Cheap(ish) rain sensor?

[3v0]

A while back I looked at the datasheet for one of the automotive type rain sensors. It was a scaled down version of the traces on PCB sort of thing.

Guessing here:
The sensor can sense a single drop of rain. At speed the air moving through the sensor dries each drop. Each drop would show up as a pulse. From this you could determine how hard it was raining. Maybe good enough to set the wiper speed.

 

[mneary]

With insulated (soldermask) traces, and a DC bias on the traces, the rain drops might show up as a sudden increase in capacitance. Charging that capacitance would cause a brief impulse of current.

 

[Rolf]

It might not........

With insulated (soldermask) traces, and a DC bias on the traces, the rain drops might show up as a sudden increase in capacitance. Charging that capacitance would cause a brief impulse of current.

Water is not a good insulator it can therefor not be a dielectric in a capacitor. See:

"ca·pac·i·tor (kə-păs'ĭ-tər)
n.
An electric circuit element used to store charge temporarily, consisting in general of two metallic plates separated and insulated from each other by a dielectric. Also called condenser."

So any change in current will be because of a change in resistance.

 

[gramo]

But if the sensors were already insulated, then the water "landing" between the two "plates" would alter the capacitance of the sensor. Similar to how touch sensors work

 

[gramo]

Water would have a completely different dielectric constant to that of air I would think

 

[Rolf]

Just a little......

Water would have a completely different dielectric constant to that of air I would think

Dielectric constants of some materials at room temperature Material Dielectric constant
Vacuum 1 (by definition)
Air 1.00054
Polyethylene 2.25
Paper 3.5
PTFE (Teflon(TM)) 2.1
Polystyrene 2.4-2.7
Pyrex glass 4.7
Rubber 7
Silicon 11.68
Methanol 30
Concrete 4.5
Water (20°C) 80.10
Barium titanate 1200

 

[Hero999]

Most water has salts dissolved in it making it a conductor therefore the dielectric constant is meaningless.

 

[Roff]

Most water has salts dissolved in it making it a conductor therefore the dielectric constant is meaningless.

Yeah, but if you just look at impedance, it's going to be low compared to air, salts or not.

 

[Rolf]

Most water has salts dissolved in it making it a conductor therefore the dielectric constant is meaningless.

Thank you for putting it so plainly, I agree 100%. I have tried to get this across in a more subtle way.

 

[mneary]

If the raindrop is salty, then it bridges a gap between insulated pads. Reducing the gap increases the capacitance. If the raindrop is more pure, then its dielectric constant increases the capacitance.

I think that you can depend on a raindrop influencing the capacitance one way or another.

 

[philba]

{snip}

The originator of this thread stated that he just wanted a rain indicator, NO amount.
I didn't think you could have dew unless the object was colder (dew point) than the atmosphere. This sensor will be at the ambient temperature most of the time so I don't think there should be any problem. If there is it can be fixed with a small heater, light bulb under it.

the sensor has to be uncovered so it will take dew in early morning.

Also, another post got me thinking that the amount of pollution will also effect the reading of this kind of sensor although presence vs absence may be ok.

Finally, I still think IR may be a better approach - look at this product for inspiration - **broken link removed**

 

[Rolf]

Experiment..........

If the raindrop is salty, then it bridges a gap between insulated pads. Reducing the gap increases the capacitance. If the raindrop is more pure, then its dielectric constant increases the capacitance.

I think that you can depend on a raindrop influencing the capacitance one way or another.

Salty or slightly acidic water is not a dielectric so there will be no capacitance!
Experiment: Take to metal plates about one square inch and place 4 layers of writing paper between them, hold them together with a close pin and measure the capacitance. It should be around 50pF, if my meter is accurate but that is of no consequence because of the next step. Now wet the paper thoroughly with saliva (to simulate moisture) and take the measurement again.
The capacitance has disappeared!
Why, because it has turned into a resistor!

di·e·lec·tric (dī'ĭ-lĕk'trĭk)
n.
A nonconductor of electricity, especially a substance with electrical conductivity of less than a millionth (10-6) of a siemens.

 

[mneary]

An important detail is missing. Your spit-saturated paper isn't even remotely similar to solder-mask insulated copper traces.

Take two metal plates and cover each with an insulator (such as a plastic bag, or plastic wrap). Then separate them with 4 layers of writing paper. Measure. Now wet the paper throughly with rain water. (We want to detect rain, correct?) Measure capacitance again.

I hope you'll forgive me for not doing this myself, but rain water isn't readily available here.

 

[Rolf]

An important detail is missing. Your spit-saturated paper isn't even remotely similar to solder-mask insulated copper traces.

Take two metal plates and cover each with an insulator (such as a plastic bag, or plastic wrap). Then separate them with 4 layers of writing paper. Measure. Now wet the paper throughly with rain water. (We want to detect rain, correct?) Measure capacitance again.

I hope you'll forgive me for not doing this myself, but rain water isn't readily available here.

So when did the traces get covered with solder-mask; I must have missed something?

 

[mneary]

So when did the traces get covered with solder-mask; I must have missed something?

8:01 PM yesterday (PDT).

You quoted it in your rebuttal.


[edited to be more positive.]

 

[mneary]

I've been looking myself for some indications of how the dielectric constant does change with impurities in water. I finally found a paper at Stanford www.slac.stanford.edu/econf/C06092511/posters/WEPO06_POST.PDF That shows that the measured dielectric constant of water rises dramatically with trace impurities.

 

[Roff]

An important detail is missing. Your spit-saturated paper isn't even remotely similar to solder-mask insulated copper traces.

Take two metal plates and cover each with an insulator (such as a plastic bag, or plastic wrap). Then separate them with 4 layers of writing paper. Measure. Now wet the paper throughly with rain water. (We want to detect rain, correct?) Measure capacitance again.

I hope you'll forgive me for not doing this myself, but rain water isn't readily available here.

Yeah, I hope that's not spit falling out of the sky!
I think the conductivity of rain is highly variable, depending on what is in the air where it forms and/or through which it falls. Apparently, most rain is very pure, and basically non-conductive, but acid rain is highly conductive. Of course, even very pure rain falling on a dirty sensor will probably be fairly conductive. The fact that water is somewhat conductive may not totally negate the capacitance of a drop of it. A dielectric constant of 80 is huge, so a drop of water may look like a parallel RC network. The time constant would be a function of the ion concentration. As I said, regardless of the conductivity, the impedance of a drop of water (at high frequency, which I meant to say) will be drastically lower than that the space which was where it fell. However, with parallel traces, the air gap capacitance is along the entire length of the trace, whereas the water droplets may be few and far between.

 

[mneary]

Other details that might have been missed; the fingers need to be biased so that the new capacitance needs an electron flow to charge it. And, the proposed sensor would detect but not measure rainfall, since drop size and purity are highly variable.

Another variable is electrostatic charge; some droplets might not be neutral.