Dust detector with light sensor.

[Haider]

So, my question is if it would be possible to use a normal light sensor for detecting if a certain place need cleaning. The plan is to put it somewhere in my computer where sunlight doesn't reach, but is still a great place for detecting dust build up, having a LED (otherwise it would reach its max resistans?) above it and when the resistans of the light sensor exceed a certain amount of ohms (because of dust covering it, darkening the light from the LED) make my start button pulse. Is there any chance this would be able to work?

What I consider a "normal" light sensor  http://assets.controlanything.com/photos/PDV-P8001-900.jpg

[Len]

I don't know if your idea will actually work, but...

There are photointerrupters with the LED & sensor in one unit:
http://www.digikey.ca/product-search/en/sensors-transducers/optical-sensors-photointerrupters-slot-type-logic-output/1967053
But it looks like the LED & sensor are recessed and less dust would gather on them.

There are also actual "dust sensors" but those detect dust in the air, not dust buildup.

Heck, your idea is simple enough that you should just try it. The main problem will be waiting for the dust to gather.

[Haider]

Thanks for the response.

Then I'm going to build it and see if I can make it work! 

[pickle9000]

Try a yellow LED, you will find it works better with that type of sensor.

You may also want to glue a magnet to a thermistor and try placing that on the power supply. As the supply gets dusty it will warm up and you should be able to use that as an indication. The magnet just allows you to mount the thermistor easily. Don't put it in the airflow / vent area and that should give a better indication of the issue.

[Zero999]

It sounds good in theory but the LED's brightness will decrease with age so it could cause a false alarm.

[bench_knob]

So, my question is if it would be possible to use a normal light sensor for detecting if a certain place need cleaning. The plan is to put it somewhere in my computer where sunlight doesn't reach, but is still a great place for detecting dust build up, having a LED (otherwise it would reach its max resistans?) above it and when the resistance of the light sensor exceed a certain amount of ohms (because of dust covering it, darkening the light from the LED) make my start button pulse. Is there any chance this would be able to work?

What I consider a "normal" light sensor  http://assets.controlanything.com/photos/PDV-P8001-900.jpg

Hi,

The first thing to determine is what type of dust you want to detect. Dust suspended in the air or dust that accumates on a surface?

Smoke detectors for instance, detect particles suspended in the air, but in any case these sensors are commonly available and are very easy to modify.

http://en.wikipedia.org/wiki/Smoke_detector

Then there are 'gas' detectors such as this one:

http://www.brighthubengineering.com/diy-electronics-devices/113566-make-a-simple-ic-741-smoke-detector-circuit-schematic-diagram-enclosed/

And an Arduino based detector,
http://www.arduinomojoe.com/projects/Security_System/arduino-diy-security-sensor-smoke-detector

So far the above detects suspended or airborne particles, if you are interest to detect or measure partical accumalution, that's an entirely different subject and the strategies will likely be very different.

If dust accumulation, the sensor need be easily cleanable.

You could measure blockage to air flow using a filter paper and a free-wheeling DC fan to act as an airspeed sensor, A/D converter output compared to a table. Or measure change in transluscency of a filter paper...etc

good luck and have fun

bench knob

[onesixright]

How about the GP2Y1010AU0F ?

Its a optical dust sensor, designed for what you want. Pretty common, round 10USD on eBay.

There are some arduino projects that show how it works.

[bench_knob]

How about the GP2Y1010AU0F ?

Its a optical dust sensor, designed for what you want. Pretty common, round 10USD on eBay.

There are some arduino projects that show how it works.

The detector you cited is essentially the same type sensor as a common-smoke detector, it detects fairly dense smoke, but not accumulated dust, like snow falling on a sidewalk, it piles up....how big is the pile type measurement.  For that type of measurement something else is needed. Years ago I worked for the physicist who developed the Cupie-Pie for the Manhatten Project. He and his son ran a gadget factory and one of their gadgets was an air sampling system. It was comprised of an indexed filter paper roll, each sheet, was stepper-motor placed into the airstream of the atmosphere return register of a nuclear power-plant office workers air-supply.  The gadget, timed each sheet, rolling up onto a takeup reel the exposed sheet, positioning a new sheet into the airstream. The takeup reel sheets were periodically measured with a scintillator spectrometer, a fancy ionizing radiation, (very sensitive) sensor.  Its purpose was the measure the possible accumulation (integrated over time) of nuclear particulate contamination.  A similar principle, yet simple system could be devised to measure dust accumulation. Only ya wouldn't be measuring radiation, but rather dust, and the sensor would be different.

bench knob

[G7PSK]

If you have a shiny surface inside such as plastic fan blades you could shine a light or infra red onto it and measure the amount that is reflected more dust less reflection.

[bench_knob]

If you have a shiny surface inside such as plastic fan blades you could shine a light or infra red onto it and measure the amount that is reflected more dust less reflection.

Yep, thats in the right ball-park...in any case, its a long period monitor.

 bench knob

[Rerouter]

another is if you have a heatsink on your cpu with straight blades, you could use a laser diode pointed through the blades, this will give you the best representation of dust as that's where its most likely to collect,