And... that kind of confirms the diagnosis I put above
Told you so!
Look at the end of part 4. Tranlated transcript:
Well, this turned into a full documentary — I wanted a short video!
Anyway — if you’re working on this system, you already know these Fiat engines are very picky.
After working on timing, the check engine light may flash, it needs a timing belt “relearn” procedure.
When I first started it, it ran terribly — I thought I broke the MultiAir, but no — it was fine.
The real issue: the crank sprocket I replaced had the keyway 2 mm off (cheap Chinese part). That threw off timing and caused RPM limit and misfire.
Also, the timing tool I bought was bent — total mess.
In the end, I timed it by eye, and it worked perfectly.
Moral: make sure you know what you’re doing. The MultiAir system is fine — the car now runs beautifully, the customer is coming to pick it up.
See you in the next video. By the way — I post my daily workshop work, including this one, on Instagram and Facebook: Mecánica Sin Tontería.
Follow me there.
So he basically disassembled the multiair unit, cleaned and polished it, renewed the worn O-rings, renewed the liquid gasket seal.
However, the four "oil bassins" shown are not high pressure oil accumulators. They are low pressure. Even if the old seal was porous in some places, this shouldn't have much impact. Oil is not compressible, therefore those accumulators don't act as some sort of "hydraulic spring", such a thing doesn't exist. Oil in the multiair is replaced all the time,
at the low pressure level (the oil pump of the engine primes oil into the multiair if the engine is cranking/running)
and at the high pressure level: after each valve lift, the very small volume in the high pressure compartment in the hydraulic valve lifters is replaced.
How could a multiair unit go "bad" and
a) throw codes
b) not throw any codes?
Let's discuss a) fist: the solenoid timing could be off. This timing is controlled by the The Magneti Marelli 8GMF control unit
indirectly using a time/current signature.
This unit knows crank position (CKP) and cam phase (CMP). It has calibrated maps:
“At X rpm, Y load, we want this intake profile ⇒ fire solenoid at angle A, release at angle B.”
From that alone you can run fully open-loop: energize at a given crank angle, de-energize at another, assuming the solenoid & hydraulics behave as characterized. Actuation of the solenoid creates a current signature.
The coil inductance and current vs. time change noticeably when the armature moves.
What the ECU/driver does:
- Apply a peak & hold style drive (high current to pull in, then reduced current to hold).
- Measure current through a sense resistor / driver feedback at high resolution.
- Detect the change in current slope that occurs when: the armature starts to move, the magnetic circuit “closes”.
From this, infer: “Valve actually energized now” (or not), approximate electromagnetic switching delay relative to the command.
This is a standard automotive trick: many ECUs use coil-current shape to confirm injector or VVT solenoid operation. MultiAir leans heavily on that: one calibrated component, one known hydraulic system, one precise driver.
So in practice: there's no separate sensor inside the solenoid which would give feed-back on the actuator timing, it’s self-sensing via inductance / current. Therefore,
if the current signature is off, the Magneti Marelli 8GMF control unit throws a code.
have a look here in order to understand the multiair:
https://www.schaeffler.com/remoteme...loads_11/Schaeffler_Kolloquium_2010_18_en.pdf
The precision of the opening and closing angles of the engine valves is essential for system function. The switching time precision of the solenoid valve makes a considerable contribution in this regard. During the assembly of the solenoid valve and their subassemblies, various function values such as flow and switching times are measured on the assembly line and the assemblies are adjusted in such a way that the function values are within the required range. This means that the manufacturing tolerances of the individual parts can be compensated by sorting them accordingly. Despite compensating the tolerances of the individual components, it is still necessary to optimize the precision of the switching times by means of an appropriate compensation function. This compensation is active during the entire lifecycle of the product and therefore also counteracts changes in switching times caused by aging. This ensures optimum balancing of the cylinder charge of one engine during the entire lifecycle.
First of all, the on and off times of each solenoid valve must be mentioned here. They are individually monitored using the current curve during each switching process for each cylinder and then readjusted depending on the operating condition using data maps in the engine control system. The special challenge in this case is the detectability of the current curve over the entire required temperature range and the oil viscosity associated with it. All solenoid valve components must be perfectly matched to each other to ensure this function.
b) How could the multiair unit fail without throwing any codes?
Simple: the actuator signature must be within parameters (because otherwise it would throw codes). The temperature sensor of the multiair should not be disconnected (open) or have a short. What if there's no oil in the multiair, wouldn't that change the current profile of the actuator? I believe it would, as there would not be high pressure in the high pressure part against which the solenoid pushes it's valve in order to close the connection between the high pressure part and the low pressure accumulator. The generation of a code might be suppressed in this particular situation, since it's also normal behavior if the car was "on exposition" on a used car dealership for months without being driven. The oil level in the multiair would drop and the car would not start immediately when cranked. Throwing a code might complicate things for the seller. I actually don't remember if it still throws a code but deletes it when the car finally starts.
Otherwise I don't really have an idea how the multiair could fail without throwing codes. Some mechanical failure perhaps in the linkage?
Anyway, I'm still not convinced that revving the engine suddenly because of mis-shifting the gear, but
without redlining would damage the multiair unit and prevent the car from running. There just isn't that much stuff in the multiair which could be damaged by high revs. I believe the multiair can actually rev higher than redline, but it would need stiffer springs for that. Therefore I'd rather expect mecanical damage somewhere else, like timing belt, and so on. Or because of the unusual vibration something electrical got disconnected or shifted in place.
He's hilarious!!!
Summary of the case
Step-by-step reasoning
What happens in an over-rev on a MultiAir engine
Evaluate the DTC pattern
Compression zero on all four
Likely root causes given those clues
What to check next (in logical order)
How this aligns with diagnostic code behaviour
So yes — your reasoning that this situation might show only cam-sensor codes is consistent with a dead MultiAir brick or a slipped exhaust cam.
My short list of top suspects
Additional suggestions
Bottom line