Author Topic: Fridge alarm design review  (Read 1567 times)

0 Members and 1 Guest are viewing this topic.

Offline TGETopic starter

  • Newbie
  • Posts: 2
  • Country: be
Fridge alarm design review
« on: May 24, 2017, 08:38:04 pm »
I’ve designed a fridge alarm to prevent the door from being left open without any of us noticing. It is a working design and it meets the following design criteria.

Design criteria:
Works off 3V CR2032 cell
Can last for years on one cell
  No (=0) current when off
  Very low current when fridge is opened but alarm is not going off (it’s a couple of uA in practice, so this is negligible)
Alarm either goes off or it doesn’t (threshold)
Alarm goes off (virtually) instantly when door is closed
Time between open door and alarm can be set (using a pot)
No micro-controller; as simple as possible

Explanation: when fridge door is opened, the capacitor is slowly charged and the voltage climbs. Once this voltage reaches the reference voltage (which can be set using the pot), the piezo buzzer and red LED activate. When the door closes (regardless of whether the alarm went off, the voltage at the base of the PNP goes low and the cap quickly discharges.

General question: can anything be improved? Are there more elegant solutions? Am I making something unnecessarily complicated?

Specific questions:
Because my reference voltage only available when the switch is closed, once the switch opens again (the door is closed), my reference voltage drops below capacitor voltage and it fails to deactivate. The OpAmp power rail is only active when the switch is closed, so this is not a problem, but I was wondering whether it is bad practice or something? It is trivial to solve this ‘problem’ by having the voltage reference on at all times, but this would consume a bit of power. Granted, it would be very little, but nothing is better than very little.

It is my first ‘design’; I’ve made this completely from scratch for an actual, practical problem that I had. 

I apologise in advance for the rather messy schematic.

Any kind of (constructive) feedback is more than welcome. Thanks!
 

Offline DrGeoff

  • Frequent Contributor
  • **
  • Posts: 793
  • Country: au
    • AXT Systems
Re: Fridge alarm design review
« Reply #1 on: May 26, 2017, 03:28:08 am »
Consider adding a temperature threshold as well. In case something else happens that causes the temperature inside to rise, and potentially spoil the contents. This could be anything from a fuse/breaker to faulty compressor, door seals etc. A reed switch will only tell you that the door is open, not that the interior is no longer cool.
Was it really supposed to do that?
 

Offline TGETopic starter

  • Newbie
  • Posts: 2
  • Country: be
Re: Fridge alarm design review
« Reply #2 on: May 29, 2017, 12:39:12 am »
I made this mainly to prevent the very common problem that I or someone else closes the door a bit too roughly and it opens up slightly. For every other problem, I probably get 100 open doors, so this was by far the priority.

You are right of course that adding a temperature sensor would be an improvement, but I'm not sure how to combine this with the low current requirements that I need in order to keep my design "stick it on and forget about it". Any ideas?
 

Offline DrGeoff

  • Frequent Contributor
  • **
  • Posts: 793
  • Country: au
    • AXT Systems
Re: Fridge alarm design review
« Reply #3 on: May 29, 2017, 01:02:46 am »
I made this mainly to prevent the very common problem that I or someone else closes the door a bit too roughly and it opens up slightly. For every other problem, I probably get 100 open doors, so this was by far the priority.

Try tilting the fridge slightly backwards? Then the doors become self-closing.

Was it really supposed to do that?
 

Offline Paul Moir

  • Frequent Contributor
  • **
  • Posts: 926
  • Country: ca
Re: Fridge alarm design review
« Reply #4 on: May 29, 2017, 01:17:29 am »
Mechanical thermostat?  If that's too pedestrian, how about a microswitch pushing on an ice cube? 

The 10uF that gets dumped by the PNP:  I'm guessing it's not much energy or maybe the PNP switches slowly so it doesn't matter, but perhaps it wants a small resistor to limit the current through the PNP? 

What's the comparitor, buzzer current, and the NPN transistor?  Little worried about the base current on the NPN transistor.
 


Share me

Digg  Facebook  SlashDot  Delicious  Technorati  Twitter  Google  Yahoo
Smf