Effective Alarm Clock

Inspiration

I was inspired to do this project by my own severe issues waking up in the morning and staying awake. This has taken a toll on my productivity, so I knew I had to do something about it. Having tried everything conventional, I thought I would try a hardware hack to help myself wake up.

What it does

My project is pretty basic. It is a conventional digital alarm clock with the tiny piezo speaker replaced with a large electric alarm bell. It is very loud, and sure to wake up even the deepest of sleepers.

How I built it

The essence of this project is the interface between the alarm clock's speaker output and the alarm bell's power input. The interface consists of a 12V "wall wart" power supply to power the bell, a MOSFET to turn the bell on and off, and an optocoupler to safely turn the MOSFET itself on and off without electrically connecting the clock to the bell.

Challenges I ran into

The main challenge I encountered was reliably turning the MOSFET on and off with clock's speaker output. These issues could have been overcome if I had used a modern MOSFET instead of one probably designed in the 1980s, but given the time constraints, I had to use what was available in my parts collection. Also, it was fun to take a break from my day job (graduate research on power conversion using modern high performance transistors) to play with some old school components. If you are familiar with power MOSFETs (especially old ones like the IRF513 I used), you will know that using them invokes two major challenges:

1. High gate threshold voltage - the MOSFET's gate input must be connected to a relatively high voltage (4 V) before it will even consider passing any current to turn on its load, and much higher (>10 V) before it will fully turn on and pass full current through its load.

2. High gate capacitance - a great deal of charge must be passed into the gate to reach reach 10 V for full turn on, meaning relatively high current to turn the MOSFET on quickly. In addition, this charge must be taken back out of the gate to turn it off.

Since the speaker output of the alarm clock was both low voltage (3 V) and low current (too low to measure), I found it useless for turning the MOSFET on. While there were obvious solutions like a power BJT (needs only 0.6 V to turn on) or low voltage MOSFET (~2 V), I had neither on hand while working on this project.

To overcome these issues, I found an optocoupler in my parts bag and connected it as an amplifier to turn on the MOSFET. If you don't know what an optocoupler is, it is essentially an LED and a phototransistor in one package. The LED takes only about 1 V at low current to turn on (easily provided by the alarm clock), and the phototransistor can pass hefty currents (sizable fraction of an amp) or block a voltage well in excess of the 10 V needed to fully turn on the MOSFET.

Since the optocoupler can only turn on the MOSFET, I ran into the issue that after the alarm was supposed to turn off, the bell would keep ringing for a long time. This was caused by the high gate capacitance maintaining a high gate voltage, keeping the MOSFET on. To fix this issue, I simply connected a high value bleeder resistor between the gate and ground to make sure the MOSFET would turn off in a reasonable amount of time once the optocoupler turned off, while not pulling enough current to significantly slow down turn on.

Since the alarm bell needed about 12 V to ring loudly anyway, I decided to use that as the MOSFET gate voltage to simplify the power supply design.

In the end, this two stage drive solution only required three basic components (optocoupler, MOSFET, resistor), keeping it very simple and inexpensive.

Accomplishments that I'm proud of

Above all else, I'm proud of successfully completing my first ever hackathon project. Secondly, I'm proud of making a simple and useful product that doesn't need the internet, a server, or even any code at all to run successfully and is based on proven, reliable technology. Thirdly, I'm proud of finishing the project very early, so I can get some much-needed rest.

What I learned

My main lesson was that you should never expect any drive strength at all from an unknown signal source (alarm clock speaker signal in this case), and always be ready to do one or two stages of conditioning before its ready to do what you need it to do (ring a bell).

What's next for Effective Alarm Clock

Next, I would like to recreate this device only using parts that are mass-produced and cheaply available, since the bell I used came out of an old piece of test gear, and I'm not sure if it manufactured anymore.

Built With

• alarm
• clock
• mosfet
• optocoupler
• solder
• wire