Chuck D. Bones
Circuit Wizard
Before I go any further, let's talk about negative feedback. People don't always like negative feedback in their personal lives, but we absolutely depend on it in electronics. Stephen S. Black is credited with discovering the negative feedback amplifier in 1927 while working at Bell Labs. The Telephone Company was using vacuum tubes as repeater amplifiers throughout their telephone network, but they had a problem: the amplifiers distorted, had poor frequency response and their gain drifted. Negative feedback fixed all of that. The cost was a few extra parts and some gain, but it was well worth it. The significance of this breakthrough cannot be overstated. We use negative feedback in pedals to set and stabilize the bias point, control gain and frequency response, and increase or decrease distortion. It all depends on what's inside the feedback loop. The last two examples in part 1 used an emitter resistor to create negative feedback and help stabilize the bias point. It works like this... Suppose the base voltage is constant and the temperature starts rising. The temperature rise makes Vbe go down which will make the collector current go up. The emitter current is almost entirely collector current, so that goes up too. Increasing emitter current raises the emitter voltage which effectively lowers the base voltage from the transistor's point of view. While a temperature increase is trying cause the collector current to go up, the increasing voltage drop in the emitter resistor is driving the collector current back down. The emitter resistor doesn't make the change in collector current go to zero, but it does make the change much smaller.
OK, on to the next example. This is from the Quarantine Fuzz, but can also be found in the BMP, DS-1 and many others. This circuit is very similar to the last one in part 1, the only difference is now R8 is connected between base & collector instead of base & Vcc. We already had a negative feedback path due to the emitter resistor (R10), but now we have a 2nd feedback path via R8. We can estimate the bias point thusly: The voltage drop in R10 is usually much smaller than the transistor's Vbe (0.65V), so we'll assume the voltage across R10 is zero. We'll also assume that the base current is much smaller than the current in R7. Not a super-accurate assumption, but good enough for estimating the bias point. The base voltage is 1 Vbe. R7 & R8 make a 5.7:1 voltage divider. So we'll estimate the collector voltage Vc = 5.7 * Vbe = 3.7V. Those of us who have measured this circuit know that estimate is a little low, the collector voltage is usually between 4.0V and 4.5V. But it's pretty close for a cocktail napkin calculation. R8 doesn't just stabilize the operating point, it helps set the gain. C5 is also in the feedback loop; it affects the gain at higher frequencies, providing some treble cut.
A variation on that circuit is to omit R7. Here's the output stage from the MI Audio Tube Zone v3 (PedalPCB Valve Stem). The only advantage of deleting the resistor from base to ground is that now Vc is free to move if Vcc changes. This pedal can be run on 9V or 18V. This bias circuit automatically adjusts Vc to keep it somewhere near 1/2 Vcc. It's not super stable, but it's good enough. The EQD Arrows (Amentum Boost) uses the same biasing arrangement.
Part 3 will focus on the Fuzz Face. Give me a couple of days to crank that one out.
OK, on to the next example. This is from the Quarantine Fuzz, but can also be found in the BMP, DS-1 and many others. This circuit is very similar to the last one in part 1, the only difference is now R8 is connected between base & collector instead of base & Vcc. We already had a negative feedback path due to the emitter resistor (R10), but now we have a 2nd feedback path via R8. We can estimate the bias point thusly: The voltage drop in R10 is usually much smaller than the transistor's Vbe (0.65V), so we'll assume the voltage across R10 is zero. We'll also assume that the base current is much smaller than the current in R7. Not a super-accurate assumption, but good enough for estimating the bias point. The base voltage is 1 Vbe. R7 & R8 make a 5.7:1 voltage divider. So we'll estimate the collector voltage Vc = 5.7 * Vbe = 3.7V. Those of us who have measured this circuit know that estimate is a little low, the collector voltage is usually between 4.0V and 4.5V. But it's pretty close for a cocktail napkin calculation. R8 doesn't just stabilize the operating point, it helps set the gain. C5 is also in the feedback loop; it affects the gain at higher frequencies, providing some treble cut.
A variation on that circuit is to omit R7. Here's the output stage from the MI Audio Tube Zone v3 (PedalPCB Valve Stem). The only advantage of deleting the resistor from base to ground is that now Vc is free to move if Vcc changes. This pedal can be run on 9V or 18V. This bias circuit automatically adjusts Vc to keep it somewhere near 1/2 Vcc. It's not super stable, but it's good enough. The EQD Arrows (Amentum Boost) uses the same biasing arrangement.
Part 3 will focus on the Fuzz Face. Give me a couple of days to crank that one out.
Last edited:
