Anyone know a simple (online) resource for impedance/phase characteristics?

[earthtonesaudio]

I'm looking for something that shows (with pretty pictures ideally) basic transistor amplifier configurations and their outputs... such as, whether the amplifier has a gain of greater than one, whether it's inverting or not, and impedance characteristics.  Something with a picture of the circuit, with a sine wave at the input and the inverted (or not), amplified (or not) sine wave at the output.

Also if there's a good site that has the same sort of info for active filters, that would be great too. 

Basically I just want phase and impedance info, but from a "fundamentals" perspective.


Thanks,
Alex

[R.G.]

It seems like a simple question, right?

For transistors, it is.

There are only three single-transistor BJT connections. The differences are in what is grounded and where you take the signal out.

Imagine a BJT in the standard four-resistor circuit: emitter resistor, collector resistor, and base tied to a two-resistor divider across the power supply. That establishes the DC conditions on the transistor by the values of the resistors. The emitter and collector resistors can vary between a short circuit (zero ohms) and quite large values. The base resistor divider MUST present the base with a voltage about 0.5V higher than the emitter. Nothing works otherwise.

But let's assume that we have these values set up well, and further than the collector and emitter resistors are each 10K, and that the base voltage is 3.5V while the power supply is 9V. That gives 3V across the collector and emitter resistors and 3V from collector to emitter INDEPENDENT OF THE ACTUAL TRANSISTOR as long as its hfe is over about 100.

We can now put signal in by connecting a capacitor to either the base or the emitter. We can take signal out by connecting a capacitor to either the emitter or collector.

If we put signal into the base, we can take signal out on the emitter or collector. If we take it out the emitter, then the gain is always just a hair less than unity. This is an emitter follower, and the value of the collector resistor actually matters almost not at all, as long as it's small enough not to interfere. The impedance is ... small... ; how small? There are equations to calculate it, but in most cases it does not matter enough to calculate it closely. Under 100 ohms is common. If you need to answer "how small" to a greater precision than that, it's outside the bounds of your question as stated.

If we take the signal out on the collector, the signal is inverted, and the output impedance at audio frequencies is the value of the collector resistor itself. The collector looks like a current source and is effectively too large an impedance to get into the calculations much. That's not exact, but it's so close that again if you need to know it to greater precision than that, it's outside the bounds of your question.

In the case of signal from the collector, the gain from base to collector is determined by the emitter and collector resistors. We can make the emitter resistor a pot and put a big capacitor on the wiper so we can bypass a variable amount of the emitter resistor. This lets us change the effective value of the emitter resistor without rebiasing for DC conditions. If the unbypassed portion of the emitter resistor is equal to the collector resistor, then the gain at the collector is -1, the gain at the emitter is +1 (as above) and we have a phase splitter. The output impedances of the two output s is different, but the voltages out are equal.

As we turn the pot and slide the wiper over more and more of the emitter resistor, the gain to the collector goes up. The gain is approximately Rc/Rex where Rex is the unbypassed part of the emitter resistor. But what happens when the wiper gets to the top of the emitter resistor?

There is an unseen resistor inside the transistor, the Shockley resistance. This resistor is a function of the emitter current, and is about Ie/26mV. This resistor cannot be bypassed, and it turns out that it determines the voltage gain in cases where the effective emitter resistor is zero. So the biggest gain you can get is Rc/Rex. It is a bad idea to rely on Rex unless you just want big gain and can't be bothered to say exactly what gain.

If we connect the input to the collector, nothing happens. The collector is too high an impedance to let signal in.

We ...can... put signal into the emitter, too. The base cannot put signal out, but the collector does. We put a capacitor between the base and ground, and insert signal current into the emitter through a capacitor. The impedance looking into the emitter is quite low, as it's also trying to reproduce the base voltage at low impedance at the same time, so it's the same under-100-ohms. So we shove *current* into the willingly-accepting emitter. The same current must appear at the high impedance collector, so the current you put in appears across the collector resistor. The gain is more of a transconductance: you put in current, take out voltage. The current is whatever goes into the emitter, the voltage is the current times the collector resistor. The collector impedance is again the collector resistor. The output can have quite high voltage gain.

Why would anyone ever do a common base (that's what this is) connection? it's for isolating the input side from the output side. It's big in RF circles. It's also used to "protect" low voltage circuits from high voltage things. Music Man used a power MOSFET to cascode-protect 20V JFETs from the 300V power supply in their amps.

No pretty pictures, but it's pretty accurate.

[R.G.]

Part 2: Active filters.

All active filters that I've seen require a low impedance source driving them. Their output impedance tends to vary all over the place.

If you mess with active filters in any form, go buy a copy of Don Lancaster's Active Filter Cookbook. It's the go-to book even for EE professionals.

[earthtonesaudio]

Thanks R.G.  I appreciate your detailed replies!  I still wouldn't mind the pretty pictures though...

[GibsonGM]

How about finding a suitable simulation program, 'building' each of the 3 basic circuits, and play with the R, C and input signal values?  Most of the programs have an Oscope function, and you can watch what happens (in the ideal world) in real time as you vary parameters.   Many will even do the graphs and charts, such as parameter, Monte Carlo, frequency info (Bode plots).   That's how I 'burned in' how BJT's operate...

[alanlan]

Easiest way in my opinion is to download the student version of PSpice (v9.1) and have a play.  It's not very difficult to do what you want, but I haven't seen anything on-line which hasn't been tied to a particular vendor's products.

You can download it from http://www.engr.uky.edu/~cathey/pspice061301.html

If you need any help driving it and getting set-up please ask.

[earthtonesaudio]

I found something that is pretty close to what I was looking for:
http://www.play-hookey.com/ac_theory/filter_basics.html

Sort of like an online textbook.


Thanks for the suggestions everyone, I will try out one the PSpice student version as long as it's free. 

[michal_k]

+1 for spice. if you use mac try www.macspice.com it's free, powerful and has a good tutorials/documentation.

to be honest i avoided it at first cause i thought software like NI's workbench let's me draw circuits faster and is generally speaking easier to use but after simulating a few circuits i found spice much more versatile and powerfull. it pays to spend a few hours figuring out how to write for spice.

if you needed any help I can try to answer your questions, I'm not a spice-master but I'm sure we could figure something together.
also, if you decide to use spice let me know which version you use(PSPICE or macspice). if it's macspice I could paste few ready to tweak OE and OC circuits.

m.

[earthtonesaudio]

+1 for spice. if you use mac try www.macspice.com it's free, powerful and has a good tutorials/documentation.

to be honest i avoided it at first cause i thought software like NI's workbench let's me draw circuits faster and is generally speaking easier to use but after simulating a few circuits i found spice much more versatile and powerfull. it pays to spend a few hours figuring out how to write for spice.

if you needed any help I can try to answer your questions, I'm not a spice-master but I'm sure we could figure something together.
also, if you decide to use spice let me know which version you use(PSPICE or macspice). if it's macspice I could paste few ready to tweak OE and OC circuits.

m.

I have some sort of spice software for Mac already, but I have been putting off learning it.  I will checkout macspice for sure.  I'll give the tutorials a look before I bombard you with the tons of questions I'm sure I'll end up having... Thanks!