I have one of those radio controlled projection alarm clocks that's supposed to keep time by picking up the radio time signal from Cumbria (it used to be Rugby within spitting distance from here but it was shifted up north recently).
Anyway, I noticed yesterday morning that the clock thought it was 3AM. Fortunately as it was Saturday I hadn't missed the alarm and been late for work however it did highlight the fact that the clock was no longer functional.
An ideal candidate therefore for repair and conversion to run off my solar installation!
Firstly I verified the fault wasn't down to something I could repair. Although the clock still worked off batteries, it didn't work off the mains adaptor despite the adaptor and mains input socket being okay. I checked what I could inside the clock but I'm not clever enough to verify where the fault lay so I went ahead with the solar conversion.
The solar conversion would be easy enough. As the clock still worked with two AA batteries inserted (3V), all I had to do was convert the 12V output of the solar battery into 3V and connect it to the battery terminals. The only other thing to do was rewire the LED used for projection to operate from this input as by default the LED didn't work from the batteries because of it's high current drain.
The wallpaper is almost nasty enough to make me want to get up in the morning without this alarm clock
In past projects I have used the trusty 7805 voltage regulator to reduce the 12V output from my solar battery to 5V but for this conversion where 3V was required I decided to try something slightly different and I built a simple Zener regulator as below:
I haven't used a Zener diode for years - I really must use them more as they're jolly useful! Mine was a 3.9V Zener that I've had in my junk box for years and its job is to convert Vs (12V) to a steady 3.9V (Vz) despite current fluctuations. It works because the Zener diode allows a variable current to flow through it while keeping the Voltage at a steady level. This means the regulator reduces the 12V source to 3.9V which is closer to what the clock needs to operate. To further reduce the voltage, a 1N4004 diode is placed in series with the clock input which has the effect of reducing the Voltage by a further 0.6V so a fairly healthy 3.3V remains to drive the clock. In parallel with this extra diode is a 100 Ohm resistor which reduces the 3.9V regulator output to 2V at 19mA to drive the projection LED separately.
Why a 220 Ohm resistor? Well, to calculate resistance value you need to know the output Voltage and current. I've based this on a maximum current of 37mA as that should be more than enough to drive this clock. The formula for calculating the value of the resistor is Ohms law, i.e.:
R = V / I ...or... R = Vdrop / Imax
In my case... R = (12 - 3.9) / 37x10-3 ... so R = 219 Ohm (220 Ohm in practice)
The power must also be calculated for both the resistor and the Zener diode to ensure the components used are with tolerance, so...
Pr = Vdrop x Imax, therefore, Pr = (12 - 3.9) x 37x10-3 = 0.3W
Pz = Zv x Imax, therefore, Pz = 3.9 x 37x10-3 = 0.14W
So the resistor and diodes used must be able to handle these power levels (i.e. a normal quarter Watt (0.25W) resistor will not be up to the job)!
With the components sourced from my box of spare bits and the circuit tested on a breadboard, construction began. I soon had a little stripboard knocked up as below.
The largest component is the high power 220 Ohm resistor while the Zener diode sits at the bottom of the board and the two outputs, one to the clock via the 1N4004 diode and the other to the projection LED via the 100 Ohm resistor, sit to the right. The other component which looks like a green resistor is actually a 250mA fuse inserted for safetys sake.
The whole assembly fits into the battery compartment of the clock and the output of the regulator is connected to the battery terminals.
As it is Sunday night now that I type this, I guess I'll find out if it works properly tomorrow morning as I'm relying on this thing to wake me up for work!