Leslie Motor Simulations

For Players

The Leslie tremolo motor can be controlled two ways: the existing TRIAC circuit that chops the mains voltage for two fixed speeds, or a proposed micro-VFD that synthesizes variable-frequency AC for continuous speed control. These LTspice simulations model both approaches against the same motor to evaluate which fits the MIDI conversion best. The takeaway: the TRIAC works fine for fast/slow switching, but the VFD opens up expressive possibilities — continuous speed via mod wheel, programmable ramp curves — for about $10 in parts.

The Motor

The Leslie Tremolo Unit (CBS/Electro Music, Part 660890) uses an 18W shaded-pole AC induction motor. It is a 20-pole design with a synchronous speed of 360 RPM at 60 Hz. Both simulations share the same electrical model: a series R-L circuit with R = 300 ohm winding resistance and L = 1H motor inductance. At 60 Hz this gives an impedance of about 482 ohm, a power factor of 0.62, and a rated current of 0.24A — yielding the 17.6W nameplate power.

The 1H inductance is the dominant characteristic. It causes current to lag voltage by about 52 degrees at 60 Hz and, as we will see, makes the motor an effective low-pass filter for high-frequency switching signals.

See Leslie Control for the physical unit, schematic labels, and wiring details.

TRIAC Phase-Angle Control (Existing Circuit)

The stock Leslie speed controller fires a TRIAC at different phase angles each half-cycle of the 60 Hz mains. Earlier firing delivers more of the sine wave to the motor (more voltage, more power); later firing chops the waveform aggressively, reducing RMS voltage and torque. A reed switch selects between two fixed firing angles — FAST and SLOW — actuated by a coil driven from the console’s three-position switch.

Results

Setting	Firing Angle	Motor RMS Voltage	% of Line	Power	Speed
FAST	25 deg	116.1V	99.2%	~17.3W	~340 RPM
SLOW	110 deg	62.7V	53.6%	~5.1W	~40 RPM

Motor Terminal Voltage

Motor terminal voltage, FAST and SLOW overlaid. FAST is nearly a full sine wave; SLOW shows severe chopping with conduction only in the last ~70 degrees of each half-cycle.

Motor Winding Current

Motor winding current for both speeds. Current lags voltage — the inductive load continues drawing current past the voltage zero crossing. The simulation models this by extending the TRIAC conduction window 55 degrees past zero to account for the power factor lag (real TRIACs conduct until current, not voltage, crosses zero).

Circuit Notes

The simulation includes a 100 ohm + 100 nF R-C snubber across the TRIAC, standard practice for inductive loads to damp the voltage spike at commutation. The power factor lag modeling is important: without it, the current waveform would show an unrealistic sharp cutoff at the voltage zero crossing.

The key observation is that the motor’s synchronous speed stays at 360 RPM regardless of TRIAC firing angle — speed reduction comes entirely from high slip and torque loss. At SLOW, the motor is operating deep in its unstable torque-speed region, fighting its own rotating magnetic field. This works, but it is thermally inefficient and gives no possibility of intermediate speeds.

ESP32 MCPWM Micro-VFD (Proposed)

The alternative approach discards phase-angle chopping entirely. Instead: rectify the 117 VAC mains to a 165V DC bus, then use the ESP32-S3’s MCPWM peripheral to generate sinusoidal pulse-width modulation (SPWM) through a full H-bridge at variable frequency. Constant V/Hz scaling — modulation index m = f_motor / 60 — maintains motor flux across the entire speed range.

Motor as Filter

The motor’s 1H winding inductance is the secret weapon here. At the 10 kHz SPWM switching frequency, the motor impedance is 62.8k ohm — switching-frequency current ripple through the winding is under 3 mA, negligible compared to the hundreds of milliamps of fundamental current. The motor itself filters the PWM into a clean sine. No external LC filter is needed.

LC Filter Resonance Trap

An external LC sine filter was tested in simulation and rejected. A filter capacitor sized for 10 kHz attenuation resonates with the motor’s 1H inductance at approximately 50.1 Hz — right in the middle of the operating range. The 5th harmonic of the 10 Hz drive frequency lands directly on this resonance, causing destructive voltage amplification. Industrial VFDs solve this with active damping firmware, but for an 18W motor the simpler and safer approach is to skip the filter entirely and let the motor winding do the work.

Results

Frequency	Mod Index	V_fund (RMS)	Motor Z	Current	Power	Sync Speed
10 Hz	0.167	19.4V	306 ohm	63 mA	1.2W	60 RPM
20 Hz	0.333	38.9V	325 ohm	120 mA	4.3W	120 RPM
40 Hz	0.667	77.8V	391 ohm	199 mA	11.8W	240 RPM
60 Hz	1.000	116.7V	482 ohm	242 mA	17.6W	360 RPM

Motor Winding Current (All Frequencies)

Motor winding current at all four frequencies overlaid. Each is a clean sinusoid — the motor inductance completely filters the 10 kHz switching. Current scales linearly with V/Hz, from 63 mA at 10 Hz to 242 mA at rated 60 Hz.

H-Bridge SPWM Output

Raw H-bridge SPWM output (100-200 ms window). The output swings between +165V and -165V with duty cycle modulated by the sine reference. This is what the motor terminals see before the winding inductance filters it.

Voltage vs. Current at 60 Hz

SPWM voltage and motor current overlaid at 60 Hz (150-200 ms). The raw PWM voltage (sharp bipolar switching) produces a smooth sinusoidal current through the motor — a direct demonstration of the motor-as-filter principle.

Comparison

	TRIAC (Existing)	VFD (Proposed)
Speed control	2 fixed speeds (FAST / SLOW)	Continuous, 60-360 RPM
Speed method	Voltage reduction at fixed 60 Hz	Frequency + voltage reduction (V/Hz)
Motor behavior at low speed	High slip, fights sync speed	Optimal slip at every speed
Interface to ESP32	Relay or optocoupler across reed coil	MCPWM direct to H-bridge
MIDI mapping	CC#80 three-state (off/slow/fast)	CC#1 continuous + CC#80 presets
Ramp profiles	Mechanical inertia only	Programmable in firmware
Additional BOM	~$2 (relay + driver)	~$10 (rectifier, caps, MOSFETs, gate driver)
Mains isolation	Inherent (relay/opto)	Requires design attention

Both approaches are viable. Scenario A (relay drive of the existing TRIAC circuit) is the baseline — proven, simple, zero risk to the motor. The VFD is the stretch goal: it unlocks continuous speed and firmware-shaped ramp curves, making CC#1 mod-wheel Leslie control musically meaningful.

VFD Candidate BOM

Component	Value	Purpose
Bridge rectifier	4x 1N4007	Mains rectification to 165V DC
Bulk capacitor	220 uF / 200V	DC bus smoothing
MOSFETs	4x IRF840 (or IRFP460)	H-bridge power stage
Gate driver	IR2110 or IR2184	High/low-side MOSFET drive with bootstrap
ESP32-S3	MCPWM peripheral	SPWM generation, V/Hz control
Bootstrap caps	2x 10 uF / 25V	Gate driver bootstrap supply

Total power-stage BOM is under $10. The ESP32 is already part of the MIDI conversion design — the MCPWM peripheral is a hardware feature that costs nothing extra.

Source Files

The SPICE netlists are in the project repository:

• leslie-triac-motor-control.cir — TRIAC phase-angle simulation, .step param alpha list 25 110
• leslie-vfd-motor-control.cir — ESP32 MCPWM VFD simulation, .step param f_motor list 10 20 40 60

Both run via LTspice under Wine on Linux. See the sims/ README for instructions.

---

Cross-References

• Leslie Control — physical unit identification, schematic, TRIAC circuit details, and MIDI interface options
• Simulations — CapSense detection chain simulations (separate subsystem)
• Approach — overall MIDI conversion architecture
• Implementation Roadmap — build sequence and phase plan