Lately i found some cheap LED modules at a german electronics junkdealer. They have some really nice stuff (mostly optoelectronics). The most of it looks like the left-overs of industrial production that is not needed any more, but most times it’s just B-class ware.
Okay, so i bought five LED modules for <2 euros and built this thingie - a radio controlled LED matrix clock! Woot! 
The most expensive part was the DCF77 receiver module (antenna+a little receiver PCB), it cost me about 12 euro which is pretty much for such a small module. DCF77 is the standard longwave time signal in germany used in most (all) radio controlled clocks.
I used an ATmega8 microcontroller to control the LED matrix and decode the output of the DCF77 module. The matrix itself is controlled by using latch registers and UDN2981 as a source driver to drive the rows and ULN2804A to drive the columns (it’s a darlington array). The row driver gets a little warm after a while, but it’s okay (I am planning to add another LED module anyways, so it is going to get even more warm).
Okay enough blah blah, here are some photos and a video of it in action, click on the image to view it in full size (warning: the images are very, very high resolution :))
I am especially proud of the way i run the databus over the latch registers. The latch registers (and the driver ICs which fit directly to the output pins of the latches) are below the stripboard in the middle. Before soldering the stripboard there i wired everything up to the LED modules by using magnet wire (the one used for winding coils). Doing it this way kept the “wire wraping” to a minimum and i didn’t end up in a mess of wires
The little crystal you see on the microcontroller is taken off a Pokémon Mini <3, it turned out to be a standard 32kHz clock frequency crystal used for the RTC of the PM. On the right side you can see a potentiometer to control the linear regulator, this way the operating voltage of the whole circuit can be adjusted to the right value and it is possible to dim the LEDs (just a little bit until the whole thing turns off :P).
Maybe you ask yourself where i put the resistors for the LEDs because you think that all LEDs need series resistors. Well, the answer is that i didn’t use them, mainly because i am lazy and secondly because they take too much board space. If you tune the voltage of the circuit right the LEDs will operate with something about 1,8-2,0 Volts because the driver ICs reduce the voltage (i don’t exactly know by what amount).
This way the voltage is high enough for the rest of the circuit but low enough to drive the LEDs with about 20mA power consumption (higher voltages would destroy the LEDs or shorten their lifetime). I was told it would not work this way, but apparently it does
The front of the module, here you can see the matrix modules. The edges of the PCB will be cut off and i will add another PCB on the right side to hold another matrix module.
One of the most annoying problems I encountered was the interference created by the rest of my circuit, it turned out the multiplexed LED matrix creates a high noise (i think it’s called ripple) in the supply voltage of the receiver module and the receiver won’t work while the LED matrix is multiplexed at high frequencies. I had to add various capacitors to the DCF77 receiver module to get it to operate more stable while the LED matrix is on. I ended up with 100nF, 47nF and 100nF capacitors and one 499 Ohms resistor used as an RC circuit. This way the module is supplied with a stable and clean voltage.
The movie shows the clock in action displaying correct informations (Xvid)
Special thanks go to Tsukasa for taking the photos and movie. Finally some high quality pics on my blog and not these blur mobile cam pics 
