Simple BJT Amp Stage

[questron]

http://youtu.be/Y2ELwLrZrEM

You will see that he doesn't use a circuit simulator, he does the design calculations on paper and then builds the circuit to see if it matches the calculated predictions.

Thank you very much Ian! i'll watch it.
I can't do it that way though, I don't have the necessary equipment, although that's surely the most fun way of doing it.

[questron]

Thank you Jay for the great work!
I learned about how to step a sine wave signal as well!

is 10mv, as in my circuit, a reasonable assumption for real applications, radio RF stages for example? 10mv seems to give too high a THD number as your study shows.

[Jay_Diddy_B]

Hi,
I decided to switch from LTspice to hardware to see if I could see the same property.

First I built the circuit on a solderless breadboard. I used a 2N3904 transistor.



If I look the output waveform with a scope, I see a reasonable waveform, this is actually 3% THD.



I then used my HP 35665A Dynamic Signal Analyzer. Here is a screen shot showing the spectrum of the output. You can see the fundamental and the second harmonic.




I then measured the THD versus signal amplitude.
With 1mV peak I measured around 1%:



With 3mV peak I measured around 2%:




With 5mV peak I measured around 3.2%:



You can see that the THD is increasing with the signal amplitude.

This makes sense, because for very small signals the non-linearity is less important.

Regards,

Jay_Diddy_B

[LvW]

Hi questron,
for your information, I like to add some general remarks to the two feedback schemes discussed above.

1.) Resistor between collector and base plus input series resistor: This principle is called "voltage controlled current feedback" and causes a DECREASE in input as well as output impedance of the amplifier stage.

2.) Using a signal relevant emitter resistance (resp. a corresponding Re-Ce combination) is called "current controlled voltage feedback". This feedback principle has the advantage of an increased input impedance. However, the output impedance (from the collector into the transistor) is also increased - however, this is not so relevant because you have to consider Rc in parallel. That means: The total output resistance primarily is determined by Rc (some kohms).   
(Remark: This feedback principle exploits the fact that the BJT is a voltage-controlled device).
------------
In many cases, these considerations are important for selection of a suitable feedback scheme.

[Mark Hennessy]

thank you very much Mark, it helped a whole lot!

i tried 50k, 10k, and 2.2k base resistors. with 2.2k and Rload changed to 100k, THD=0.17%, and the gain is about 13, that's much better. with base resistor being 500ohms, THD=1.1%, and the asymmetrical output waveform is obvious in LTspice, but i get the idea now.

It's partly a factor of the output swing. In many applications, the output is much smaller.

Thank you those who contributed to that point while I was snoozing last night

say in radio RF stages, what's the normal interstage output voltage roughly speaking? just want to have a general idea. why do designer not fully utilize the gain capability of BJT's, is it because of negative feedbacks?

In a radio, the situation is complicated by the interstage coupling transformers - these improve the impedance matching between stages, which reduces losses. They are also tuned to allow only 10.7MHz through, so distortion is less of an issue - these will attenuate 21.4Mhz and 32.1MHz, etc, by a large amount.

Also, don't forget that FM IF amplifiers (at least the last one) are run into saturation to reduce the sensitivity to AM.

At the input to the discriminator coil, it's not uncommon in my experience to see several volts of 10.7MHz.

Rest assured; the designers did all they could to maximise sensitivity, and they generally were very successful. When I restore old transistor radios, I'm always very impressed at the performance achieved from just 5 or 6 BJTs. The challenges were noise and stability (the risk of amplifiers turning into oscillators), not signal level. To understand this, imagine 4 transistor amplifiers connected in series, all with a relatively low gain of 10 to keep the numbers simple, and with an overall output voltage of 5 volts. The input to the last amplifier will be 0.5V. The input to the penultimate amplifier is 50mV. And the input to the antepenultimate amplifier will be 5mV. So as you can see, signal "swing" is only an issue for the very last amplifier in a multistage circuit.

I have not investigated Jay's circuit yet, becasue I know the emitter resistor is going to reduce the gain, then I won't be able to tell whether any THD reduction is due to reduced gain, or circuit configuration, or the BJT itself. I'd like to nail down the BJT first, the gain of a particular circuit configuration the next, and then different circuit configuration the next.

Quite simply, the distortion comes from the exponential transfer characteristic of the BJT, and NFB - however applied - reduces this distortion.

To further lower the distortion, you need more open-loop gain. But this is limited by the gm of the transistor (set by the collector current) and the load resistance. To increase gm, you could double the collector current. But if you've kept the supply voltage the same, this will require a lower value of Rl, so you're back where you started!

Don't forget that the impedance of the following circuit (or 3k9 in your first schematic) is in parallel with Rl, reducing the gain. Everything is against you

To take the performance to the next level, wouldn't it be nice if we had something that passed a current, but had a very high impedance? That, in a nutshell, is why current sources or current mirrors are often used as a collector load for transistor amplifiers. Or, back in the days when BJTs were expensive, we did "bootstrapping" - an extremely elegant technique.

[onlooker]

It's best to watch Shahriar's video tutorial...

If I did not miss something, my comments would be:

The video did not talk about the signal input and output impedances in relation to the voltage gain. If voltage gain were the only goal,  a transformer can do it simpler.

The 1st important effect of the input impedance is the loading down of the signal source. The input impedance is depending on the Bf of the transistor, and espetially so if the design were to have max voltage gain (no negative signal feedback). This Bf dependency is to be avoided in any real design.

Lastly, a minor point, the vedio demoed a 2 stage amp, but titled as a single stage amp.

[questron]

thank you Jay for your great work again, it's really nice to see the real thing at work plus a real frequency spectrum! too bad i don't have fancy equipment like that.

as per Mark's radio example, 10mv input for the last stage of a radio circuit is not out of the ordinary, so probably some sort of formal NFB is a must for a reasonable low THD figure suitable for audio applications, i'm guessing. so let's wait for Mark to get a chance when he could give a lesson on NFB, that'd be great at least for me

[questron]

Hi questron,
for your information, I like to add some general remarks to the two feedback schemes discussed above.

1.) Resistor between collector and base plus input series resistor: This principle is called "voltage controlled current feedback" and causes a DECREASE in input as well as output impedance of the amplifier stage.
...

thank you LvW, alright we are finally in NFB land now!

one question, if BJT's are voltage controlled devices, how does a whatever-controlled-CURRENT-feedback work? doesn't any controlling thing have to use voltage to control a voltage-controlled device?

suppose the given is that a BJT is on already, then we pump currents into its base, that will cause a corresponding change in Ic, could we say that this additional pumped-in current caused the base-emitter P-N junction resistane to drop, as the BJT is behaving as a trans-resistor, which in turn caused the Ic to rise? so at least for this part, with the issue of how a BJT is turned on ignored, the working mechanism is based on current, could we say that?

[questron]

thank you very much Mark, a lot of useful information again!

Also, don't forget that FM IF amplifiers (at least the last one) are run into saturation to reduce the sensitivity to AM.

didn't know that one before, one more pending topic to be explored. i'm mainly interested in audio related topics such as radios and audio power amps, etc., so i guess i'd be interested in learning about how to have control over THD and noise, on top of all necessary basic circuit topics. when you get a chance, please do give a lesson on formal NFB, many thanks in advance!

To further lower the distortion, you need more open-loop gain. But this is limited by the gm of the transistor (set by the collector current) and the load resistance. To increase gm, you could double the collector current. But if you've kept the supply voltage the same, this will require a lower value of Rl, so you're back where you started!

i see, thank you! no wonder people talk about high supply rails with discretes, they say high supply voltages is the whole point of using discretes. so increasing the supply voltage is certainly one of the legitimate ways of lowering THD, yeah?

To take the performance to the next level, wouldn't it be nice if we had something that passed a current, but had a very high impedance? That, in a nutshell, is why current sources or current mirrors are often used as a collector load for transistor amplifiers. Or, back in the days when BJTs were expensive, we did "bootstrapping" - an extremely elegant technique.

wow, bootstrapping, current-mirror, these are certainly the topics i'd like to learn about. i've just finished my reading on current-mirror, now i can apply that to lower THD, just great!

all great topics, very interesting to me. could we talk about each one by one? i'd like to hear formal NFB first, but it's all up to you, whichever topic, whenever you feel like it. thank you very much!

[LvW]

thank you LvW, alright we are finally in NFB land now!
one question, if BJT's are voltage controlled devices, how does a whatever-controlled-CURRENT-feedback work? doesn't any controlling thing have to use voltage to control a voltage-controlled device?
suppose the given is that a BJT is on already, then we pump currents into its base, that will cause a corresponding change in Ic, could we say that this additional pumped-in current caused the base-emitter P-N junction resistane to drop, as the BJT is behaving as a trans-resistor, which in turn caused the Ic to rise? so at least for this part, with the issue of how a BJT is turned on ignored, the working mechanism is based on current, could we say that?

No - we never "pump" a current into the base. Base current is always a current than cannot be avoided, but in any case - as mentioned by you - it is the base-emitter voltage that controls Ic. However - how is this voltage developped?
Remember the classic inverting opamp circuit: Two currents meet at a common node (inverting terminal) - the current driven by the input voltage and the current driven by the output voltage (to be computed using the superposition principle). This combined current produces a voltage at the inverting terminal which causes a correponding output voltage.

And exactly the same applies in our case to the signal currents (only signals because of the capacitor): The collector voltage drives a current through both resistors and the signal source which adds to the signal input current (in the reverse direction). Both currents generate a signal voltage at the base which is amplified.
(In this description, I intentionally have neglected the base current which also contributes up to a certain degree to the base signal voltage). 
The "working mechanism" of a BJT is always based on physical laws and cannot depend on any external circuitry. 

These considerations do even apply for the DC current into the base. It is common to say that such a single resitor Rb between collector and base "injects" a DC current Ib into the base. But this is not true! It is the collector VOLTAGE which is connected to a voltage divider consisting of Rb and the base-emitter region of the BJT where Vbe is developped.