One of the companies on which my family relies has a problem: people keep leaving their front door open. I've offered to make them a simple buzzer to alert them to a door that's been left open for more than a minute.
I'm taking this opportunity to brush off my analog electronic skills, which have almost entirely rusted since I was a child. I could just buy an off-the-shelf item, but think it'd be more fun to build it from scratch :) I'll keep a detailed build log on this page ...
I've assembled the veroboard; I'm particularly pleased with my design which fits comfortably on the cut-out board, and also involves no track breaks except for those under the IC.
Assembled circuit on veroboard
I used an hacksaw to cut the veroboard to fit the enclosure (and around the piezo buzzer). I then used a fine point on a dremel to clean between the tracks at the point of the cut, as the saw blade had torn the copper about and bridged several of them.
The enclosures arrived today. I'm very happy with the manufacturing quality for such a low-cost item. However, and this is entirely my own fault, the enclosures are too shallow for the 1000µF capacitor I selected for the circuit. So I've increased the resistance from 47kΩ to 470kΩ, and decreased the capacitance to 100µF using a lower-profile axial lead capacitor.
Here is a photo demonstrating the height problems:
Capacitors clearly too tall for the enclosure
... and here is the fifth prototype, featuring the new components:
I'm fairly sure the replacement capacitor dates back to my Funway 1 days as a child. The leads needed some cleaning up, but the timer tests out at ~ 59 seconds so I presume they're still good.
DSE Funway Into Electronics kids
Some additional good news is that the reed switch magnets are very strong. This means that I can mount the reed switch entirely inside the case, and the door-frame-mounted magnet will still switch the device from more than a centimeter away. This will allow for a much cleaner and more robust assembly.
The reed switches arrived - nifty three-lead units which can be used in either NO or NC mode. Here's the fourth prototype, with the power routed through the reed switch:
Once installed, this will mean that the unit draws no power at all when the door is closed. When the door opens and the magnet moves away from the reed switch, the switch will supply power to the timer which will in turn start the countdown.
I've ordered two of the enclosures pictured below. It's now a matter of waiting for the enclosures, reed switches and bezels to show up so I can determine how large my veroboard can be. That will in turn determine how I lay out the circuit. I anticipate at least a week's wait at this point.
Found a nice enclosure from plastboxuk.com (now defunct) that should work. It features a separate 9V battery compartment, which is ideal.
Here is the third prototype, complete with flashing red LED to indicate that the timer is running:
I've measured the current at < 6mA when the buzzer is not active, and < 20mA when it is. The readings fluctuate a fair bit due (I think) to the flashing of the LED, but running with the maximum figures I should get > 90 hours out of Duracell 9V, or 29 hours if the buzzer were operating constantly.
A thread on the mAh rating of 9V batteries
That's assuming that the circuit would function normally with a nearly depleted battery. I'm not sure if it would (I doubt the voltage and / or current supplied would remain high enough) but it suggests I'm on the right track by powering it from a 9V battery.
I've also ordered some chrome LED bezels, to make the whole thing look nice. Finally, I've ordered some reed switches to detect whether the door is open.
_Edited to clarify_: Armin has pointed out that wasn't too clear. In this design the power is routed through a magnetic reed switch, so it only draws from the battery if the door is open. So the 90 hours I mention represents 90 hours with the door in an open position, which hopefully will represent at least weeks of operation, possibly months. We'll see.
I've built the first prototype of the door timer, which illuminates an LED after around a minute (58 seconds, actually).
It's built from the monostable timer circuit on the 555 Timer Circuits site, but with the LED connected from Vcc to pin 3 (known as sourcing).
As it turns out, that green LED was under-rated for the current flowing through it; it started to turn orange as in [this video](https://www.youtube.com/watch?v=foNi-keKGs4). I've subsequently replaced it with a blue LED that works fine.
Here's the second prototype, with the actual buzzer, and an LED indicating that the unit is operating (so you know that it is, for safety reasons):
Found a handy resource full of 555 timer info. Will peruse before continuing. Also want to find out about 555 power usage; would like to power the beastie off a 9V battery but I'm concerned about battery life.
---
