### The Power Amp

The Fender Bassman 5F6-A push-pull power amp has two 6L6-compatible 5881 tubes operating Class AB.

Historically the plate, screen, and grid bias supply voltages are

VPP = 432 volts
VSS = 430 volts
VGQ = -48 volts

(When driven by modern AC power the voltages tend to be a bit higher.) The parts values are

CG = 0.1uF
RG = 220k
RS = 470

The DC grid bias of minus 48 volts is supplied by a tap on the power transformer. After rectification by a solid-state diode the AC ripple is attenuated via a tradional RC filter. The voltage drop across the screen resistors is negligible at idle, so we can assume the screen voltage, measured relative to the cathode, is 430 volts minus the voltage across the cathode resistor. Since the plate and screen voltages are nearly identical, we can compute the cathode current, representing the sum of the plate and screen currents, using 6L6GC triode-connected curves.

From the DC operating point marked by a red circle, the cathode current is 43 milliamps. From the transfer characteristics, the screen current is only about 1 milliamp:

We conclude that the DC operating point is approximately

VGQ = -48
VPQ = 432
VS = 430
IPQ = 33mA
IS = 1mA

The plate current is only about half the electron flow in the Vox AC30, which uses smaller EL84 pentodes. The Bassman is much more coldly biased.

Plate and screen dissipations are

P = (432)(33mA) = 14 watts
P = (430)(1mA) = 0.4 watts

This compares to 6L6GC limits of 30 watts and 5 watts, respectively. The tube heater produces 6 watts, so heat production almost triples when the standby switch is on.

## AC Characteristics

Published plate characteristic curves for the 6L6 are available for 400-volt screens, not 430-volt screens. There are methods of scaling the curves to the higher voltage1 but we can get a rough estimate of power output by considering how the amp performs at 400 volts and then scaling our conclusion. Here are the 400-volt curves:

The 5F6-A uses a 4k output transformer which, for Class AB, represents 1k for the conducting tube. The red line represents the load line corresponding to 1k. Notice that it passes though the knee of the zero-grid-voltage curve, which is typical for push-pull designs using beam power tetrodes like the 6L6. If it were possible for the plate voltage to swing all the way to zero then at full power plate voltage would swing from an idle value of 400 volts to a minimum of zero while the net plate current swings from zero to 400 milliamps. This creates a power output of

P = (400)(400mA)/2 = 80 watts

In reality, however, the plate voltage and current swing only as far as the blue circle, representing a voltage swing of 400 - 90 = 310 volts and a current swing of zero to 320mA. The more realistic power output, measured at the primary, is

P = (310)(320mA)/2 = 50 watts

This is 40 percent less than the simpler calculation. An actual 400-volt amplifier would produce substantially less than 50 watts because of output transformer losses.

A load line for a plate voltage of 432 volts and the same output transformer is shown by the green line. The curves are invalid for the higher screen voltage, but in this case all we are only interested in extremes of the graph. For a plate voltage swing of 432 volts and a plate current swing of 432 milliamps the power output is

P = (432)(432mA)/2 = 93 watts

Subtracting 40 percent we get an estimate of 56 watts RMS, measured at the primary before output transformer losses. This assumes, however, that the power supply can sustain 430-volt screens and a 432-volt plate supply at full power. In reality the power supply sags well below 400 volts, as described in "The Strange Effects of AC Ripple on a Class AB Power Amp."

## Bass Response and Bias Excursion

Our Grid Bias Excursion Calculator shows a bass cutoff frequency of 6 Hertz and a bias recovery time of 29 milliseconds. By comparison the Marshall Model 1987 "Plexi" has a bass cutoff frequency of 25 Hertz and a bias recovery time of 6 milliseconds. The key difference is that Marshall uses 0.022uF coupling capacitors instead of Fender's 0.1uF. The short excursion time, 7 milliseconds, which is a characteristic of both amps, indicates that the power amp moves rapidly closer to Class B when overdriven. Marshall's DC grid bias, however, recovers much more quickly when the guitar player lightens up on the pick. When it comes to the dynamics of overdrive, the question of whether "response" represents an improvement over "sustain" is a matter of personal preference.

Since the lowest note on a guitar with normal tuning is 82 Hertz, there is no significant effect in the 20-Hertz difference in bass response. Unless, of course, your original intent was to create a bass guitar amplifier...

## References

1Richard Kuehnel, Circuit Analysis of a Legendary Tube Amplifier: The Fender Bassman 5F6-A, 3rd Ed., (Seattle: Pentode Press, 2009).

2Richard Kuehnel, Vacuum-Tube Circuit Design: Guitar Amplifier Power Amps, (Seattle: Pentode Press, 2008).