thebugger

modified 10 years ago

Basic BJT Topologies

201

4723

2d 19:11:07

Almost all kinds of amplifiers are composed from one of these 3 basic preamplifier topologies which offer different features like voltage or current amplification and they have different qualities concerning frequency response distortion rate phase shift etc...

Common emitter amplifier is one of three basic single-stage bipolar-junction-transistor (BJT) amplifier topologies, typically used as a voltage amplifier.In this circuit the base terminal of the transistor serves as the input, the collector is the output, and the emitter is common to both (for example, it may be tied to ground reference or a power supply rail), hence its name. The output is an inverted image of the input.

Common collector amplifier (also known as an emitter follower) is one of three basic single-stage bipolar junction transistor (BJT) amplifier topologies, typically used as a voltage buffer (current amplifier) it has a voltage amplification less or equal than 1 because the emitter voltage is less that unity (max voltage minus base voltage drop). Thanks to Secuture i recently learned that after a few modifications this kind of amp ceases to subdue to emitter degeneration and the negative feedbacks and actually could be used as a voltage amplifier too.

In electronics, a common base (also known as grounded-base) amplifier is one of three basic single-stage bipolar junction transistor (BJT) amplifier topologies, typically used as a current buffer or voltage amplifier (both voltage and current amplification). This arrangement is not very common in low-frequency circuits, where it is usually employed for amplifiers that require an unusually low input impedance, for example to act as a preamplifier for moving-coil microphones.

published 10 years ago

thebugger

10 years ago

I dont know if ive designed the first (common emitter) topology right but its non responsive to frequencies above 1Khz. This almost has no use in its curent state because even for voice frequencies up to 3,4Khz are used.

Secuture

10 years ago

I notice a frequent use of far too low collector resistance in common emiter. That's why it works bad in higher frequencies. I never go down below 2kohms and if higher power is need it's always better to use combo wit voltage follower. Common emiter do not like to amplify power very very much unles target load is in collector itself in series with transistor instead of resistor or u use transformer then u can use it to power amplification. Most high frequency operation fails due to over loaded base when someone want to drive output swing almost to cut off level and load is heavy requiring high base currents. More bias resistance used is frequently too large like in your example. Uoperate in close to peramnet saturation and ammount of current is high but as we can push currents through low ohm input base a after push a pul is made but this time we can only load cap by chrges suplied by bias resistors cuz base is unidirectional. And if some load point is crossed not all push charges will be equally compensated by pull charges and make our cap at input polarize in one direction and this is equal to smaller capacitance seen by signal source . THis technique is used to make fixed capacior to act as variable one by over polarize it by some current source working in synchrony with supply during operation. ANd this what happens here if bias resistance is too high. Rule of thumb says that bias resistors should conduct currents of ten times higher value than expected average value of base current. It not must be ten times more as max and u can set it to 15x more also. Just nt use some insane rattios like a 100x more cuz this current flush signal out like tsunami wave.

rbrtkurtz

10 years ago

I don't understand what the heck is going on here. I deleted everything except the common emitter. Then I deleted the caps, and the AC source, and began setting up quiescent voltages and currents based on a 12V supply voltage, and a 10K ohm load. Then I did all the math. I pretty much threw standard values to the wind, and put the exact values from my calculations into the sim. I put 10mA at the emitter with a 100 ohm resistor. Then I setup the voltage divider to give me 1.8V at the base. Since Vbe is 0.8V, that leaves 1V at the emitter, at 10mA. That's pretty standard. I only wanted about 1mA through the divider, so I went with 1K and 5.41K to get 1.8V @ roughly 1-2mA. So far, so good. Finally, I set the quiescent voltage at the output to 6V (centered between 12V and 0V, so the signal can swing fully from rail-to-rail) with a 606 ohm resistor. That's a gain of 6. Not bad. Then I did the math for the two coupling caps and the bypass cap. With these, I did round the values slightly, but kept them close to the actual values from the math. I figured for a lowest frequency of interest at 20Hz. So I put 22µFs on the input coupler and the bypass, and a 2.2µF on the output. Then I hooked up an AC input signal of 10KHz at 100mV, that's reasonable, right? Apparently not. The output was slamming against both rails, hard. Curious, with you having mentioned frequency issues, I set the frequency to 100Hz. It was running at a gain of about 12, which still isn't right, but it was no longer clipping. As I played around, I noticed that the amplitude is increasing freakishly as frequency increases. WTF?

thebugger

10 years ago

Weird,right?

pip

10 years ago

It's because you have frequency dependent resistor on the emiter the 4.7uf cap has 3.3R at 10khz and 330R at 100hz in parallel with the 100R resistor so cap has a dominant effect at higher frequency and is less dominant at lower freq. Plus the rolloff at lower frequency imposed by input cap all conspire to lower the gain at low frequency and increase it higher up the band. If you are looking for a wide band amplifier remove the dominant Pole's.

rbrtkurtz

10 years ago

I call BS. A 4.7µF cap in parallel with a 100 ohm resistor has a capacitive reactance of 0.3374 ohms at 100KHz, 3.275 ohms at 10KHz, 25.29 ohms at 1KHz, and 77.2 ohms at 100Hz. That's in parallel with the 100 ohm resistor. On its own, it's .3386 ohms @100KHz, 3.386 ohms @ 10KHz, 33.86 ohms @ 1KHz, 338.6 ohms at 100Hz. The formula for capacitive reactance is Xc = 1/2πfC. The formula for calculating parallel resistance is Rtotal = 1/(1/R1)+(1/R2)+(1/R3)... The higher (ie, faster) the frequency, then less charge can build up between the conductors in the capacitor, therefore there will be LESS opposition to current. It's science.

pip

10 years ago

Increase the input cap and the 4.7uf to 1000uf then set output of the signal gen to 10mv and sweep the frequency from 10khz to 100hz notice how the gain remains pretty constant ? That's because those cap's no longer play a dominant role. That's experience.

pip

10 years ago

No output falls with frequency. Ie gain is reducing. Now repeat rinse and condition with those 2 cap's at 1000uf starting both tests at 10khz the sim settles faster high to low sweep the frequency slowly down gain now remains the same !! Yes these are large cap's and yes you wouldn't normally see them in a well designed amp but this isn't well designed.

rbrtkurtz

10 years ago

Oh wait. I'm dumb. AC gain is like 150+ for this. Derp. That's way too much to be putting in 100mV, with only a 12V supply. I usually have a resistor before the signal hits the voltage divider to step it down. That's my bad, there. pip, you're still very wrong. Big caps will give you better low frequency response. With a 220µF bypass cap, this thing is going to suck balls below 100Hz. But it's fine at high frequencies. Change that to like 470+, and it will be good to go down below 100Hz.

pip

10 years ago

The resistor on the input is bad design. A couple of thens of ohms for other reasons is doable but it does nothing to gain. Gain is high now you reduce input signal a lot if you are looking for that gain of six or anything near. So you amplify noise in the transistor instead plus you have added resistive noise in the worst possible place input to a high gain amplifier. A well designed amp could care less about low or high frequencies it will do a good job with them all . The problem with big cap's is the need for electrolytics and everything that's inside them is not at it's best for higher frequency but even this is doable with an amp that is designed well for the task.

pip

10 years ago

What I'm saying is the gain is to high and the input cap and the emiter cap roll off the lower frequencies so high frequency slams the into the rails but lower frequencies because of the rolloff don't so the amp's design is wrong and any fudge to correct this with a resistor in the input is making this worse but maybe you can get away with this in some circuits because the increased noise is an acceptable tradeoff in some circumstances. And if you have to use huge cap's to straiten the gain the amps design has bigger problems than what comes out of the cap's. Reduce the gain or use a lower input .

Secuture

10 years ago

Or get rid of caps at all make design a direct coupled amp. Yes there were required high precision in calculating bias resistances cuz any resistor will have influence over whole amp not only single stage but this is worth a try cuz any frequency related problems disapears for audio range.

rbrtkurtz

10 years ago

You can also just not use a bypass cap, then AC gain is the same as DC gain.

thebugger

10 years ago

Wont direct coupling pass DC through the speaker. Maybe an inductive coupling will do best?

sonypl

10 years ago

Nice :) fajne

GomezAddams

5 years ago

Nice summary.

MatthewKessler

1 year ago

Sure

add comment in editor