Making Device to Send Data to Smartphones: 100m range, 1kb/s bandwidth, low power

Question

What I want to make: A device that immediately sends a notification if the device incurs a moment of high acceleration. The notification needs to denote the acceleration. There is no guarantee that there will be a The device needs to be very small, about the size of a penny, and not much thicker. The device should last for a year using a standard watch battery (about 80% of the time, the device will not be moving at all, the device wouldn't send more than 100 times a day).

Bonus qualities: - This notification would be sent to a smartphone for convenience and possible future visualization purposes. - The notification would be visible from up to 100m away from the device (this could be achieved via wireless signal).

What I'm asking for: Advice on through what medium this notification could be sent (Bluetooth, Zigbee, Wifi), and if it can't be sent to a smartphone, what it could be sent to to achieve the goals of the device (most notably, immediately notifying that the device was accelerated, and how much it was accelerated).

Answer 1

You indicated Bluetooth doesn't satisfy your requirements, and I agree. But that's classic Bluetooth. Are you familiar with BLE (Bluetooth Low Energy, part of the Bluetooth 4.0 spec)? The maximum range is anywhere from 50m to 450m, depending on the module used. (That last figure is not a typo, here is the BLE121LR -- LR for long-range. And it still runs off a coin cell battery.)

Apple added support for BLE starting with the iPhone 4S. BLE is also supported by Android OS (such as used on the Galaxy S series) version 4.3 or later.

I have seen BLE chips with maximum transmit power ranging from 12 mA to 39 mA. I'm currently using one that is 18 mA, so let's use that.

No, it will certainly not be transmitting continuously. The trick is to get it to transmit for a brief a time as possible.

If one assumes a connection is made and data is sent in 50 ms with a 10% duty cycle, and that is done 100 times per day, then the power required is 18 mA * .05s * 0.1 * 100 = 9 mAs / day (where mAs is milliampSeconds).

Spreading that over a year, gives 9 * 365 / 3600 = 0.9 mAh per year. Even if I got the communication time off by five times, that would still be around 4.5 mAh per year.

Assuming the microcontroller is only active when sending data, it should use no more than what the BLE device uses, so that is 0.9 to 4.5 mAh / year.

The problem, I believe will be the accelerometer.

You said 20% of the time, the object will be moving, and there will be less than 100 peak events per day. I am going to assume that there is some sort of trigger in the accelerometer that allows it to remain in low power until the 50 G event is seen, and only then wake-up the microcontroller.

Assuming the low-power (but active) mode of the accelerometer takes 25 µA, then 0.025 mA * 24 * 365 = 219 mAh. You haven't provided the part number of the accelerometer so I don't know if the 25 µA figure is reasonable. I took it out of a spec sheet for a Freescale accelerometer I have used. I am not including the extra power needed when it is connecting to the microcontroller, but that will be very small and can be ignored.

219 + 4.5 + 4.5 = 228 mAh, less than 240 mAh, but not much.