A fully assembled and partially assembled circuit board next to the top and bottom of the key fob. (Click to enlarge)
Back to the main page for this project.
A fully assembled and partially assembled circuit board next to the top and bottom of the key fob. (Click to enlarge)
So close. One mistake on the PCB. One more feature I want to add to the PCB. The buttons are too tall. And the LED current limiting resistors are too small. In the scheme of things, I'm almost done. See the previous entry for more information.
Over winter break, I made another prototype that replaces all digitial part of the circuit with an AVR microcontroller, which drastically cuts down the part count and power consumption. I've ordered a PCB that will fit everything inside a keyfob, so this project may be very close to completion! I've moved back to an LED display since the small LCD proved to be too hard to use. To get around the LED current draw program (which caused me to try LCDs in the first place), the microcontroller turns off the LED while it's using the random bit generating part of the circuit. The LEDs are also pulsed on and off to further lower current draw. This design also has a new feature -- the ability to choose the upper end of the RNG range using a potentiometer.
Here are pictures of the prototypes mentioned in the previous update. The smaller LCDs came in some time ago. I haven't had time to make much progress on this project, but that may be changing.
Well, I've got two working prototypes. There are two major problems left to solve. First, I need to find a small LCD -- the smallest single digit LCD I've found has a 1.36" large character. There are smaller 2.5 digit displays, but they're still larger than I'd like. However, I've bought a batch of very small LCDs from American Science and Surplus. They haven't arrived yet, but hopefully they'll work out. The other problem is that using 7400 series integrated circuits takes up too much space, since I need a bunch of them. I'd like to use a simple programable logic device, but they use to much power. So, I'm thinking about using a microcontroller -- maybe AVR rather than PIC (which I used for the previous RNG project). Either way, I may be able to get away with only one chip, and use less power to boot. Sweet.
I've been working on this project off and on (more of the former than the latter) since the last update. The basic idea is the same although some of the details have changed. For example. the chip I was using to drive the LCD wasn't very good for the task (it turns out LCD elements aren't supposed to just be turned on, they are supposed to be pulsed on and off very quickly). So I switched to a different display driver which made this easier to do. To drive the signal off and on, I needed another clock, so rather than adding another 555 timer, I used a 556 which has two timers on it. Finially, the method I'm using to get randomness out of the voltage regulator has changed again -- I'm using a transistor to amplify it. The device works in that it displays numbers that are apparently random, but I have doubts -- it looks like the circuit that's supposed to be random is just a really fast oscillator, so it's not truely random. However, the whole shebang now runs off a 20mm coin Li battery, and includes a smaller LCD, so progress is being made... slowly.
I now have a working prototype using an LCD display, instead of an LED display. Pictures and schematics to come. Here are the datasheets that matter.
Well, it's all working, but not correctly yet. The LEDs draw enough current to change the voltage across the battery enough to screw up the analog part of the circuit. Because there's a capacitor between the op-amp input and ground, I can't use the op-amp to reject these voltage fluctuations. I'm reworking the analog part to get around this problem... I'll post more information when it's working...
