M40b16 engine cutting on aceleration

[minesapint]

I’ve been chatting with my friend GPT about your problem & he suggests this……..

Given the facts now confirmed:
• BMW M40B16
• Manual
• Cuts cleanly at ~4000 rpm and above
• Also happens free-revving
• New crankshaft position sensor
• New ignition coil

There are only two realistic remaining causes on this engine. One is by far the most common.

⸻

MOST LIKELY CAUSE: Ignition Power Stage (Ignition Control Module / ICM)

This is the number one cause of a hard 3.8–4.2k rpm cut on M40 engines once CPS and coil are ruled out.

Why this fits perfectly
• The ICM switches the coil at high speed
• At ~4000 rpm the switching frequency becomes high
• A failing ICM:
• Works perfectly at low RPM
• Cuts spark instantly at a repeatable RPM
• Causes a clean cut (no misfire)
• Engine immediately drops to idle and recovers

This is textbook M40 behaviour.

Key point

A bad ICM will do this even with a new CPS and new coil

Resistance tests do not reliably detect ICM failure.

⸻

How to confirm (best test)

Freeze-spray test:
1. Run engine until it cuts at 4000 rpm
2. Spray the ignition module (aluminium heat-sink unit near the coil)
3. Immediately try revving again

✔ If it now revs past 4000 → ICM is faulty

Substitution with a known good unit is the definitive test.

⸻

SECOND POSSIBLE CAUSE (LESS COMMON): Throttle Position Switch WOT contact / wiring fault

Not “TPS worn”, but WOT contact shorting or mis-signalling.

Why it can cause this
• At high throttle angle the WOT contact closes
• If the switch or wiring shorts:
• ECU momentarily cuts fuel/ignition
• RPM collapses
• Engine falls back to idle

Quick isolation test
• Unplug the TPS
• Free-rev engine

✔ If it now revs cleanly → TPS or wiring fault
If unchanged → TPS is not the issue

(Unplugging it is safe; ECU defaults.)

⸻

What it is NOT (with certainty now)
• Fuel pump or filter (would not cut free-revving)
• Rev limiter
• ECU (extremely rare)
• Lambda sensor
• Idle valve
• CPS or coil (already new)

⸻

Final Diagnosis Probability

Component Probability
Ignition power stage (ICM) -80–85%
TPS WOT contact / wiring -10–15%
AFM signal dropout <5%

Still trying to help.
Cheers

[blood69]

Hi, thanks for your help.
The ICM on Bosch Motronic Systems is a transistor inside the ECU that is controled by the CPU depending on the signals from diferent sensors, mainlythe CPS sensor. It's not an external module.
I've tried 3 different ECUs already, they all work on another e30s.
About the TPS, I've tried with the TPS disconnected and still have the same issue. If I disconnect the TPS in other E30 there's no issue whatsoever other than the WOT map will not engange.

[Speedtouch]

Is it possible that the ignition coil has developed a fault, even though you replaced it last year? Also worth measuring the resistance of the HT leads to ensure they're within spec.

Do you have the CID lead fitted?
https://www.e30zone.net/e30wiki/index.php?title=CID

It's also worth checking out the main engine loom connector (C101):
https://www.e30zone.net/e30wiki/index.php?title=C101

[blood69]

Is it possible that the ignition coil has developed a fault, even though you replaced it last year? Also worth measuring the resistance of the HT leads to ensure they're within spec.

Do you have the CID lead fitted?
https://www.e30zone.net/e30wiki/index.php?title=CID

It's also worth checking out the main engine loom connector (C101):
https://www.e30zone.net/e30wiki/index.php?title=C101

Hi, thx for your reply.
Thr CID is not fitted since 2017 on this car, and I never had any issue.
The C101 I cleaned the conector and test the CPS signal but did not test the other connections.
The Coil I've tested out with an old Bosch working coil and it does the same behavior.

[Speedtouch]

Hmm, new engine time, perhaps? How many miles has it done?

[minesapint]

Been chatting with my pal again……….
⸻

AFM failure (why it matches perfectly)

When AFM signal drops out at a certain flap angle:
• ECU suddenly “loses” load info
• Injector pulse collapses
• Spark is momentarily invalid

That causes an instant, clean cut at a specific RPM, regardless of load.


⸻

Bottom line
• Bad AFM or ECU power = hard RPM wall

Your symptom profile is classic AFM carbon track failure or DME relay voltage drop.


Now I know you’ve “swapped out” (as you say) your AFM, but are you certain the one you tried was a known good one?
Swapping out parts is not a good fault finding technique, Have you examined the track on your AFM, the pick-up arm can be moved slightly to wipe a different area of the carbon track, best way to test this is not swap but get your meter on it, chase the fault!

Cheers

[blood69]

Been chatting with my pal again……….
⸻

AFM failure (why it matches perfectly)

When AFM signal drops out at a certain flap angle:
• ECU suddenly “loses” load info
• Injector pulse collapses
• Spark is momentarily invalid

That causes an instant, clean cut at a specific RPM, regardless of load.


⸻

Bottom line
• Bad AFM or ECU power = hard RPM wall

Your symptom profile is classic AFM carbon track failure or DME relay voltage drop.


Now I know you’ve “swapped out” (as you say) your AFM, but are you certain the one you tried was a known good one?
Swapping out parts is not a good fault finding technique, Have you examined the track on your AFM, the pick-up arm can be moved slightly to wipe a different area of the carbon track, best way to test this is not swap but get your meter on it, chase the fault!

Cheers

With the AFM disconected it happens the same. On other e30 I disconected the AFM and it doesn't cut.
I also tested the AFM and have smooth voltage throughout the carbon track.

[blood69]

Hmm, new engine time, perhaps? How many miles has it done?

There is no way to know the milage of an E30
The cluster always fails. This one was busted 7 years ago but I fixed. It have now 155000 km on the cluster.
3 or 4 months ago this engine was open to change a failed headgasket. The engine was ok and strong until now.

[Speedtouch]

Have you carried out a compression test?

[blood69]

Have you carried out a compression test?

Unfortunately I haven't. I don't have the kit. I'm gonna try next week to send the car for a repair shop. And let the guys have some fun

[minesapint]

You’d not mentioned about the head being off!

Asked my pal if this could be a possible cause……….
Yes — it absolutely could have, and on an M40 there are very specific ways a head-gasket job can create exactly the fault you’re seeing, even months later.

This doesn’t contradict the AFM theory either — the two can be linked.

⸻

How a head gasket job can cause a hard 4k cut on an M40

Engine ground strap disturbed or left marginal (VERY common)

During a head job:
• Intake off
• Engine harness moved
• Ground straps often loosened or refitted poorly

Why this matters
• ECU, AFM, and CPS share ground reference
• At higher RPM → more electrical noise → ECU loses signal → hard cut

Classic sign
• Runs perfectly until a specific RPM
• No fault codes
• Repeatable limit

What to check
• Main engine-to-chassis ground (near engine mount)
• Ground eyelets on intake manifold
• Clean to bare metal, retighten

⸻

AFM boot / intake leaks introduced

After head removal:
• Rubber intake boot refitted
• Small cracks missed
• Vacuum lines swapped or left loose

Why this shows up at ~4k
• At higher airflow, unmetered air overwhelms correction
• AFM signal goes out of range → fuel cut

Check carefully for:
• Split intake boot underside
• Brake booster line
• Idle control hoses

⸻

Timing belt installed one tooth out

This one is critical.

If the belt was installed one tooth retarded:
• Engine runs “fine”
• Revs cleanly but hits a wall around 3.5–4.5k
• No codes
• Feels like a limiter

M40 is interference
If belt age or install accuracy is unknown → verify cam timing now.

⸻

AFM carbon track wear accelerated

If the engine was overheated before the HG:
• AFM track often already worn
• New airflow characteristics after repair push flap into the worn zone

So the HG job didn’t “break” the AFM, but it made the failure obvious.

⸻

TPS misadjusted after refit (secondary effect)

Not your primary fault, but:
• If TPS never hits WOT, engine feels worse at high RPM
• Makes AFM-related issues more noticeable

⸻

What I’d check first given the HG history (in order)
1. Verify cam timing (do not skip this)
2. Clean and re-secure all engine grounds
3. Smoke test / inspect intake boot and vacuum lines
4. Unplug AFM test
5. Swap DME relay

⸻

One key question (important)

When the head gasket was done:
• Was the timing belt replaced?
• If yes — by who (DIY vs shop)?
• Any overheating before the HG failed?

Your answers will tell me whether this is an AFM/ground issue or a timing issue, and which one to chase first.

Cheers

[Speedtouch]

Nice summary of diagnosis! I recently had to correct the timing on my 2009 Abarth Grande Punto after a previous owner/garage fitted the cambelt one tooth out, causing the EML light to keep coming on, even though the engine ran surprisingly well - but even better after I'd corrected it.