I'm beginning in electronics and have found a moisture sensor that appears to be cheaper but similar to this one

http://www.redcircuits.com/Page18.htm

Which was suggested in this thread:

http://forum.allaboutcircuits.com/showthread.php?t=3645&highlight=moisture+sensor

The sensor circuit can be found here:

http://nurne.blogspot.com/2008/01/water-me-man.html

In the subsequent blog entries someone suggested that instead of a 2n2222 they use a MOSFET tied to a high ohm pull down resistor:

Great concept. Love the idea of a dirt-cheap, dirt-sensing circuit. As someone who could can plead guilty to involuntary plant-slaughter, I can see the utility.

Unfortunately, the circuit as drawn is going to consume a little over 7mA in the idle "off" state, which is bound to be a battery-killer. Dead battery could mean dead plant. Consider using a p-channel MOSFET with a multi-megaohm pull-down resistor. That will take close to zero current (pA) when off and still keep the part count very low. You'll have to pay a little more for the MOSFET than the 2N2222, but what's the cost of a new plant nowadays?

My question is assuming he is looking to replace the 2n2222 with the mosfet, where would I place the mega-ohm resistor for the pull-down?

I've sent the blogger an email asking for clarification based on that reply but haven't received a response so now I turn to you all, the giant pulsing brain of the interweb. This circuit was interesting to me because of the apparent cheap cost and with the MOSFET the long lasting nature of the battery (especially when tied to a blinking led).

P.S. I found that when I measured the resistance of freshly watered soil, I still got resistances of 6-8k ohms, in his original design it appears to me that it would fire when the resistance got above 1k - which didn't seem right from what I was measuring. If anyone else finds this interesting I would be greatful for an outside opinion as to the general resistances measured for your various flora. Thanks!

 

RED Circuits made a Plants Watering Watcher version 1 circuit that didn't dim like it was supposed to. So I fixed it by adding a capacitor so the dimming works with PWM.

But it used an ordinary Cmos IC that has a minimum operating voltage of 3.0V and he used a 3V battery that drops to 2.0V. So I used a 74HC IC that has a minimum operating voltage of 2.0V.
His new version still uses an ordinary Cmos IC. It doesn't dim.

I wanted my battery (two AAA cells instead of the larger AA cells) to last a long time so I made the LED blink to save power. RED Circuits copied my idea.

So my LED is off when the soil is damp and the battery current is very low.
When the soil begins to dry then the LED blinks dimly.
When the soil is drier then the LED blinks brighter.
When the soil is very dry then the LED blinks very brightly.
Here is my project: http://www.electronics-lab.com/projects/science/018/index.html

My stripboard is the same size as the two AAA cells battery holder under it. It has a daughter board on top.

I think my tap-watered soil measures 47 ohms when damp and about 50k ohms when dry.

 

Hi... rather than starting a new thread, I thought I'd try a reply to this one first. I'm interested in finding the project you posted above, Audioguru, but the link is broken - is it still available somewhere online? I'm working on a project that interfaces a moisture detector circuit with an HM-11 Bluetooth module, and I'd like to "invert" the output of the moisture sensing circuit (above) so that the output signal/voltage increases with an increase in moisture between the probes (rather than having the output increase with decreasing moisture between the probes). Also, I'd like to know how easy it would be to decrease the measurement/output rate to, say, once a minute. The Bluetooth module will run on 3V too, and this is the only circuit (above) I've found so far that is low voltage.
My background is mechanical engineering, and while I can read a schematic and assemble electronics, design-wise I'm hopeless!