can't find good complementary transistors

[janneman]

Originally posted by sawreyrw
Tim_x,

You must be the guy who posted those links. That was a nice gesture; I think keantoken looked at them, but obviously, he didn't spend much time trying to understand them. keantoken may be smart enough to figure out how to bias transistors, but he clearly hasn't demonstrated it yet. keantoken, I would like to see you show a properly biased common emitter stage. Make it simple, one transistor with a current gain of 100, 3 resistors Rb, Rc, and Re with a couple of supplies; Vcc and Vbb. If he can't do this, he's lost.

Rick

I think keantoken is just having fun.....

Jan Didden

[poobah]

Kean,

Most engineers would tell you not to learn transistors on a simulator.

Do you have some transistors you can experiment with... resistors, batteries, and a voltmeter?

I think, like sawrey, that you need to start with some basic circuits and learn those first. A complimentary output amp is probably not the first place to start. The common emitter and the emitter follower are the most important circuits you must learn first... then a differentai pair.

I have 1000's of transistors... maybe not too many PNPs. I could maill you some...free! Enough so that you could blow some up... you will!

We are interested in helping you learn... some people are not responding to your posts because your circuits are very strange and they don't wish to be rude. I have less problem with that I 'spose.

[keantoken]

I know how to bias a transistor and I don't, I guess... I know how to get the voltage at the right level but I don't know how to adjust the gain. I would have read up on this except for the fact that I have lost my engineering book... I want to learn! I want to make my own amp like symasym and make my own ham radio station! I want to build circuits for other people and understand people when they throw long terms around and explore possibilities of tweaking their designs! I read those links, tim_x, and understood them but I wanted to do something less major and much simpler so that I knew what I was doing when I tried to build one seriously! Usually you can just tell me what values will make it work and I'll get the idea, but when I'm left to figure it out the first time I try it, things don't work... I tried using the two-battery method on the common-emitter, but I just couldn't get things to work, But I tried the standard common-emitter design and I was able to do things somewhat without errors... maybe. This is my circuit in LTSpice/SWCADIII format:

[keantoken]

What an awesome offer, poobah! I will email you my address and assume that you won't contact any third-party organizations. I suppose that I could break out my old breadboard kit and experiment. I happen to have an oscilloscope, the tektronics 561B, so if I could make an impedance matching circuit so that I could use my computer output as a test source, that would work. I will try to understand these concepts. here is the circuit for all you MS Paint users:

[sawreyrw]

Do a .op on this circuit and then tell us you know how to bias a transistor. :-)

For starters.

Delete R3. You don't need a resistor in series with your supply. It just wastes power.

Remove R5; it's worthless.

Connect the upper lead of R2 to the battery, not the collector.

Add a resistor to the collector circuit.

A single stage common emitter circuit won't be able to drive much signal into 100 ohms. Delete C1 and the 100 ohm resistor in series with it.

[keantoken]

Thanks guys! I think I can truly say that I know how to bias a transistor now! Just in case, here's my biasing technique for a common-emitter circuit:

Take a resistor that is around ten times the supply voltage and connect it between +V and the collector of the transistor, call it R1. Then take one of the same value and connect it between -V and the emitter, call this R2. Next, take a capacitor and connect it in parallel with R2, let this be around 4.7uF-10uF and call it C1.
Now connect a resistor between +V and the base of the transistor, let it be 10k-930k (around 86K for a supply voltage around 10V)and call it R3. Now take a resistor and connect it between the base and -V, the value should be the resistance of R3 divided by your supply voltage and call it R4. Now, connect the power source and measure the voltage across the base of the transistor and -V, it should read around 800mV. Now take R4 and change it to the next lower value, and if the voltage is above 710mV, lower it again. If this time it is lower than 690mV, move it to the last value and start lowering R2 until it is within 690mV-710mV. You can achieve a good bias by fine-tuning with R2. If the base current is too low, repeat the steps with the next lower value of R3. If the current is too high, repeat the steps with the next higher value of R3.

Is this a good technique? Will it work? Thanks everybody!


Here is the revised circuit:

[Tim__x]

Your newest schematic is missing the collector resistor, but you mentioned it in your post, so I'm sure it was just a drafting error.

The voltage across R3 determines the bias current (V/R=I).
Let's say we want to use a 10v single supply with 1k collector and emitter resistors (R1 and R2, I'll stick with the names you were using).
For maximum voltage swing we want to bias the output at around V+*0.75, that means we need 10v-7.5v (2.5v) across R1, 2.5v/1k=2.5ma.
We need to bias the transistor such that it holds it's emitter 2.5v above V- so that 2.5ma will flow though R2.
So we add one Vbe ~0.7v, we need to hold the base at ~0.7v+2.5v=3.2v.
10v/3.2v=3.125 (not volts but a ratio, the V divided out) we set R3 to 10k (ten times R1 as a rule of thumb) and R4 such that (R3+R4)/R3=3.125, to find that we take R3*3.125-R3=21.25k.
C1 isn't absolutely necessary, it simply serves to raise gain without upsetting biasing by reducing the impedance seen by the emitter at audio frequencies.

P.S. Without C1 the gain is R1/R2+1. When the input goes up 1v, the voltage across R2 goes up one volt, so the current through R2 goes up by 1v/R2, so the current through R1 goes up by the same amount and the voltage across R1 goes up by 1v/R2*R1.

[poobah]

OK,

This is going to be a busy week, but I'll see what I have laying around. I know I have plenty of NPN and N-channel stuff I can send you... gimme a few days.

Tell us, in general, what parts you have already.

As far as biasing a transistor is concerned... as gently as possible... you have much to learn. Remember we are ALL still learning... this is engineering... it NEVER stops.

Follow along with Tim_x and sawrey, and put the amp design on the back burner for now. Do you know Ohm's Law inside and out?

Now, that amp that sreten gave you would be a great project for the simulator.

[keantoken]

ok, I understand your method, but do you think mine will suffice?
I am not that mathematical, but I can be even more mathematical than that when I want to... I don't like having to use equations in everything I do, so when I can, I opt for the more methodical methods.

[poobah]

Get used to the math... learn to love it. Throw some on the floor and roll around in it... the ONLY way.