diy Airflow Sensor (Sensing when dust filter is clogged)

[dfidler]

Hi Guys,

I'm thinking of taking on a chassis build project and was thinking of moving air through it. However, I wanted to put a large dust filter in there to prevent dust buildup.

Unfortunately, dust collects on filters so I want to figure out a way of sensing when the airflow is reduced through the chassis. The first thing that comes to mind is to use an airflow sensor that, when the airflow (either measured velocity or mass) drops below a certain threshold, it lights an LED (at first and, later, when my electronics skills improve, sends a trap to a logic circuit that deals with it).

Has anyone done anything like this? Can you recommend any cheap parts for the sensing? Or can you suggest an alternative?

Cheers,
Dave.

[davidsrsb]

A small paddle on a microswitch in the airflow should work.
The fancy solution is measuring the voltage drop on a thermistor or diode with a heating current.

[Enzo]

If the filter is such that you normally can see through it, use an optical detector to sense when the light coming through is blocked.


Or just note how long it takes to get dirty and clean it now and then pre-emptively.

[dfidler]

Originally Posted by Enzo

Or just note how long it takes to get dirty and clean it now and then pre-emptively.

Great ideas so far, but as to this last one, I'm just not that organized (or pre-emptive).

[Ron E]

The airflow is for cooling, presumably. Maybe all you need is a temp sensor.

[dfidler]

Originally Posted by Ron E

The airflow is for cooling, presumably. Maybe all you need is a temp sensor.

Yeah, I was thinking of that as well. Of course, dust can impact the accuracy of temp sensor, no?

[wakibaki]

You could try a pressure sensor. Here's a 0-0.5psi one from RS:- Honeywell | Process Control | Sensors and Transducers | Pressure Sensors | Piezo Resistive Sensors |24PCEFA6D

A sensor of this type costs about UKP12, they come in absolute and differential types.

It all depends on how much differential pressure you get in the chassis. This is fairly easy to determine with a manometer. You just need a pressure-tight spigot on the chassis leading to the atmosphere that you can connect some clear plastic tubing to. You make a u-bend in the tube and pour in some water. Turn the fan on. The water will move in the tube and you can measure the difference in inches between the top level and the bottom. 32 feet of water is approximately 1 atmosphere or ~14.7 pounds psi, so 0.5 psi is about 12" of water pressure.

You could actually fabricate a complete sensor with a bit of tube and a couple bits of wire shoved into it, with a circuit to sense when the water is shorting the wires. The problem with this is that the water needs some ions (disinfectant?) to make it conductive, and the lifetime of a homemade sensor will probably not be long due to corrosion or contamination.

The simplest way to use one of the off-the-shelf sensors is to include it in a Wheatstone Bridge, but the circuit may need some temperature compensation, which can be tedious to get right.

w

[Steerpike]

At university we used an ordinary lightbulb with the glass broken away - just the naked filament in the air pipe. It's heated by applying constant current & the resistance measured by monitoring the voltage across it. Reduced air flow makes it heat up fast, making the voltage rise.