leak detector

Back in the '60s a hobby magazine used a TV antenna clip and an aspirin to detect water. The aspirin dissolved when it got wet, and allowed the contacts to close. No fancy electronics needed, and it was a good way to get rid of out of date aspirins, too. The hard part today would be to find the TV antenna clips made for the VHF terminals.

on

en

t,

und

me

es

uoted text -

Alligator clip? Gotta love the low tech solution.

George H.

quoted text -

No. A TV antenna clip. They looked like a plastic clothespin, with a set of notched metal tips that went around the screws for the 300 ohm twinlead. I haven't seen a new one in years.

Interesting. I used my Fluke 189 and got about 30k ohm. That's a big difference.

We may be seeing the effects of the probe material and bias voltage. The 'cell potential' that Phil mentioned earlier in this thread. A full IV plot might be informative.

I used to randomly selected probes. Maybe I'll try some others.

George H.

Electronics are over rated.

Get an Aspirin tablet and use it to hold a wooden clothes pin open. You have thumbtacks pushed into the pin jaws with wire underneath the heads of these thumbtacks.

When the water hits the Aspirin, it crumbles and allows contact between the thumbtack heads. Think of it as a headache free, one shot relay. For the cost of a large bottle of pills, you have a hundred or two resets available. None of that ionized/distilled water quandary to trouble you.

mike

Mousetrap?

Not unless you had 300 ohm mice.

Yes, the danger with water is due to the SIZE of the conductor! Even a 18 Mohm conductor that is 6 inches in diameter can carry a lot of current, and has low effective resistance. So, if you are in water, or standing in it, it can handle plenty of dangerous current. Especially since your body has a lot lower resistance than the water around you, if it is fresh.

Charlie

If your body isn't fresh, you need to spend more time in the water. ;-)

Now attach each wire to a thumbtack, press each one into the jaws of a clothespin and put an aspirin tablet between them.

When water hits the aspirin, it dissolves. The clothespin spring closes the contacts and the alarm goes off.

quoted text -

I saw this done with a wooden spring-type clothespin with a couple of screws in the jaws - somebody recommended putting the aspirin tablet sideways between the notched parts of the clothespin so that the aspirin stuff wouldn't get between the contacts.

Cheers! Rich

Put the aspirin sideways between the notches in the jaws - that way the aspirin stuff doesn't foul the contacts. ;-)

Cheers! Rich

source.

"a 1-10M resistor"

means.. ANY resistor value between 1M to 10M? a variable resistor? just one 10M resistor?

thx.

source.

Depends on the situation, especially on the area of the electrodes. You want normal humidity, minor amounts of acidic decay products, etc. not to allow triggering. You also want actual water to turn the FET on hard.

As long as it works with a range of 3-10x on both sides, it'll probably be pretty robust.

Cheers

Phil Hobbs

source.

I'm curious why nobody (unless I missed it) suggests AC resistance measurement for water level? Got memories of doing something like that with CMOS a long time ago, simple oscillator with capacitor isolated probe to wet spot, detect when signal disappears due to wet spot being wet, conducting signal to ground. Possibly similar to the 'modern' cap sensing buttons?

The series cap was such value where accidental contact with 240V 50Hz wouldn't kill anything.

Grant.

source.

The reason I didn't suggest it is that the OP's circuits knowledge appears to be pretty rudimentary. One FET and one resistor seemed about his speed, and should work fine. (To the OP: We all had to start sometime! Trying stuff and figuring out why it works or doesn't work is a good approach a lot of the time.)

Cheers

Phil Hobbs

Thinking about it, my first choice would probably be to use an (insulated) capacitive probe driven by the VCO of a CD4046 via a resistor, with the two inputs of the XOR phase detector connected to the two ends of the resistor. The output would rise from near 0 to near Vcc/2 when the probe got wet, and if the amplitude at the probe end got too low to switch that end of the PD, it would still be near VCC/2, which would make it more forgiving. Using that to drive a low-threshold FET to turn on the alarm would probably work pretty well.

The simple FET+resistor solution should have a lot lower battery drain under normal conditions.

Cheers

Phil Hobbs