitvma.mit.edu?body=SET%20PICList%20DIGEST