When the master volume is after the tone stack (stock 2203 location), turning the master down attenuates everything that's already passed through the tone stack and phase inverter. The signal between the master and the speaker is small; hum and RF pickup at that point are inaudible.
When the master is moved before the tone stack (the Jose mod), the entire downstream path — tone stack, phase inverter input, the wiring between them — becomes a wide-open signal path even when the master is at zero. Per Dave Friedman, Headfirst Amp Panel #14 (April 2026):
*"As soon as you move [the master] before the tone stack, the whole tone stack's just wide open into the phase inverter… it's going to pick up everything. [Hum] running to the master volume."*
The phase-inverter input is the most-sensitive node in the chassis — *"like a big aerial"* per Friedman. With a pre-tone-stack master, anything that picks up hum or RF between the master and the phase inverter feeds straight into a fully-amplified power-amp path.
Observable consequence on real Jose amps: Friedman in Amp Chat #17 (Feb 2025): *"Original Joses bleed like a bitch when that master is at zero."* This is a documented characteristic of period Jose amps, not a fault of any one chassis.
Modern remediation: Friedman's own production "honoring Jose" amp [Panel #1, Feb 2025] solves the bleed via careful layout — short leads, controlled grounding around the master pot, shielding where needed. *"We built a fully loaded two-channel Jose with two master volumes, switchable that does not bleed."* The hazard is a layout problem, not a topology problem; it's solvable with care.
Why JMIL surfaces this: the lab's audio engine models the signal-path topology correctly (pre-tone-stack MV produces the right tonal behavior). What the simulation can't capture is the physical-build hazard: builders re-creating the Jose MV mod on a real chassis need to plan for the bleed risk. Documenting the hazard is part of honest pedagogy.