hi there
i have made this as shown on this site:
http://hackaweek.com/hacks/?p=380
all works fine but i would like to reverse the operation of the LED
when voltage is detected LED is low
What would be a good way of doing this?
Thanks Joe
Or wire the last transistor with the resistor between the collector and the supply ( just like the other two ) and then wire the LED between the collector and the emitter.
and then wire the LED between the collector and the emitter.
?
Are you sure you meant to say that ?
Well anode to collector and cathode to the emitter, that is what I meant.
right ok, should i be fine using the same 220R or should i go for something more like 2K?
should i be fine using the same 220R
Yes.
You may need to add a "test" pushbutton (or power switch) to prevent the constant current draw when not being used.
the LED does not go out, it duels a bit but that is it, i did increase the 220R but made no difference?
Then put a normal diode in seriese with the LED.
Joes:
the LED does not go out, it duels a bit but that is it, i did increase the 220R but made no difference?
Joes:
the LED does not go out, it duels a bit but that is it, i did increase the 220R but made no difference?
This circuit detects mains AC (detects in the radio sense, rectifies and amplifies), so that the current
through the 220R load is pulsed, which the size and length of the pulses increasing as the copper
strip picks up more stray field. This means that most of the time the output transistor is off, so
that simply complementing the output won't work well.
You would need to low-pass filter the output voltage, then use a comparator to threshold the signal,
then drive an LED to get the inverted sense (light only when little input, no light when strong input).
[ Its probably possible to add low pass filtering earlier on in the circuit by adding a capacitor to
one of the transistor bases, that might help ]
Then put a normal diode in seriese with the LED.
Seemed to make no difference 
there must be a way you can do this with keeping a low component count?
thanks Joe
Tried the capacitor yet? This is a very high gain circuit (gain around 10 million), so it picks up mains
interference from any nearby object, which means it is pulsing - you need dc to be able to invert
the sense of the LED.
I think the output could be inverted like this, but that would mean constant drain on the battery when no electric field is near the copper strip ... so a "test" switch might be needed.
Also note that its much more difficult to perceive impulse activity on an LED that is normally ON than an LED that is normally OFF. Using a capacitor as MarkT has suggested could resolve this.
Please confirm, for the record,
you did this...
and "no joy", right?
BTW, with 9V, I'd go with >= 470ohms, not 220. [He used a rechargeable batt., "7V"]
Please confirm what happens when you short the copper strip to your circuit GND.
right my circuit looks like this:
i did try your theory (dlloyd) but led just stayed on.
The 220R i have increased it but it did not seem to help.
i have been using a voltage from 7 to 24v from a power supply.
shorting the copper strip to the circuit GND made no difference but if i short it to vcc (through a resistor) the LED go's out.
This is what i am seeing over the LED when voltage is present:
so are we looking at a capacitor now?
so are we looking at a capacitor now?
I would try 100µF across the LED ... this will give a time constant of 22ms.
shorting the copper strip to the circuit GND made no difference but if i short it to vcc (through a resistor) the LED go's out.
According to the schematic, shorting the strip to Vcc should blow out the base of the first transistor because there is no base resistor.
This is what i am seeing over the LED when voltage is present:
You must be in Europe. That's 50 Hz.
The led is always going to be dim when driven by a 50 Hz signal.
Is there any waveform with nothing touching the copper strip ?