Hi all,

Last night, I was working on one stage of a multistage transistor amplifier that uses a 2N2222 with a gain of ~2:

I had my guitar running through it and outputted it to a DAW to look at the waveform (best I can do without an oscilloscope), and noticed that the bottom half of my waveform was getting cleanly clipped off like so:

I had set the q-point of my circuit to be 4.75 with a supply voltage of about 9.5V, so this didn't make any sense. I kept messing around with different capacitors and other point of the circuit before I figure I would just set the q-point a bit higher, to about 6V (I just switched out R3 with a 300k). This worked and my waveform did not clip.

Does anybody have any sense of why my circuit was even clipping in the first place, and having a q-point above 1/2 Vs stopped the clipping?

I had a few guesses but they seem relatively outlandish and I could not find any information about them online. For instance, is there some kind of lower voltage limit caused by the PN junction of the transistor? This seems incredibly unlikely since this would affect the performance of every transistor amplifier.

Or, does the bias point actually depend on the Collector-Emitter voltage drop, which at high gains is essentially the Collector voltage w/r/t the ground? I.e. in my circuit, the q-point was 4.75, but Ve was ~2.8V, which would make Vce only 2. Raising the q-point to 6 makes Vce ~3V which I have seen as about normal for an operating point in a transistor amplifier (1/3Vs < Vc < 1/2Vs).

This is all conjecture though. I'm pretty confused still, and if anybody can shed light on this problem I would greatly appreciate it.

Thanks,
-Carl

 

I recommend you to slightly improve the schematic with a single transistor. An alternative would be to switch to an audio preamplifier with two transistors. The Elektor magazine collection of the 1980s and 1990s was a schematic that could help you!
On the other hand, in my website I made an article for sizing circuit with transistor in common emitter connection. I recommend you to use Google Translate, you will find many interesting things about your problem. See link:
http://www.tehnium-azi.ro/page/arti...lul_simplificat_al_preamplificatoarelor_audio

 

The circuit you posted has no bias on the base. Hopefully, it is only because the schematic is drawn wrong (missing the dot on the intersections of the voltage divider and the base.)

Bob

 

The circuit you posted has no bias on the base. Hopefully, it is only because the schematic is drawn wrong (missing the dot on the intersections of the voltage divider and the base.)

Bob

Hi Bob,

Yes, that's exactly it. Thank you for pointing that out. I'm still learning LTSpice. R3 and R4 are both connected to the base.

-Carl

 

I should also mention that I was able to verify the q-points I had set with voltmeter and they were within ~.05V of what I had calculated.

-Carl

 

I recommend you to slightly improve the schematic with a single transistor. An alternative would be to switch to an audio preamplifier with two transistors. The Elektor magazine collection of the 1980s and 1990s was a schematic that could help you!
On the other hand, in my website I made an article for sizing circuit with transistor in common emitter connection. I recommend you to use Google Translate, you will find many interesting things about your problem. See link:
http://www.tehnium-azi.ro/page/arti...lul_simplificat_al_preamplificatoarelor_audio

Hi Don,

I'm not sure what you mean by "improve the schematic with a single transistor." There is only one transistor in this. That being said, this is not a standalone circuit, but will be integrated into a later project. I am just trying to understand why there is clipping even with such a low gain and a well defined q-point.

 

A better solution is to connect R3 from the base to the collector of the 2N2222. This is why a scope is a valuable aid in designing circuits. Sometimes you just have to forget the math and fine tune by visual inspection.
SG

 

A better solution is to connect R3 from the base to the collector of the 2N2222. This is why a scope is a valuable aid in designing circuits. Sometimes you just have to forget the math and fine tune by visual inspection.
SG
View attachment 148863

Wouldn't the gain of the circuit change since now there is negative feedback through R3?

 

Yes, so change R2 to 3.3K. The gain will also be affected by the impedance of the following stage. The 3.3K resistor gives a gain of 2 into a 47K load. See revised schematic in post #7
SG

 

The negative clipping is due to the voltage drop across R2, which means the output signal can't go to ground, and thus can't give an output that is symmetrical around a bias point half-way between V+ and ground.
So, for a symmetrical output, the bias point needs to be set about halfway between V+ and the emitter voltage.

 

The negative clipping is due to the voltage drop across R2, which means the output signal can't go to ground, and thus can't give an output that is symmetrical around a bias point half-wave between V+ and ground.
So, for a symmetrical output, the bias point needs to be set about halfway between V+ and the emitter voltage.

Thank you so much! This is exactly what I was asking. I had a feeling it was something along the lines of this. So let me see if I get this: when the output voltage gets down to the emitter voltage, it simply stays that voltage because the whole collector-emitter path basically becomes a voltage divider between R1 and R2?

 

Thank you so much! This is exactly what I was asking. I had a feeling it was something along the lines of this. So let me see if I get this: when the output voltage gets down to the emitter voltage, it simply stays that voltage because the whole collector-emitter path basically becomes a voltage divider between R1 and R2?

I suppose this is why it seems that higher gain amplifiers don't have this issue: the emitter voltage is so much smaller that using a bias point between Vcc and the ground is essentially the same as using one halfway between Vcc and Ve.

 

Thank you so much! This is exactly what I was asking. I had a feeling it was something along the lines of this. So let me see if I get this: when the output voltage gets down to the emitter voltage, it simply stays that voltage because the whole collector-emitter path basically becomes a voltage divider between R1 and R2?

You got it; Grasshopper.

 

I suppose this is why it seems that higher gain amplifiers don't have this issue: the emitter voltage is so much smaller that using a bias point between Vcc and the ground is essentially the same as using one halfway between Vcc and Ve.

Usually the amplitude of the signal is not that high in an audio amp expect for the output stages.
And those are typically push-pull, and designed to swing to near the supply rails.

 

The push pull output stage is made of two voltage followers with no voltage gain, so the previous stage needs to swing just as far as the output stage.

Bob