# Weird Transistor Behavior

Discussion in 'The Projects Forum' started by alfonsofernandeztakaki, Aug 21, 2012.

1. ### alfonsofernandeztakaki Thread Starter New Member

Aug 21, 2012
3
0
First I have to say that my knowledge about electronics is based only on a high school class a couple of years ago and the pages I read here about diodes and transistors.

I'm trying to make a computer controlled TV-remote, using an Arduino processor. After connecting the IR LED I realized I didn't have enough power, so I tried to connect it to an AA battery and control the led with a transistor.

My circuit looks as attached.

The initial problem was that the LED is constantly on, shining a little brighter when the Arduino leg is HIGH. Then I connected the transistor base through its resistor to ground (outside the Arduino), creating a power-free base-emitter circuit.

Oddly, the LED is still on. When I completely disconnect the base-emitter circuit, the LED turns off. To the best of my understanding, there shouldn't be current through the base in either of the cases, so the transistor should not permit current through the LED circuit.

Am I using the transistor wrong? Overloading it? Did I just create a source of unlimited free energy?

Thanks for any help.

File size:
9.9 KB
Views:
30
2. ### #12 Expert

Nov 30, 2010
18,076
9,676
Like this...

• ###### IR_remote.jpeg
File size:
10.8 KB
Views:
22
Last edited: Aug 22, 2012
3. ### alfonsofernandeztakaki Thread Starter New Member

Aug 21, 2012
3
0
Thanks for the response. I didn't test it yet because I need to get the resistors, but I have a couple of theoretical questions:

First, how do you calculate the currents in a circuit like this? (I know how to calculate currents in a parallel circuit with resistors, but not how to treat the transistor).

And second, why does my circuit behave the way it does?

4. ### #12 Expert

Nov 30, 2010
18,076
9,676
Why does a 44 milliamp arduino fail to control a 20 milliamp LED?

I don't explain why bad designs don't work, only why good designs do work. I could work it out, but why? So I can use it to make another circuit design fail? I don't think so.

Expecting 5 volts input, the transistor will use up 6/10 of a volt for its base to emitter. That leaves 4.4 volts for the input resistor. You need 1/10 of the LED current in the base of the tansistor to make it switch sharply so E=IR
4.4 = .002 X R
That's 2200 ohms but,
you need a resistor from base to ground to make sure the transistor shuts off.

Make both of them 1800 ohms and 4.4/1800 = .00244444 amps input
.6/1800 = .00033333 amps wasted to ground. Remainder is .00211111amps to switch the transistor on. Sufficient.

The transistor is either on or off. When it's off, the LED doesn't light up. When it's on, the LED does light up. When the transistor is on it uses up 2/10 of a volt. The LED uses up 1.5 volts. There are 1.5 volts leftover. E=IR
1.5 = .02 x R
R=75 ohms, at least.
You used 100 ohms. Good enough. The LED will not smoke.

alfonsofernandeztakaki and absf like this.
5. ### alfonsofernandeztakaki Thread Starter New Member

Aug 21, 2012
3
0
I think understanding our mistakes is much more important than understanding our successes... But a computer simulation of the circuit agrees with what I predicted should happen, so I guess it's all due to an imperfection. I bought an ampermeter to study the transistor and the Arduino pin more closely.

Anyhow, I tested your design and it works. Thanks for the solution and explanation.

6. ### #12 Expert

Nov 30, 2010
18,076
9,676
Correction: .2 volts from collector to emitter of the transistor (when it's on) plus 1.5 volts for the LED leaves 1.3 volts for the resistor.

1.3V/.02 amps = 65 ohms minimum to keep the LED from smoking.

Sorry. I slipped a digit.