Single transistor oscillating?

[bitshift]

Hey Guys,

So I'm trying to learn more about transistors. I've breadboarded the common collector configuration using a jellybean 2N3904 NPN. My setup is as follows:

• 4V from LM317 into the base
• 12V from SMPS wallwart into the collector
• 100E resistor from the emitter to ground

I'm observing some rather large (up to 3V p-p) oscillations on the collector and the emitter of the transistor. When I put a 1uF capacitor on either the collector or the emitter the signal becomes smooth as a babies bottom.

I've heard that transistors and op amps can start to oscillate and I think this is what I'm observing. Unfortunately I haven't had too much luck in finding anything explaining this on google. I would appreciate if you guys could point me in the right direction.

I've attached a picture of the setup and screenshots of the measurements on the scope.

[TimFox]

What is the purpose of your test circuit?
The 4 V DC power supply and 100 ohm emitter resistor should form a 33 mA = (4 - 0.7)/100 current source, meaning 33 mA flowing out of the 12V supply into the collector.  Is there anything connected between the collector and the wallwart?
When you see high frequency (like 70 MHz here) parasitic oscillations, the usual solution is to add an appropriate series resistor at one or two of the three transistor leads.  The high-frequency impedance presented by the wallwart and its cable is unknown, and could have a resonance at your frequency.  Try adding 10 to 50 ohms in series with the collector, as close as practicable to the transistor.

[Howardlong]

Adding to TimFox, see w2aew video here: http://youtu.be/81C4IfONt3o

[bitshift]

The purpose was just to measure the Hfe really (can confirm the rough 33mA  ) I'm very much new to electronics and I'm trying to get a feel for transistors.

I measured the oscillation on the collector and the oscillation on the emitter. They seem to be exactly 180 degrees out of phase. I've attached a screenshot of this. I've been reading a bit about linear regulator compensation so the 180 degree out of phase seems to be exactly that problem. Is this likely the same problem?

I tried a 22E and a 56E resistor from the 12V to the collector but the oscillations remain.

Surprisingly enough, I actually have a couple of ferrite beads lying around. I've added one from the collector to the +12V and the signal is perfectly clean. I've attached a picture of this.

So you guys seem to know whats going on here. Do you know of any resources that could teach me about this kind of stuff?

[Howardlong]

Aside from the first w2aew video (particularly the second half, but the whole video is worth watching), there's another one here http://youtu.be/oZ1Dv2dVGkU

Part of it is due to the layout on a breadboard where controlling the capacitances and inductances can be difficult, and the other is due to the gain of the transistor itself, the base (input in this configuration) picking up the output (collector) that's in anti-phase. So the transistor is acting like a switch, turning on and off, being largely controlled by it's own output, as fast as it can.

That is also why you are seeing anti-phase on the collector and emitter. The transistor is acting like a switch, on and off. When off, you get a large Vce, when on Vce is small.

[KJDS]

One of the first things I learnt about transistors was

If you design an amplifier it will oscillate
If you design an oscillator it won't

[bitshift]

Aside from the first w2aew video (particularly the second half, but the whole video is worth watching), there's another one here http://youtu.be/oZ1Dv2dVGkU

Part of it is due to the layout on a breadboard where controlling the capacitances and inductances can be difficult, and the other is due to the gain of the transistor itself, the base (input in this configuration) picking up the output (collector) that's in anti-phase. So the transistor is acting like a switch, turning on and off, being largely controlled by it's own output, as fast as it can.

That is also why you are seeing anti-phase on the collector and emitter. The transistor is acting like a switch, on and off. When off, you get a large Vce, when on Vce is small.

Really nice videos thanks for sharing 

The switch explanation seems to make sense. I've attached two screenshots from my scope. The scope was set to trigger when I switch the circuit on. The yellow is the collector, the blue is the emitter and the purple is the base. Here is my interpretation of the attached graph:

• The transistor starts in the forward-active state Ve < Vb < Vc
• The collector voltage drops and the transistor transitions to the saturation state Ve < Vb > Vc
• The collector then rises again and the transistor transitions back to the forward-active state Ve < Vb < Vc
• Since Vc is the largest voltage, the transistor seems to be oscillating between the forward-active and cut-off states

Is this interpretation correct? If so, what causes step 4? What causes the emitter voltage to become larger than the base (the start of the oscillation?)? As I understand it, in this configuration Vb = Ve + ~0.6?

[Howardlong]

Three things...

O Can we have the schematic you're using please?
O Try to think in terms of voltages relative to the transistor's emitter;
O Remember that the scope probes have significant reactance at 70MHz and will definitely have some effect of the circuit (I also assume you're on 10X probes, it will be much worse at 1X).

You may find it worthwhile adding a math function or two to measure the voltages relative to the emitter.  It is also almost certain that the probe loading is affecting the circuit, especially at 70MHz, consider the reactance of your ~15pF probes at 70MHz.

[T3sl4co1l]

When you have common collector, and inductance on the emitter, the base has an apparent negative resistance.  Just add LC tank, and you have yourself an oscillator.

Usually around 20 ohms in the base or collector, at frequencies where this is relevant (i.e., 1 to 100 times below fT), dampens the negative resistance.  A ferrite bead does a good job, having a largely resistive characteristic at RF.

Tim

[Hideki]

I tried a 22E and a 56E resistor

E? I have never seen anyone using that before. It's usually just "ohm" or the ohm symbol. In a schematic it's often dropped, so you see nothing but the number for values < 1k (just 22 and 56) or you might see 22R and 56R.

[radix]

E is quite common around here. A local resistor factory used it as a designation for Ohms. I think Philips also did that.