Go Down

Topic: pressure sensor (Read 2351 times) previous topic - next topic


I want to be able to read data based on airflow, more specifically someone blowing. I read somewhere that a pressure sensor might work if you're blowing into it (which wouldn't be a problem). Has anyone here done something like this before?



There are several questions you need to answer here, beyond 'measuring pressure'.

1) What's the range?  Lung pressures are actually rather low, but you'd be in better shape
if you measured your own pressure to get an idea what pressure you're talking about.

2) How much resistance should the test subject blow into? It's a different act to blow into
a plugged tube (like blowing into a commercial sensor) vs blowing up a balloon vs. blowing into free space.

3) the biggy - what sort of accuracy do you need?  I'm always suprised by the number of people who build gizmos to measure accurately, then only use it for:

if (pressure > THRESHOLD_PRESSURE) {
     do something...
} else {
    do something else....
If that's what you're doing, you can get away with a simpler system.

4) How fast do you need to take measurements?  Are we dynamicly measuring lung action, or just measuring 'how hard can you blow'?

5) What precautions do you need to use to protect subject from sensor, and vice versa?

ok, as for the actual sensor:

low pressure pressure gages are of three types -
a manometer is just a tube with a liquid in it. You
measure the change in height of the liquid. This could be as simple as a resistive element dipping into a conductive solution, like a graphite pencil lead hanging in a salt water solution in a U tube.
You can increase the sensitivity by tilting one arm of the U tube til it's almost flat, so a small rise in the level of the liquid means a large change in the position of the meniscus

Another common type is a venturi type. Make a conical hole in a plastic block, like a narrow funnel. Put air into lower end, drop a lightweight ball into the top, ball rises further with more pressure.  Good idea to put a crosspin in at top to retain ball.

Finally, you can use a diaphram - some sort of rubber/plastic/metal disk in the end of a container. As the pressure increases the diaphram moves a bit. Usually you have to mechanically amplify this - this could be as simple as a lever arm arrangement. To measure the position of the end of the arm, use a card with a slit of varying width, and a photodetector.

Most commercial gauges use a strain gage permanently bonded to a diaphram. This is a bit hard to interface to. Making your own sensor is easier.    :D


I am not an expert, but I think it would be better to use a THERMISTOR, calibrated at the room temperature, assuming the room temperature is stabel. This way, running or waving people (who move air, but aren't blowing) doesn't get meseared.

Hmzz thinking about it, Just place it in the end off a hollow tube or something, so it will  only detect straight blows.. than you can use apreasure sensor as well.


 I'm not sure how using a thermistor will help with the problem of incidentally air movement (people waving, etc).

 Do you mean a thermistor and a resistor used as a heater?  That works, but seems
like a lot of added complication.

  A simpler, heat based system is to put a piece of nichrome wire in a tube. Heat the
wire with a regulated current, with a resistor in series. Measure the voltage drop across
the wire, which will vary with the temperature. The temp in turn will be some amount
above ambient that varies in a monic down function of the air velocity.
  You also need to enforce the user blowing for long enough for the system to stabilize, but
it's pretty fast.


;) I am new to this, but I read that a thermistor is a resistor that changes when the temperature changes.
Running and waving people do not change temperature, only blowing people do.
So just read the restance change and you would know if some one is blowing (or not).


My impression was that the original poster wanted to actually measure pressure, not
just 'know if someone is blowing'. If that's all they want, a diaphram that's deflected into a contact will suffice.

Go Up