423 DA amplitude ladder is throttle-selected: light-gate array (4-4) BS1-BS4 to booster switch (4-14) to relay-tapped VIC primary (4-20); gate rate is crank speed (4-8)

COMPLETE AMPLITUDE-LADDER CHAIN (Memo WFC 423 DA, verified from scans of Fig 4-3, 4-4, 4-8, 4-20):
accelerator pedal -> LIGHT-GATE MODULE (32~200 light-circuit array, Fig 4-4: SEP8526 emitters at 220 ohm, movable light-gate, SDP8601 optoschmitt receivers, output HIGH when light activated, isolation diodes into 3.3K ladder, 5K IDLING SPEED trim, "crossover switch function provides smooth switch action") -> taps BS1/BS2/BS3/BS4 -> DIGITAL BOOSTER SWITCH CIRCUIT (4-14) -> COIL SWITCH-IN CONTROL (4-20B: 4050 buffers, H11D1 optos, TIP120, 1N4001-clamped relay coils) -> RELAY ARRAY tapping segments of the VIC PRIMARY COIL -> amplitude rungs Vo/Va/Vb/Vn ("VARIABLE DIGITAL VOLTAGE AMPLITUDE CONTROL", "BOOSTER VOLTAGE").

KEY CONSEQUENCE: the amplitude ladder is selected by THROTTLE POSITION, four bits of it, and it works by changing TURNS RATIO, not supply voltage. Confirms the Fig 4-20 note "APPLIED VOLTAGE POTENTIAL REMAINS THE SAME WHILE COILS ARE SEQUENTIAL SWITCH IN OR OUT" and the Fig 4-2 note that the booster coil holds primary input amplitude constant while raising secondary level.

GATE RATE IS CRANK SPEED (Fig 4-8, "INJECT TIME SEQUENCING (POWER STROKE STAGE)"): trigger-input repeat range @800 RPM /2 /60 = 6.6 pulses/sec; @3K RPM = 25 pulses/sec; @8K RPM = 66.6 pulses/sec (the /2 is four-stroke). So the relaxation/gate envelope is 6.6-66.6 Hz, not a free parameter. Other Fig 4-8 data: 5 degrees firing displacement; 3K RPM @ .347 ms; 800 RPM idling 6.65 ms; .0694 ms per degree; max speed (b) 135 degrees displacement 9.369 ms. Input from Laser Distributor Pulse Shift Control (4-18), or direct to Pulse Expander Crk't (4-7A) when the laser distributor has a mechanical advance clutch assembly. Inject Driver (4-8A): H11D1 -> TP120 Darlington -> RL, 1N4001 -> water inject coil.
ERRATUM ON THE SHEET: "25 MS = 3K R.P.M." is wrong; 3000 RPM is 20 ms/rev. The .0694 ms/degree figure corresponds to 25 ms/rev = 2400 RPM. The pulses/sec column is arithmetically correct.

CRANK-SYNC CITATION: use Figures 4-8 and 4-18. NOT 4-25, which is the Emergency Shut-Down Circuit.

ALSO ON 4-4: LM131 voltage-to-frequency converter, 10 kHz full scale, 5 V half scale (this is the "Data Link V/F ~ F/V" that feeds the LM111 in Fig 4-21); LM3914 dot display as Laser Accelerator Scan Circuit, LED 1-10. Fig 4-3 "Opto Circuit" is the primitive: LED, movable light-gate, optoschmitt receiver with internal voltage regulator.

FIG. 20YID PROVENANCE (now verified from a readable scan): sheet reads "COPYRIGHT (C) 1982 BY STANLEY A. MEYER", page marker G39, figure numbered "20YID" with NO Memo WFC header. It is LEGACY 1982 material predating both 422 DA and 423 DA, and must not be presented as part of the 423-DA-2 architecture. Items confirmed: Electron Extractor Grid (51), insulation gap (45), Gas Resonant Cavity (110), laser injection port (44), combustible gas inlet port at bottom with cavity body grounded, Voltage Intensifier Crk't 9XA driving pulse (61).
EXTRACTION PATH CORRECTION to notebook #3367: the grid is not merely biased positive — the extracted electrons return via 56 -> 52 -> ENERGIZED LIGHT BULB (180) / filament wire (53, 55), annotated "CONSUMING ELECTRONS", then blocking diode (57) to B+ (54). The bulb filament IS the current-limiting ballast and the extraction current does useful work; the diode blocks B+ back-feeding the grid. Build the extractor with a resistive load in the return leg.