Moisture Sensor

[Novice]

New to this so here goes - I have breadboarded a moisture sensor which is very simple and works as I want it to... I have designed a PCB circuit which doesn't work at all - I suspectr it's the design rather than the components or the soldering which I have checked (I think) - any suggestions? Design Attached.

[daveshah]

What type of transistor are you using?

[SeanB]

The 1k resistor should be connected not to the 1M resistor but to the switch, so that the reduced resistance of the probe switches the transistor on. At present it will light the LED always.

[JOERGG]

The 1k resistor should be connected not to the 1M resistor but to the switch, so that the reduced resistance of the probe switches the transistor on. At present it will light the LED always.

Yes, but has the switch not to be moved between minus and emitter also? Turning the switch on, at its current position, will shorten the circuit. Correct me if i am wrong.

[Novice]

Thanks for your input
the transistor is a 108  - but the Led  should be on when the moisture sensor is dry most of the time in this situation when it is wet it goes off which when it's in a plant pot most of the time - The 1K resistor was put in to protect the base of the transistor. Like i said the breadboarded circuit worked really well and was quite sensitive to moisture suprisingly.
I'll look at the switch - where do you think it should go?

[Novice]

You were right about the switch I just breadboarded it and the switch between the emitter and the minus works on the bread board - I tried it with the 1K resistor between the emmitter and the switch but it didn't work - I still think I need that at the base. One thing I did think of was - as i'm not getting anypower accross the 1M would the power take the easier route through the led as I am getting power here  but it's not going through the transistor to the negative rail so doesn't come on.

[JOERGG]

My intention to move the switch was to let the circuit be without current when the switch is turned off. You can do this by putting the swich in series with the plus or the minus rail directly. If it is an on-off switch. I supposed the diode to light when sensor is wet, now i know that i was wrong and have no solution so far.

[Novice]

I've added a sort of diagram of the bread board if that helps - ive been trying to get the PCB circuit to work for weeks without success. any other suggestions?

[SeanB]

Plant watering detector..... Your original circuit will work, but you will need to use 2 stainless steel probes, not a copper board that will corrode very easily and turn into an insulator. Get 2 stainless steel knitting needles ( very cheap at the OHL mall) and crimp the wires to them or use a 2 piece 5A screw terminal block, as they will not solder if they are really stainless steel. As you are likely using either 2AA cells or a coin cell I would recommend using a 2M2 or 3M3 resistor for the positive connection so as to reduce current flow through the probes, or use a push button switch to only connect power when the button is pressed to increase both battery and probe life. Using a second transistor in a darlington configuration will also make it more sensitive to dry soil, though this is not going to be very precise in any case.

[JOERGG]

Have build up the circuit myself and must say it works fine on the breadboard. Make shure that you did not confuse the order of the transistor terminals. If in your first drawing this is the solder side the BC 107 will match when placed on the other side of the board. I used 9V Voltage and as an sensor i took a decade resistor box. The light begin to come slowly at 70K Ohm and lits full at 200K Ohm. The current is max 1,5 mA.

[Novice]

Thank you for taking the time to breadboard the circuit - I did consider using a darlington pair but i wanted it to be simple and because the circuit worked didn't extend it. I like the idea of using the knitting needles and the terminal block brilliant idea I will try that.
I have checked all the components in the circuit.. the transister is the right way round as is the LED. I'll look at the sensor it's self to see if I have a short accross the copper lines... haven't found one doesn't look like any dry joints. the LED lit up when I accidently shorted the circuit using side cutters the flat bit of the side cutters touched the negative rail near the switch!!! my eyes light up as well cos i thought it had worked.... then hit the floor when i realised what it was. 
Sorry to persist but any other suggestions for the circuit or should I pitch it and start again?
frustrated novice

[Novice]

Just to let you know I got the PCB circuit to work - it was a faulty piece of tracking and the copper on the moisture sensor itself had a short  - one of the fingers was touching the oposite side. Circuit design was fine. Will consider redoing at some time with the knitting needles and the darlington pair thought.
Thanks for all your help - what a brillaint lot you are!
Novice