An iDrive controller that stops answering leaves the screen frozen on whatever page it was showing, and the knob feels dead in the hand. On an E90 the controller feeds a head unit in the dash, and on an F10 the same part sits under a larger trim with a touch surface on top. The checks below tell you whether the knob or the unit behind the screen has failed.
The screen shows the menu and will not move. The radio stays on one station and turning the knob does nothing at all. Some drivers report that the system worked after the car stood overnight and failed again once it was warm. Others say a drink went over the console a week before the fault. Write that history down, because the controller sits right beside the cup holder on both models.
Ask whether the screen has ever gone completely black, since a dead display points at the head unit rather than at the knob. Ask whether the steering wheel buttons still change the volume, because that proves the head unit is awake while the knob is not. Note the build date from the label on the door pillar so the right software is used for the test. Ask also whether the car has been jump started lately, since a spike on the supply can leave the head unit in a state it does not recover from until the battery is disconnected.
Switch the system on and watch the screen while you press the controller down. A screen that changes page and then freezes has a live head unit and a failing input. A screen that never changes at all, even with the steering wheel buttons, has a head unit that has hung or lost its supply, and the knob is then a witness.
Turn the knob slowly through a full revolution with the engine off and listen for the detents. A knob that turns with no feel or that grates has a broken encoder inside, and the fault is mechanical before it is electrical. A knob that feels correct but does nothing sends you to the plug and the bus.
Open the head unit address with the manufacturer software and look at the input page. With the system awake, the page carries a counter for the knob and a status for each button around it. Turn the knob and watch the counter. A counter that moves while the screen does not is a display or software fault, and a counter that never moves is the controller or its wiring.
Check whether the controller appears in the module list at all. A controller that answers on the bus is alive and its supply is good, so the fault sits in the encoder or in the button pad. One that does not answer has no supply or has lost the bus, and you go to the plug with a meter.
Lift the console trim and pull the plug from the controller. With the ignition on, measure the feed on its pin against the body earth, then measure the earth pin against the battery post. You want the feed to hold near battery voltage while a lamp is used as the load, since a meter alone will read a good voltage on a circuit that cannot carry current.
Look into the plug and the socket for a film of dried drink. Sweet liquid leaves a sticky residue that bridges the pins and it also corrodes them. Clean both sides with contact cleaner and a soft brush, then dry them and test the knob before you go any further. A controller that comes back to life after a clean needs no part. Write the pin numbers you cleaned on the sheet, because the same pins will fail again if the source of the spill is not dealt with.
Scope the data line at the controller plug with the ignition on and the system awake. You want a clean signal with sharp edges and a steady rest level. A line that sits at one rail, or one that shows rounded edges with noise, has a wiring or a module fault further up the bus, and the controller then reports nothing.
Wiggle the loom at the back of the console while the scope is connected. A trace that breaks as the loom moves is a broken strand, and it usually fails where the harness passes the hinge of the armrest. Repair the wire with a soldered joint and a heat shrink sleeve rather than with a crimp, because that spot flexes every time the console is opened. Compare the trace with the same line at the head unit plug, since a signal that is clean there and poor at the console puts the fault in the loom between the two.
With the controller on the bench, power it from a stable supply and watch the output while you turn the shaft. The encoder should give a clean pair of square waves that change in step as you turn. A channel that stays flat or that misses steps has failed, and the part is replaced rather than repaired. Watch the supply at the same time, because a unit that drops on each step gives a clean trace on a bench supply and fails once it sits in the car.
Press each button in turn and watch the status line in the tester. A pad where one button never answers has a worn membrane or a cracked track under it. On the F10 the touch surface on top is a separate layer and it fails on its own, leaving the knob working while the writing and the swipe do nothing.
If the screen never changes and the controller answers on the bus, test the head unit. Read its memory and look for an internal fault, then check its supply and earth at the plug in the dash. A unit that reboots in a loop has a supply that drops when the amplifier draws current, and that shows on a meter as a dip each time the display flickers.
Check the display cable where it passes the hinge of the dash panel. A cable that has been trapped during a repair gives an intermittent picture that looks like a failed head unit. Reseat the connector, then flex the cable gently with the system on and watch for the picture to drop. Look at the screen itself in a dark workshop with the system on, since a backlight that fails leaves a faint image you can see from an angle and a black screen from the driver seat.
A replacement controller needs to be coded to the car so the head unit accepts it. Fit the part, then run the coding step for that address and clear the memory. On the F10 the touch surface has its own calibration, and it is run from the same software page once the knob answers.
Test the car with the engine running and the console trim back in place, since a plug that is not latched will fail once the trim pushes on it. Turn the knob through every menu page, press each button and use the touch surface if the car has one. Then let the car stand for an hour, switch it on cold and repeat the test, because a failing encoder often works while warm.
| Symptom | Likely cause | What is measured |
|---|---|---|
| Knob dead, screen still shows the menu | Failed encoder or button pad | Knob counter in the input page of live data |
| Screen frozen and no response at all | Head unit hung or lost supply | Supply and earth at the dash plug, memory in the head unit |
| Works cold and dies when warm | Broken strand in the console loom | Data line on a scope while the loom is flexed |
| Knob works, writing does nothing | Touch surface layer on the F10 | Button and touch status per area in the tester |
It drives normally, and the steering wheel buttons still cover the volume. Settings that sit only in the menu are out of reach until the knob works.
It needs coding so the head unit accepts it. Fit it first, then run the coding step for that address before you test the menus.
The controller sits beside the cup holder and the liquid runs down into the plug. Residue bridges the pins and corrodes them, and a clean often restores the part.
The knob counter moves in live data while the picture stays still, or the steering wheel buttons change the volume. Either shows the unit is awake and the input is the problem.