Touch activated torch - increase sensitivity

[Cirkit]

I've been given the attached circuit to improve. The LED should light when a finger is placed across two contacts spaced around 1cm apart. However, the LED doesn't light with everyone's skin.

I have reduced the value of R2 from 150 Ohms to 47 Ohms which has slightly improved the sensitivity.

Is there anything else I can do to make the circuit work with anyone's skin?

Also, the brightness of the LED depends on the amount of pressure/skin resistance. Is there a way to just make the LED light up at full brightness when the contacts are touched and off when your finger is removed?

The circuit is powered from a CR2032 3V battery.

Thanks.

 

[kellys_eye]

Look at the parameters of the switching FET.

You need one with a very low gate threshold voltage.

Search the usual sellers sites using their parametric search options.

 

[Bluejets]

Darlington transistor.

 

[Audioguru]

Some white LEDs need 3.5V to be bright. Not from a 3V battery cell.
Some Mosfets need 10V to fully turn on but your Mosfet has no part number.
If you try that circuit in winter when the air has low humidity then static electricity will destroy the Mosfet that has no protection.

 

[Cirkit]

The LED does light brightly when some people touch the contacts.

The MOSFET has the marking 'A77E' on it. I'm guessing it's this part:

http://p.globalsources.com/IMAGES/PDT/SPEC/255/K1007321255.pdf

Would a darlington work down to lower voltages <3V?

 

[Bluejets]

Darlington is sensitive but, you are stretching it with the applied voltage as Audioguru says.
Did you read his comments?

 

[Cirkit]

Yes I did read his comments. So should I be looking at a transistor with integrated transient absorbers on the gate?

What is the advantage of using a P-channel FET over an N-channel?

 

[Bluejets]

I'd say the advantage would be in favour of the N channel because of availability.
Fact still remains that with some white leds the voltage required is more than you have available let alone any switching element loss, hence "did you read the reply"?
Do you have a problem with adding another battery?

 

[Cirkit]

I'd say the advantage would be in favour of the N channel because of availability.
Fact still remains that with some white leds the voltage required is more than you have available let alone any switching element loss, hence "did you read the reply"?
Do you have a problem with adding another battery?

Yes, the LED used illuminates sufficiently with a 3V CR2032 and a 150R resistor in series.

The LED does light brightly when some people touch the contacts.

 

[Audioguru]

Did you look at the datasheet of the tiny battery? It is designed for a load current of only 0.19mA but can supply 6.8mA pulses for a duration of 2 seconds a few times per day, for a few days.

You are lucky that your LED uses only a few volts, the Mosfet is more sensitive than most, the tiny battery has not gone dead yet and the Mosfet has not been destroyed by static electricity yet.
Increase the sensitivity to people and the chance of damage with static electricity by increasing the 1M resistor value to 10M.

 

[Cirkit]

Is there another way I could switch the LED with touch contacts without risking damage to a MOSFET?

Could I use a low-power Op-Amp designed for battery powered circuits instead?

 

[Audioguru]

Opamps with Jfet or Mosfet inputs are also damaged by static electricity.

 

[BobK]

Changing R2 will not effect the sensitivity. To make it more sensitive, increase R1.

Bob

 

[Cirkit]

Thanks for the suggestions.

is there an alternative circuit/components that I can use that wouldn't be prone to static damage?

 

[Audioguru]

There are probably Mosfets available with zener diodes between the gate and source or you can add the zener diodes yourself.