Do I need a transistor in my moisture sensor circuit?

Hi

I am designing a moisture sensor circuit whereby if there is moisture between the two probes it will complete the circuit and a buzzer will sound. The buzzer requires 4Ma so I am wondering if I can safely send 4Ma through a pan of water or should I use a transistor as a switch?

Thanks :slight_smile:

Do you have any technical information on the sensor?

You may need "more" than a transistor. Are you going to use an Arduino?

Are you sensing water in a container, or is there a chance of something being damp?

Note that pure water doesn't conduct. You need salt or some other impurity before you'll get current.

I just tested some water from my tap in a cup with my multimeter. The readings were jumping-around quite a bit, but the readings were mostly between 300-400K Ohms with the probes about 1/2 to 1 inch apart. (The spacing didn't make that much difference.) With my water and that particular setup, it would require more than 1000V to get 4mA. So, you can't power the buzzer directly through the water.

Thanks for the info, that was a lot of factors that I had not considered. I have tried a plant moisture sensor with the circuit and I am now working on developing a leak sensor without an Arduino.

The circuit will in theory consist of a battery, transistor, switch, two probes and a buzzer. I will be using tap water so this should conduct as it is not distilled? What transistor would work well at these low milliamps and what gain would I be looking at? Would the voltage that I will consequently be inputting into the water be safe so that you don't get a shock if you put your hand in the water?

If you put DC onto probes and then put them into water over a day or two you will get bubbles forming round the electrodes and increase the resistance you measure.
It is not entirley clear what you are doing, is the water getting one electrode, two or four?

I am designing a circuit which when put in a substance it will identify whether it is water by beeping, as air is more resistant than water the electricity will not be able to travel from one electrode to the other so the buzzer will not go off. I am planning on using a design like this: https://hwstartup.files.wordpress.com/2013/09/101.jpg?w=957&h=626 - with two electrodes. The device will only be run for around 30 minutes and then it will not be used for a day. Is the style of measurement I have shown the most effective for detecting water or is there another method?

That chip on the photo looks like an IC rather than a transistor. Have you anymore information on those?

Does this have anything to do with an Arduino?

I used a version of their circuit with an Arduino and I am now looking at making a final Arduino version before I upgrade to a standalone circuit. My main problem is the risk of the two electrodes being connected together by something metal (by accident) and frying the transistor. Is there anything that I can build into the circuit other than a fuse which will protect against this? Obviously I can't us a resistor as the water ends up acting like a resistor.

Wouldn't it be much easier and safer to use a low power signal to detect the presence of water and amplify it to feed your buzzer?

For instance http://www.faludi.com/2006/11/02/moisture-sensor-circuit/

I used a version of their circuit with an Arduino

Which you don’t want to share with us.

My main problem is the risk of the two electrodes being connected together by something metal (by accident) and frying the transistor

Then it is a bad design.

Is there anything that I can build into the circuit other than a fuse which will protect against this?

A fuse is totally useless for protecting a semiconductor device it takes far too long to blow.
Sure there is probably a modification you can make to your circuit but only you know this circuit so only you can make the modification.

To begin with I constructed this circuit, using my Arduino as the MCU - https://i0.wp.com/www.fact4ward.com/wordpress322/wp-content/uploads/2012/06/circuit.gif - however I am now exploring how I would be able to make this circuit more simple for a final product which would obviously not use an Arduino. I have considered using the same circuit with a much less powerful Atmel MCU, however I was trying to determine if it would be possible to accomplish the same result with a circuit that didn’t use an MCU in an attempt to reduce manufacturing cost and complexity.

The circuit I came up with is attached (there would be more transistors in the final version however it is just to give a rough idea), although I quickly realised the flaw with the circuit that if a metal object accidentally connected the two electrodes, the water resistor would be bypassed and the transistors would be fried. Is there some way which I can prevent this in my circuit or another circuit design which accomplishes the same task without a MCU?

Sorry for any confusion

Thanks

To begin with I constructed this circuit, using my Arduino as the MCU - https://i0.wp.com/www.fact4ward.com/wordpress322/wp-content/uploads/2012/06/circuit.gif

OK let's start with this one. A very poor circuit.

  1. There is no protection on the base like you say
  2. It uses an emitter follower configuration which might be fine for an analogue input is not for a digital one.

Your circuit is better but you have nothing in the base, therefore you need a resistor to limit the current if the probes were ever shorted. This would be designed so that when shorted less that the allowable base current flows. Note this transistor is very old and very weak, it only has a 50mA collector current and a gain of 20 and I would not recommend it.

An alternative would be to use a small signal FET like a 2N7000 that would need a 10K pull down on the gate but it would not matter if the supply was connected directly to the gate.

Of the two alternatives, which would you recommend? If I am understandingly correctly, the resistor connected to the probes would reduce the current both when the probes are shorted and if there is water between them. Would this mean that the current from the probes if water was present, would be even lower?

the resistor connected to the probes would reduce the current both when the probes are shorted and if there is water between them.

Yes but this is the base current here so what is the problem? As long as the current with the water is sufficient to turn on the transistor it does not matter. If you use a decent transistor with a gain of say 300 then that would not be an issue.

Of the two alternatives,

Proper transistor.

I am browsing data sheets looking at transistors IB and I am confused about why a transistor can have multiple IB and IC values. Is it somehow related to the max voltage?

The multiple Ic and Ib values are the conditions at which the other parameter is measured. So when testing they set Ib and Ic and then measure the saturation voltage. You will see the saturation voltage changes depending on how the transistor is being driven.

So I am planning on using this transistor - Intelligent Power and Sensing Technologies | onsemi - the Ib is 0.1 Ma and the Ic is 1 Ma. So as I have a 3v battery between the battery and base I will add a 4000 Ohm resistor so that the max current if shorted is 0.75Ma which is below Ic, then if the water is present it will add a maximum of 5000 Ohm resistance which added to the existing 4000 Ohm would result in 0.3Ma which is above Ib.

Between the battery and the Buzzer I will add a 750 Ohm resistor as the buzzer uses 4Ma.

I have based all of my calculations on there being 3V both going to the base and the buzzer, based on the circuit I have drawn (earlier post) is this correct, or is the voltage split, in which case I would need to redo my resistance calculations?

he Ib is 0.1 Ma and the Ic is 1 Ma.

These are only the minimum values in practice you can put much more, it is mA by the way not Ma.

If you have 1mA Ic then you will not have enough to turn on the load (buzzer), how much current does your load take? I would allow for at least twice that in your calculations.
A 4K resistor sounds quite high ( by the way you can't actually get 4K resistors the closest you will get is 3K9 but the values are very none critical.
The data sheet does not mention an Ib max but the On characteristics have a test parameter of Ib = 5mA so I would limit the base short current value to 10mA.

I have found out that the data sheet that I referenced is no longer widely available so I am looking for a transistor with a saturation minimum Ic of less than 4mA (current used by buzzer) and minimum Ib of less than 0.5mA (due to resistance of the water). Currently I am just trawling through datasheets, is there a better method?

My other concern is that: I have based all of my calculations on there being 3V both going to the base and the buzzer, based on the circuit I have drawn (earlier post) is this correct, or is the voltage split, in which case I would need to redo my resistance calculations?

or is the voltage split,

No.

so I am looking for a transistor with a saturation minimum Ic of less than 4mA

??
Not sure I understand that, those are just currents they test with, there is no need for your design to copy this.

I am just trawling through datasheets, is there a better method?

Most major distributors have selection pages.
Transistor selection

I thought that the IC and IB stated were relational pairs, however the fact that they are not makes things easier!

I am going to use this transistor - http://www.farnell.com/datasheets/1496353.pdf - as there is lots of availability as well as small quantity ordering available.

It doesn't reference max and min Ib, however a Ib of 0.5mA and 5mA are referenced so I will use a maximum of 0.5mA to be on the safe side, however I will obviously test this. At 3V if shorted a resistor of 560 Ohm will ensure that I don't breach that max Ib when shorted. When water is present, the max water resistance (5000 Ohm) and the extra resistor (560 Ohm) will result in a current of 0.54 mA flowing to the base which should be more than sufficient. As I need 4 mA flowing from collector to emitter as the buzzer is my only load, I will use a 750 Ohm resistor. When choosing a battery for this circuit I will obviously need to take into consideration the discharge curve to ensure that the battery successfully powers the circuit for a reasonable length of time.

Thanks for your help :slight_smile: