🎁 New account: buy your first 10 credits, get 10 more free.Avg. 25–45 min turnaround · Files written on your original readCreate free account

A continuously variable transmission (CVT) has no fixed gears. Instead of switching between ratios like a conventional automatic, it changes the ratio smoothly between a lowest and a highest value, so the engine can stay at the speed that suits the driver’s request. Many small and mid-size cars, several hybrids and almost every automatic scooter use some form of CVT. For anyone who works with transmission control units (TCUs), diagnostics or engine calibration, it helps to understand how the variator works and how closely the TCU depends on information from the engine ECU.

Diagram of a CVT variator showing primary and secondary pulleys with the belt running radius in low ratio and in overdrive
CVT variator in low ratio and overdrive (original diagram)

The variator: two pulleys and a belt or chain

The heart of a car CVT is the variator. It has two pulleys: a primary (drive) pulley connected to the engine side and a secondary (driven) pulley connected to the wheels. Each pulley is made of two conical halves called sheaves. One sheave is fixed to the shaft and the other can slide along it, pushed by hydraulic pressure.

When the movable sheave moves towards the fixed one, the V-shaped groove narrows and the belt is forced to run on a larger radius. When the sheaves move apart, the belt drops to a smaller radius. Because the belt length is fixed, the two pulleys always work in opposite directions:

Two main types of belt are used. The steel push belt is built from hundreds of thin steel elements held together by packs of thin steel bands; it transmits torque by the elements pushing against each other. The chain type uses plate links joined by pins whose ends contact the sheaves; it transmits torque by pulling. Both rely on friction between steel surfaces lubricated by a special fluid, which is why fluid type and condition matter so much.

Around the variator: launch device and direction

A variator cannot start the car from standstill on its own, and it cannot reverse. Most car CVTs therefore add:

An oil pump driven by the engine supplies the pressure for all of this, regulated by solenoids in the valve body.

How the TCU controls a CVT

Concept diagram of how the TCU controls a CVT: target engine speed, ratio control, clamping pressure from the engine torque signal and speed sensor feedback
How the TCU controls ratio and clamping force (original diagram)

The TCU does two jobs at the same time: it sets the ratio and it makes sure the belt does not slip.

Ratio control

Instead of a shift map with gear changes, a CVT calibration usually works with a target primary pulley speed, which is effectively a target engine speed. It is looked up from pedal position or torque request and vehicle speed, then adjusted for drive mode, fluid temperature, gradient and other conditions. The TCU calculates the actual ratio from the primary and secondary pulley speed sensors and moves the ratio-setting pulley until actual matches target. Many modern calibrations also simulate stepped “gears” during hard acceleration or in manual mode, because drivers dislike the constant-rpm feel.

Clamping force

The other pulley’s pressure sets how hard the sheaves squeeze the belt. Too little clamping and the belt slips, which quickly damages the belt and the sheave surfaces. Too much clamping wastes energy in the oil pump and adds wear. The TCU therefore calculates the required clamping force from the torque that reaches the variator, plus a safety margin. That torque value normally comes from the engine ECU over CAN, which links CVT control directly to the engine torque model. Which pulley handles ratio and which handles clamping depends on the design.

Protection and fail-safe

If the TCU detects slip, a ratio it cannot reach, high fluid temperature or a sensor fault, it can request an engine torque reduction, store a code and switch to a fail-safe strategy with a limited ratio range. The fluid temperature sensor and the pressure control solenoids play a central role here.

Scooters, ATVs and tractors

Automatic scooters, many ATVs/UTVs and snowmobiles use a rubber V-belt CVT that is purely mechanical. Centrifugal rollers or weights in the drive pulley push the sheaves together as engine speed rises, while a spring-loaded, torque-sensing driven pulley resists. There is no TCU; the behaviour is tuned with roller weights, springs and clutch settings. Tractors use a completely different concept, the hydro-mechanical power-split CVT, covered in our guide to how a tractor CVT works.

Diagnosing common CVT complaints

Six-step workshop check order for CVT judder, slip or limp mode covering codes, fluid level and type, live data, software and adaptations, and wiring
Check order for CVT judder, slip or limp mode (original diagram)

Typical complaints are judder when pulling away, a whine or drone that changes with ratio, engine speed flaring without matching acceleration, and a fail-safe mode with weak acceleration or limited top speed. A sensible order of work:

What this means for tuning

Because clamping force is calculated from the engine torque signal, an engine remap on a CVT car must respect the transmission’s torque limits. If the actual torque rises but the signal sent to the TCU no longer represents it accurately, clamping may be too low and the belt can slip. Many CVTs have conservative torque capacity, so a responsible calibration keeps transmission protection in place and is built on the original read of the exact engine ECU and TCU software. Always keep full original backups of both units.

Related fault codes (DTCs)

These generic codes are often seen with CVT faults. Many manufacturers also use their own codes for belt slip, ratio errors and pulley pressure, so always read the TCU directly.

CodeMeaningTypical causes
P0700Transmission control system (MIL request)The TCU has stored its own fault and asked the engine ECU to light the MIL; read the TCU for the actual code
P0711Transmission fluid temperature sensor “A” circuit range/performanceSensor drift, implausible reading versus running time, wiring or connector fault, sensor on internal harness
P0715Input/turbine speed sensor “A” circuitPrimary pulley or input speed sensor fault, wiring, connector, damaged sensor target
P0720Output speed sensor circuitSecondary pulley or output speed sensor fault, wiring or connector problem, metal debris on the sensor
P0730Incorrect gear ratioActual ratio not matching the commanded ratio: belt or chain slip, low or degraded fluid, pressure control problem, speed sensor error
P0741Torque converter clutch circuit performance or stuck offLockup clutch wear or judder, TCC solenoid or valve fault, wrong or worn fluid
P0746Pressure control solenoid “A” performance/stuck offSticking solenoid or valve in the valve body, contaminated fluid, low line pressure
P0776Pressure control solenoid “B” performance/stuck offSticking solenoid or valve, debris in the fluid, internal leak
P0841Transmission fluid pressure sensor/switch “A” circuit range/performancePressure sensor fault, wiring, actual pressure out of range from a pump, valve or leak problem
P0868Transmission fluid pressure lowLow fluid level, worn oil pump, clogged strainer or filter, internal leakage, pressure regulator fault
P0218Transmission fluid over temperature conditionTowing or long climbs, blocked or bypassed cooler, low fluid, sustained slip, temperature sensor fault

Generic SAE J2012 code meanings shown. Exact wording and enable conditions vary by manufacturer; always check service data.

Need a tuning file built on your original ECU read? Upload it to ECUFileMarket and get a file prepared for your exact hardware and software version.

Leave a Reply

Your email address will not be published. Required fields are marked *

⚡ New Free Workshop Resource: Hybrid & EV Powertrain Engineering & ECU Diagnostics Training (8 Free PDF Manuals) →