How this gain stage works

This is the common-emitter amplifier, the BJT counterpart of the JFET booster, built here around a 2N5089. The signal couples in through C1 to the base. Unlike a JFET, whose gate just sits at ground, a BJT needs its base held about 0.65 V above the emitter and it draws real base current, so R3 and R4 form a voltage divider that props the base up at the right voltage. That divider, together with the base itself, also sets the input impedance, which is why a common-emitter stage loads your pickups far more than a JFET or op-amp stage does.

The divider and R5 set the collector current. R6 then converts that current into voltage swing, and it also positions the collector's DC voltage: it does not set the bias current the way the JFET's drain resistor had to be matched to the transistor, but make it too big for the current flowing and the collector runs out of headroom and saturates. Your part ranges here include combinations that would cut the transistor off entirely or slam it into saturation with fixed parts, so this page gives R4 the same treatment as the JFET calculator: whenever you move R3, R5, or R6, it re-solves R4 so exactly half the supply (4.5 V) sits across the transistor. That rule is the same as "collector at half voltage" when R5 is small, and it keeps the stage working across every combination, including R5 at 0. The R4 it lands on for the stock values is essentially the 1 MΩ on the schematic.

Gain is the drain-to-source story again with different names: A ≈ Z_L / (r_e + R5), where r_e is the transistor's own internal 26 mV/I_E resistance. Leave C4 out and R5 stays in the signal path, trading gain for cleanliness and stability. Switch C4 in and it shorts R5 for AC, and the gain climbs toward the maximum the stage can give. The size of C4 sets how much of the band gets that boost: big caps lift everything, small caps lift only treble, and the fz and fp markers show the transition. Pull R5 all the way to the 0 Ω jumper and you get maximum raw gain, but in a real build that point is fragile: bias then depends entirely on the transistor's own spread and temperature, which is exactly why R5 exists.

The math uses a typical 2N5089 (β ≈ 500, V_BE ≈ 0.65 V, ideal hybrid-π model, V_CE pinned at 4.5 V). Real 2N5089s run β anywhere from about 450 to 1800, which barely matters while R5 is in circuit and matters a lot at the jumper setting. C2 couples the output and sets a bass corner against R7, and R8 isolates the stage from whatever comes next. The dashed curve is the nearest-E12 build, including the E12 value of the auto-computed R4.