Robotized Transmission Peugeot 107: Electronics plus Mechanics

Telling about this interesting repair case is much faster than performing the repair itself. Thus, the story will be short in spite of the fact that searching for and eliminating the fault required a lot of time and effort. But the case is so interesting and instructive that it would be useful to add it to your “piggy bank of experience”.

Here we see a Peugeot 107 car with a robotized gearbox. Such a gearbox is essentially the same as the manual transmission, but declutching was servo driven, with a servo controlled by the electronic unit.

Unfortunately, the car’s year of manufacture and the engine make were unknown. The available diagnostic program Lexia does not require such data, and there was neither time nor desire to wait for the owner to find out. We only had a VIN-code: VF3PNCFAC88233201.


The car arrived at a fellow repair shop with a bad clutch disk. The mechanics replaced the disk and cleaned the declutching servo rod. The rod had gone bad to the extent that it was impossible to move it by hand. This was likely the reason for the disk wear. To all appearances, alkaline water and dust from our roads will provide us with a broad field for making money from French cars.

After things were put together, the engine started, but the car did not respond to gear shifts. The display continuously showed neutral “N”. Only the reverse gear worked – and that only with a heavy jolt. The car jumped up and started moving. This is the condition that I found this car in.

I connected a DiagBox scanner.

There were no errors in the ECU memory. I proceeded to actuator control. This is where I found an interesting peculiarity: during declutching, the scanner duly showed the “done” message, but the rod was almost unmoved. Training the gearbox actuators yielded no results, but this will be discussed later. It is hard to escape the conclusion that there is a failure either of the clutch servo or the ECU itself.

We supplied power directly to the servo. It worked quite well; declutching was happening. Therefore, the ECU failed. It is located beneath the dashboard, on the front post of the left door, and looks like this:

I’ll say at once that the photo was taken after car repair and inspection, so don’t be surprised at the absence of fasteners and the already replaced parts.

The ECU connector looks like this:

The left connector portion with solid contacts is the one responsible for ECU power supply and actuator control.

To aid our discussion, we will number the outputs for convenience. Our numbering does not match the numbers in the database, but this is not required:

Outputs 5,6,7,8 are responsible for clutch servo control.

Despite the fact that the ECU requires replacement, there is always a chance to be able to restore it to working order. That’s why we opened the unit. This is the view with the connector looking toward us:

This is the view with the connector looking leftward, respectively:

For the full picture, this is the back side of the card:

We removed the radiator and numbered the elements for convenience:

To control the servo, field transistors are used, which are called metal-oxide-semiconductor field effect transistors in the technical literature, or simply MOSFETs. They have three outputs named G – gate, S – source, D – drain.

So the МOSFET designated with the number 9 had failed. Its make was 2SK3069. Searching on Internet and local electronics stores was rather disappointing. The part could be purchased, but the delivery period was at least two weeks. We found a datasheet for this element and learnt that IRL2505S/L and STP60NF03L transistors are similar in every way to the one we were looking for. Fortunately, we faced no large problems with the IRL2505S/L. Not taking any chances, we replaced the entire group of 4 pieces.

Why did we replace the entire group? This is a crucial issue, and we will take a closer look here.

The clutch servo was connected directly to connector outputs 7 and 8. In turn, inside the ECU, output 7 is connected with D of transistor 10 and S of transistor 4, while output 8 – with D of transistor 9 and S of transistor 3. Therefore, the failed transistor could easily “run down” its neighbour, and we didn’t want to take that risk. In such situations, it is better to spare no expense and time and to replace all the elements.

We refitted the unit in place and re-connected the DiagBox. At launch, the DiagBox requires a VIN. This is not a problem – it is on the arch of the rear right door. We go to the Repair section, Reparation package, “setting the clutch actuator into mounting/dismounting position”, and so on according to the DiagBox Manual.


We go to the Repair section again, Reparation package, “training the actuator of the clutch and gearbox”. Then we again follow all of the DiagBоx prompts. The gearbox training process in my case was quite rough. Something clicked, chattered and moved, but upon process completion, the car worked as though nothing had happened, which evoked almost childlike enthusiasm from everyone present.

This procedure is also required when replacing the gearbox control unit. Somewhat later, after a repair, I tried it for the sake of curiosity. It works.

And something else. It should be noted that exactly the same ECUs (with the same troubles) may be installed in the Toyota Corolla and Toyota Aygo. If you don’t have a DiagBоx scanner on hand, you can use another scanner for gearbox “training”. For example, the ScanDoc scanner also successfully handles this procedure.