/diy/ - Do It Yourself

>quiescent current: 0.1µA
>operating current 20mA
Boy is that a big difference. The q current is so low that I could run this thing for a year on a 1F supercapacitor, but only a few minutes when woken. Which would be fine if I wasn't going to wake it for a while, so I guess I'll be going for a couple NiMH cells.

Bike lock designing anon here, after looking into how to set the MCU to sleep to preserve battery, I found that you basically have to do it with an external interrupt. So without an RTC or something similar burning my power, it would be impossible to check continuity every second or so. Hence instead I've decided to go for my FET-buffered continuity sensor, pic related.
R0 is the only resistor that current is flowing through in standby, so I want to minimise this, for which using one of my 10MΩ resistors should work fine, provided the MOSFET's gate plays nicely in that instance. I could possibly use a JFET if that would make things easier. I've tested this part of the circuit with a 1MΩ and it works fine.
The switch on the right is a key switch that goes in the end of the handlebar that disables the entire system, while the two wire loops come out of the other end. The batteries will be in there too. Not pictured is the alarm, which will need to be somewhere protected but also audible, and the charging system, which will basically be an inductor, rectifier, filter cap, and 5V regulator, with a magnet tied to a spoke to pass near the inductor on the fork.
It's pretty hard to test whether my off-the-shelf inductors and magnets will pack enough punch to get a decent voltage off a magnet, so that will have to wait until I have a bike to test it with.