Technical really stubborn multi air bleed or failed multi air?

Currently reading:
Technical really stubborn multi air bleed or failed multi air?

I think the OP and you are both jumping
to conclusions.

Now let's assume for a minute that both of you are right and the multiair is "bad", whatever that means.

So one hypothesis is it can't pressurize all of the four high-pressure chambers and/or the actuators don't engage and let that (insufficient) pressure provide the hydraulic link which pushes down on the admission valves.

So basically your saying the admission valves don't open at all.

This is normal if the multiair is lacking oil and hasn't been primed.

But it's not normal otherwise.

However, without removing anything, there's an easy way to test this hypothesis.

Just put a condom over the exhaust and crank the engine. If it aspirates, the condom will blow up. If the engine does not aspirate any air because the admission valves never open, it won't.
I am not jumping to any conclusion. This is what every test is pointing to. If it was not a problem with the MultiAir brick, it would have never been bled dry. The ECU would have disabled the solenoids but not the oil going to the brick, so after clearing that up it should have started or at least had compression. Also, if my intake valves were opening, I would get some kind of compression, not just zero psi. I will go ahead and send high pressure through the oil while it is cranking, but putting a condom on the exhaust is just wild.
 
I think the same. Plus you have to consider how the problem appeared, because you know exactly what that moment was. It was that high revving. That plus all that's shown after does point to a problem inside MultiAir, it just cannot hold pressure anymore.
putting a condom on the exhaust is just wild.
I think that was the purpose of condoms being invented, to allow you to be wild... 😁🤣
 
This is crazy, when I had a look around the forums several years ago in order to research how to diagnose and/or change the multiair, everybody said that the brick can't be serviced or repaired, there are no replaceable parts inside and so on. So opening the multiair would basically doom it.

Now look at this guy (the English auto-dubbing is terrible) but the first video gets interesting at about 22 min in as he rambles on quite a lot

first part

second part

third part

fourth part
 
Wow, the guy in the video is wacko! He made me laugh so hard, watching second part min ~ 3:30 when he tells about mosquitos... 😅🤣😂 He's hilarious!!!
And is great work what he's done there, and great work from you too @Zardo for finding and sharing it. And... that kind of confirms the diagnosis I put above 😁 Told you so!
 
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.
 
Last edited:
What you say is not wrong. But my opinion is that it is the MultiAir, there is the problem. 0 bar compression means those valves don't open at all. That is my guessing, I'm maintaining it. I could be wrong, of course. But I don't think I am wrong.
 
ok if the admission valves don't open at all, or they have incorrect timing which prevents the engine from igniting the mixture, let's list all of the possible causes (I'm brainstorming this):

- no oil in the multiair
- no high pressure in the four high-pressure compartements because four some giant leaks in the four valve actuators (like busted O-rings)
- all four solenoids are faulty and are not actuating
- the four solenoids are not electrically connected
- the cam position is off
- the temperature sensor is off
- the Magneti Marelli 8GMF unit is faulty or is disconnected from the can bus / ECU
- some connection problem with the Earth cable

now how many of these would not throw any codes?
 
But how many of those can occur after high revving?
No oil in the multiair or no high pressure there doesn't throw any code, does it?
 
Should test that little release pressure valve after cranking. Push it down with something small and have one finger as close as you can there. Can you feel any pressure coming out when you push the valve down?
 
As this now clearly goes beyond my own experience, I asked ChatGPT after sharing the history and two Schaeffler documents.

It might be of interest:

🔧 Summary of the case​


  • Manual gearbox → mis-shift (for example, 5th → 2nd) → sudden over-rev by back-driving the engine.
  • Driver says: engine instantly stalled, never restarted.
  • Timing belt apparently intact and marks line up.
  • 0 psi compression on all cylinders.
  • Stored DTCs: cam-position sensor–related, cleared, one returns.
  • Suspect: intake valves never open → MultiAir brick not operating.



🧩 Step-by-step reasoning​


1️⃣ What happens in an over-rev on a MultiAir engine​


During an accidental downshift-overrev, engine speed can exceed 7 000–8 000 rpm mechanically (even if the tach never redlines).


Effects:

  • Hydraulic MultiAir circuit is driven beyond its designed cam frequency (solenoids can’t react fast enough).
  • Oil aeration and cavitation inside the high-pressure chambers.
  • Hydraulic hammer in the pressure accumulators and the small internal check valves.
  • Possible valve float on exhaust or intake.
  • Sometimes the hydraulic lifter brakes (the “slave” units) are damaged or collapse, resulting in permanently closed or open valves.

The outcome can be:


  • Stuck intake valves (closed) → zero air, zero compression.
  • Or bent valves (usually exhaust) if the piston kissed them during float.

Zero compression across all four cylinders means no valves are sealing correctly, or they are all permanently held open/closed by loss of hydraulic coupling.




2️⃣ Evaluate the DTC pattern​


three cam-position codes, cleared, one returns

That implies the ECU sees a plausible crankshaft signal but no coherent phase signal from the cam sensor — i.e. the camshaft is turning but the timing relationship or signal shape is wrong.


Causes that fit:


  • Broken camshaft tone wheel / key slip.
  • Sensor damaged by debris (metal, oil aeration, over-rev shock).
  • Excess camshaft torsion due to seized MultiAir pump element.

If the timing belt still aligns by eye, but the phase seen by the sensor is bad, it could be that the exhaust cam gear has slipped on its hub (the MultiAir drive cam is on that same shaft). That alone can kill valve phasing and compression without a broken belt.




3️⃣ Compression zero on all four​


If all four read 0 psi, that’s almost never four bent valves. It’s either:

  • The intake valves not opening at all (no air drawn → the gauge sees vacuum only).
  • Or the valves stuck open (no sealing).

You can distinguish:

  • No air drawn: gauge needle barely flicks, intake manifold stays in strong vacuum during cranking.
  • Open valves: gauge rises slightly then leaks instantly; you’ll hear air hissing through intake/exhaust.



4️⃣ Likely root causes given those clues​


CategoryProbable specific failureNotes
HydraulicMultiAir unit internally damaged (collapsed accumulator, stuck plungers)Common after mechanical overspeed; causes all intakes closed.
MechanicalSlipped exhaust cam pulley / broken woodruff key on the cam phaserGives cam-sensor codes and incorrect MultiAir timing; possible zero compression.
ElectricalCam sensor itself dead (but engine would still have some compression mechanically)Would cause no-start but not 0 psi. So secondary symptom.



5️⃣ What to check next (in logical order)​


  1. Physical cam timing– don’t trust painted marks:
    • Lock crank at TDC.
    • Use proper cam-locking tools or verify cam flats position.
    • Ensure exhaust cam tone wheel alignment (for the sensor).
    • Check the key/keyway on the pulley.
  2. MultiAir brick feed – remove the oil filter housing plug and confirm oil flows into the MultiAir gallery during cranking.
  3. Internal hydraulic integrity– if possible, remove MultiAir cover and inspect for:
    • Cracked top cover or accumulator (signs of pressure spike).
    • Collapsed pump plungers or blown o-rings.
    • Presence of aerated oil / metal debris.
  4. Electrical sanity– scope the cam sensor output while cranking.
    • If flat-line or nonsensical pattern → sensor or trigger wheel failure.
    • Verify that the MultiAir solenoids get supply (they won’t fix compression, but ensures driver stage OK).



6️⃣ How this aligns with diagnostic code behaviour​


If the MultiAir brick has failed mechanically or hydraulically, all four solenoids may still look electrically fine.

  • The ECU will see plausible current traces (so no solenoid DTCs).
  • Because there’s no combustion, many plausibility monitors never finish.
  • Only the cam sensor error remains → because the ECU can’t reconcile CKP/CMP correlation while the engine spins with no valve events.

👉 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​


  1. Exhaust cam phaser / tone wheel slipped
    → explains the sensor DTC, zero compression, “timing belt looks fine”.
  2. Hydraulic failure inside MultiAir block (loss of pressure / stuck valves)
    → explains uniform zero compression and absence of other DTCs.
  3. Both combined – over-rev caused hydraulic lock and sheared cam key.



🧠 Additional suggestions​


  • Pull valve cover and check if the pump pistons in the MultiAir unit move when cranking — no motion = broken drive or jammed plunger.
  • If you have a borescope, watch an intake valve through a plug hole during crank; if it never moves → MultiAir not actuating.
  • If mechanically OK, scope CKP vs CMP waveforms to confirm phase angle.



⚙️ Bottom line​


Your brainstorming is solid.
From the pattern — over-rev, immediate stall, zero compression, only cam-sensor DTC persisting — the most likely culprits are:


  • Mechanical slippage of the exhaust cam / tone wheel, or
  • Catastrophic hydraulic collapse in the MultiAir unit.

Both can kill compression on all cylinders without triggering clear OBD faults beyond the correlation code.


If this were my diagnostic job: I’d verify mechanical cam timing with tools, then inspect / swap the MultiAir unit before condemning the ECU.
 
It might be of interest:
That is very good and sums it up perfectly.
If you have a borescope, watch an intake valve through a plug hole during crank; if it never moves → MultiAir not actuating.

That is what I've mentioned earlier. Looks like is not a bad idea. @beebe101, do that if you can.
My money's on the o-rings in this one!
 
Also, ChatGPT claims that there is no error code thrown for the multiair in this case since such an error needs the engine idling for some time so the current profile can be assessed. The engine not starting prevents that assessment and therefore no code is thrown even if the brick is actually faulty. I don't know if that's true but when my multiair failed on one cylinder, it sometimes took 100 - 300 m of driving before the (manually cleared) code of the faulty solenoid timing would reappear. Sometimes it reappeared almost immediately after the engine was idling for 5 seconds.
 
Last edited:
Back
Top