Pull down resistors

[paulca]

I was playing around with trying to make AND and OR gates with transistors last night and was getting LED flickering and random bi-directional voltage on the transistors.  In an epiphany moment I realised when my switches were off the base pin was floating in the breeze picking up all manor of noise.  I added a 1M resistor to ground and the circuit was silent.

It just so happens I was watching this video and I note that he does not have pull downs on the base pins, but... strangely his LEDs do not appear to flicker and the circuit looks stable and quiet.



What is going on here?  Is he using transistors with a a higher minimum B-E current or is he just lucky?  (My transistors are cheap and nasty chinese 2n2222's).

[danadak]

Depends on transistor leakage and its beta. The leakage can be
multiplied by the beta to produce enough collector current to
turn on transistor. Temperature also a big effect on leakage.


Regards, Dana.

[David Hess]

The Nylon commonly used with solderless breadboards may have enough leakage to go either way depending on layout, humidity, and temperature.

[paulca]

I'm putting it down to dodgy transistors, or I have different kind to the video.  If I wire a floating pin on mine then waving my hand around the board produces flickering LED.  If I grab the base pin with my finger the LED comes fully on, even with a 10K resistor.

Then again I was always the one to light the mains tester the most in science class which raised a few eyebrows.  Yet the touch screen coffee machine in work often ignores me.

Also my YrRobot 5V breadboard supply is dodgy as ____ I noticed tonight that an LED on the board that wasn't even connected to the ground rail by any component would light dimly while I was holding the board.... the PSU power was OFF!  The LED was connected to the 5V rail via a 330R, that took a bit of bewilderment.  When I disconnected it from the wall wart the LED went out.  Something really screwed up happening there.  I must have been grounding the board via the PSU module somehow, but that still doesn't explain the fact the LED wasn't connected to ground on any breadboard rail.

Maybe I'm a walking battery/capacitor and electrons like me or hate me. <shrug>

Can I assume for such transistor circuits to run cleanly one should have a pull down instead of a floating pin?

(I am also replacing my cheap Chinese discovery kit bit by bit with actual branded components)

[csar]

What is the real reason to use a resistor at the base of the BJT transitor?

[paulca]

What is the real reason to use a resistor at the base of the BJT transitor?

Beginner to beginner (I assume)....

For the input signal on the base you don't need to supply very much current.  Milliamps, to work out just how many you need the datasheet, the "gain" across the collector/emitter and the current you want to flow across them in your circuit.  Divide the current you want by the gain, apply that to the base.  You can work out the resistor you need using Ohm's law.

If you don't put a resistor (as I found out) the Base->Emitter current will try and power the circuit beyond the trannie and the trannie will get hot.

The "pull down" I am referring to however is what happens to the Base of the transistor when you disconnect the signal to the base.  With the base disconnected the current/voltage on the base is "undetermined" and subject to interference/RF/noise.  By connecting the base to GND/0V you know that zero current will flow into the transistor.  It should be a pretty high value resistor (1M sort of thing) so that when the signal is connected you aren't "throwing away" a load of current to ground.

With a pull down in place, if your base current needs to be precise, you will need to factor the pull down into the equation to get your base current.

Some of that will be wrong or non-pedantic, so I'd be grateful for corrections.

[David Hess]

What is the real reason to use a resistor at the base of the BJT transitor?

Usually a resistor is included between the base and emitter to sink leakage current which could otherwise turn the transistor on.  This becomes more important at high temperatures.

A resistor between the base and emitter also serves to help remove charge decreasing storage time and rise time.

A high value resistor in series with the base or emitter limits base to emitter current when the transistor is driven by a low impedance source.

A low value resistor in series with the base may be used to prevent parasitic oscillation do to a capacitive load reflecting through the transistor to become a negative resistance at the base.

Integrated injection logic can be implemented with no resistors at all.  The example there shows a rather poor PNP current source with one resistor but a superior integrated version could be implemented without any resistors.  A discrete version of this logic would require low value resistors in series with every emitter not to limit current but to allow a large variation in Vbe from using unmatched transistors.