Technical Note: PDK Transmission Speed Sensor Operation and Diagnosis
Technical Note: Speed Sensors in PDK Transmissions
The speed sensor is a critical component of Porsche's PDK dual-clutch transmission. It continuously measures the rotational speed of internal shafts and feeds this data to the transmission control unit. This real-time feedback allows the control unit to synchronize shift operations with precision and confirm that each gear engagement has been completed correctly. Should the sensor fail—the gearbox loses its ability to verify shaft speed. In that state, the control unit refuses to perform further shifts; it holds the current gear until the vehicle slugs to a stop, after which it will not permit the car to move off again.
How the Transmission Speed Sensors Work
The transmission uses two input shafts, with each shaft connected to its own clutch.
The engine flywheel is directly connected to the clutch pack input. As a result, clutch input speed is measured by the engine speed sensor (crankshaft position sensor), while clutch output speed is measured by the transmission speed sensors.
For this reason, it is often more intuitive to think of these sensors as “clutch output sensors” rather than input speed sensors.
- Speed sensor channel 1 measures the speed of a notched impulse wheel pressed onto shaft 1.
- Speed sensor channel 2 measures the speed of the individual gear teeth on shaft 2.
How the TCU Detects Clutch Slippage
Clutch slippage is determined by comparing engine speed with the transmission shaft speed reported by the speed sensor.
When a clutch is fully clamped, engine speed should match the transmission speed of the corresponding shaft.
Transmission output speed is calculated using an average of the wheel speed sensor signals. Based on the gear currently engaged, the TCU (Transmission Control Unit) knows exactly what wheel speed should correspond to the speed sensor output.
This process is known as gear ratio monitoring, and the TCU performs this check continuously during transmission operation.
Because the PDK pre-selects a gear on the clutch that is not currently engaged, the TCU can monitor the speeds of both transmission shafts at the same time—not only the shaft connected to the active clutch.
Speed Sensor Voltage and Signal
The transmission speed sensor operates from an 8.5V supply provided by the TCU. It is the only transmission sensor that uses this 8.5V supply.
Both speed sensor channels share the same 8.5V supply, but each channel has its own individual output signal to the TCU.
When electrical power is supplied to the TCU with the ignition switched on, each speed sensor channel sends a baseline signal of approximately 0.75V back to the TCU.
This baseline voltage allows the TCU to confirm that the speed sensor is connected, online, and operating.
When the shaft begins to rotate, the corresponding speed sensor detects each passing tooth and sends a small electrical pulse to the TCU. The TCU counts these pulses to calculate shaft speed.
These pulses are superimposed on the constant 0.75V baseline signal. This allows the TCU to determine both whether the sensor is online and how fast the shaft is rotating.
PIWIS displays the calculated shaft speed as RPM, although internally the TCU is determining that RPM by counting the individual sensor pulses.
What Normal Speed Sensor Readings Should Look Like
A common question is what the correct speed sensor readings should be during normal operation.
The basic principle is straightforward.
When a gear is engaged on an input shaft, that shaft is mechanically connected to the rear wheels. Therefore, if the vehicle is stationary, the speed of that shaft should be 0 RPM.
If no gear is engaged on a shaft, the shaft is free to rotate.
Even with the clutch disengaged, residual clutch drag can cause the shaft to rotate at a speed below engine RPM. The exact shaft speed relative to engine speed depends on several factors and will also change with transmission temperature.
Examples
P selected
First gear is engaged on shaft 1. Earlier 987/997 software versions may engage reverse gear instead.
In this condition:
- Shaft 1 should read 0 RPM.
- Shaft 2 should rotate at a speed lower than engine RPM.
R selected
The readings should be the same as when P is selected.
N selected
No gear is engaged, so both shafts are free to rotate.
Both shafts may therefore rotate at a speed lower than engine RPM due to residual clutch drag.
D selected
First gear is engaged on shaft 1, while second gear is engaged on shaft 2.
Because both shafts are now mechanically connected to the rear wheels, both should show 0 RPM while the vehicle is stationary.
Testing With the Vehicle Lifted
It is possible to lift the vehicle and operate the drivetrain with the wheels rotating, simulating normal road operation.
Under these conditions, the transmission shaft speeds should also show rotation.
The shaft connected to the engaged clutch should rotate at a speed equal to engine RPM.
The shaft connected to the non-engaged clutch, with a gear pre-selected, should rotate at a different speed. This difference corresponds to the gear ratio relationship between the gears engaged on the two shafts.
Manually Testing Shaft Speed With PIWIS
It is also possible to rotate the input shafts manually and monitor their speed using PIWIS.
Shaft 1
Testing shaft 1 is relatively straightforward.
Remove the upper black cap at the rear of the transmission.
Install an M24 × 1.5 mm bolt into the visible shaft, which is shaft 1, and rotate it anti-clockwise.
PIWIS should display the resulting shaft speed.
Shaft 2
Testing shaft 2 is more involved.
The rear transmission casing must first be removed.
Engage 2nd gear on shift rod 4 by pushing the lowest shift rod/fork toward the engine.
Then install an M22 × 1.5 mm bolt into the lower shaft and rotate the bolt clockwise.
This causes shaft 2 itself to rotate anti-clockwise, and the corresponding speed should be visible in PIWIS.
A Note About Rotation Direction
If the shaft is rotated in the opposite direction during testing, the TCU will still register the shaft speed correctly.
The pulse generated by the speed sensor is different because it indicates reverse rotation, but the TCU still appears to calculate the speed correctly.
For this type of manual test, it is best to shut down the vehicle while the transmission is in N. This ensures that no gear is engaged and that the park lock is also disengaged.
Speed Sensor Faults and Diagnosis
There are many different speed sensor-related faults. Some clearly point to the speed sensor itself, while others require additional knowledge and testing to determine whether the problem is caused by the sensor or by another part of the system.
As discussed earlier, the speed sensor outputs are continuously compared with wheel speed data to verify that the calculated transmission speeds are plausible.
If a wheel speed sensor is failing and causing the fault, you would normally expect to see corresponding PSM faults related to the wheel speed sensors. If no such wheel speed sensor faults are present, the transmission speed sensor becomes a more likely cause.
Gear Ratio Monitoring Faults
Typical fault codes include:
- 0731
- 0732
These faults relate to gear ratio monitoring.
The TCU continuously compares transmission shaft speed with wheel speed to confirm that the actual transmission ratio matches the expected ratio for the selected gear.
If the calculated ratio is outside the permitted range, the speed sensor may be reporting shaft speed incorrectly.
Transmission Input Shaft Speed Plausibility Faults
Typical fault codes include:
- 1743
- 1744
These faults relate to the transmission input shaft speed plausibility check.
In this test, engine speed is compared with the speed sensor output from the transmission.
If the relationship between engine RPM and transmission shaft speed falls outside the expected limits, the cause may be an incorrect speed signal from the speed sensor.
Clutch Slippage and Clutch Overspeed Faults
There are many fault codes associated with clutch slippage and clutch overspeed, so it is not practical to list all of them here.
Examples include:
- 1745
- 1757
- 1758
- 17B5–17B8
With these faults, actual clutch slippage may be occurring and may be the reason the fault code has been stored.
One of the most useful diagnostic methods is to observe engine RPM while driving.
If clutch slippage is occurring, the engine speed will normally show behavior consistent with the clutch failing to maintain the expected relationship between engine RPM and transmission shaft speed.
If no clutch slippage can be observed during driving, a failing speed sensor is often the more likely cause.
For this reason, fault codes related to clutch slip, overspeed, or shaft-speed plausibility should not automatically be interpreted as proof of a mechanical clutch problem. The speed sensor signal should also be checked before reaching that conclusion.
Clutch Fluid Over-Temperature Faults
Although it may seem unusual, a failing speed sensor is often accompanied by clutch fluid over-temperature faults.
There are two main conditions that can trigger an over-temperature fault:
- The actual clutch fluid temperature is too high.
- The TCU calculates that the clutch steel disk temperatures are too high.
The second condition is particularly important when diagnosing a suspected speed sensor problem.
If a speed sensor channel produces an incorrect or implausible signal, the TCU may interpret the data as excessive clutch slippage. Based on this apparent slippage, the TCU can calculate that the clutch steel disks have reached an over-temperature condition.
In this situation, the clutch itself may not actually be overheating. The over-temperature fault can simply be the result of incorrect speed data from a failing speed sensor.
How to Diagnose the Fault
If clutch over-temperature faults are present, check the following:
- Actual clutch fluid temperature
- Other indications of clutch slippage
- Any additional speed sensor-related fault codes
If the clutch fluid temperature is normal and there are no other signs of actual clutch slippage, the over-temperature fault is more likely to be caused by a speed sensor that is reporting shaft speed incorrectly.
This becomes even more likely if the over-temperature fault appears together with other speed sensor faults.
Neglected eddy current effect
The sensing surface of the OEM speed sensor is made from a plastic material known as PEEK (polyether ether ketone). Over time, this material can deform; however, this has minimal impact on performance. The primary issue is that it becomes more brittle with age, and the component itself is relatively expensive.
Many aftermarket speed sensors on the market use aluminum alloy as the sensing surface. However, a critical factor is often overlooked: the use of aluminum introduces eddy current effects.
Eddy currents can significantly affect the rotational speed signal, particularly at high RPM. Based on calculations:
- At 6000 RPM:
A 0.2 mm aluminum sensing surface results in approximately 10% signal attenuation and about 20° phase lag.
A 0.3 mm aluminum sensing surface results in approximately 20% signal attenuation and about 26° phase lag.
- 8000 RPM, the impact becomes even more pronounced.
- 1000 RPM, the effect is negligible.
To the best of my knowledge, some brands use aluminum alloy sensing surfaces. While they perform adequately under normal driving conditions, they are not recommended for track use, as higher RPM leads to increased signal instability.
Some manufacturers use stainless steel for the sensing surface, which eliminates eddy current effects. However, CNC machining costs for stainless steel are roughly three times higher than aluminum. A potential solution could be a hybrid design—using an aluminum alloy body with CNC machining, combined with a stainless steel sensing surface.
Recommendation:
When selecting a PDK transmission speed sensor, it is essential to verify the material of the sensing surface to avoid potential performance issues.
PopoRacing PDK Speed Sensor Innovative design:
- Integrated assembly - no wiring required
- Dual axis induction design - strong anti-interference, zero temperature drift
- Mid chip layout - signal distortion free, beyond the perception range of the original factory
- Highly integrated automotive grade chip - stable and reliable performance
- 0.2mm ultra-thin Peek window design - eliminates eddy current effects and enhances signals
- CNC aluminum alloy shell - can withstand harsh environments (such as high temperature, vibration, etc.), improving the original sensor's susceptibility to bulging and thermal failure issues
