740 B204GT, LH2.4, missfire, code 2-1-4...

I tore off old mixer/carb style LPG system off of the car and installed propper computer controlled LPG injection (LPi) system.

During test/mapping runs, car would drive fine on petrol, but would hit a wall around 4100 rpm with something that would seem like a mild spark cut/rev limiter. Switch back to petrol, and it would rev past it. However, it doesn't happen when revved in neutral.

After Alfa RPM sensor fiasco, I repaired wiring on the old one and put it back in until new one arrives. Plaatic part that holds mounting hole was broken, so sensor was basically unsecured in its hole, so I reckoned it hits some resonance at 4000 and sends erratic RPM signal, which fouls spark enough for LPi, but petrol seems to cope with it. Today I put a new sensor in, secured it well...ALAS, problem persists. Same issue, still gives code 214 - erratic RPM signal, osciloscope logs on LPi interface show crazy jumps in rpm signal around 4000 rpm under load.

Tonight I've been driving around on petrol alone, first I would notice light hesitation around 4000, but it grew worse and worse, now it happens almost same as on LPi, it shudders, hesitates, pulls noticably worse than it should, stumbles, misses, etc. Sometimes it it gets better and pulls to redline, sometimes it does rev-limiter like thing, etc. It is most noticable when going into boost, doesn't in vacuum.

Things so far:
- No air leaks
- Known good cap, rotor and plug wires off of a B204FT engine
- Cleaned grounds
- Ignition amplifier/Power stage replaced with a known good one from my crashed LH2.4 a couple of days ago because original was soon diying with intermittent no start, erratic run, etc.
- Radio relay seems fine.
- ~15 days old Bosch plugs

No idea what next to look at. Any ideas?

No more code 2-1-4, but it still does it. It does go into full boost, but doesn't pull as it should. Sometimes it stumbles, even backfires at ~2000rpm, especially up a hill - stumble, no power at all, then it bites and pulls almost hard enough. Usually no stumble and lack of power in first gear on flat road...

Now code 3-4-4 again...that's EGT sensor, but I've noticed it's throwing it every time something TCU related is wrong, or just at random when something goes wrong...

No more missfire and code 3-4-4, but no boost above base wastegate setting as well. I'm out of ideas...


Code 111 on both A2 and A6 terminals. However, on A5 (what's that for, anyway?) I get codes 342, 423, 431 and 421, which I can't find anywhere on the net. Anyone?

I seem to remember those exhaust temp sensors (located in the back of the exh manifold near cil4 exh port) have a tendency to fail .
They are very expensive to replace and almost unobtainable.

Once that sensor is no longer providing the TCU with a decent Exh temp signal, the TCU will not allow high boostlevels.
If you ask me that whole 'add-on' TCU system is a pain in the ass.
Maybe you can just install an aftermarket EBC that is not based on exh gas temp? That way you can bypass that whole TCU system and that dreaded exh temp sensor.

Yeah, from what I've read on them, they seem to fail always... New one costs around 1.000?!
Though, I remember reading something about EGT sensors from VEMS.hu being compatible, but nothing specific. I hope someone with B204xT engine will chime in...

Anyway. Low pressure issue was because ofa melted hose going from intake to TCU's MAP sensor. In a really difficult to see place. After replacing that I again have full boost, but issues under load. When revved in neutral, it will happily rev all the way to redline. Or on wet road when tires break loose. Or if I'm really gentle on throttle, and don't allow it to build boost, it will get to redline with no hiccups whatsoever.

However, as soon as I get into full boost around 3k rpm, or in best case around 4k rpm, I run into a wall, something weird happens to ignition, car starts jerking and bouncing around between 3 and 4k rpm.

It's driving me crazy. Reset codes yesterday, today diagnostics again says this:
A6 (ignition) socket - clear
A2 (injection) socket - code 344, EGT signal missing or faulty.
A5 (TCU) socket:
3-4-2 A/C blocking relay. Signal faulty.
4-2-1 Pressure sensor. Faulty boost pressure.
4-2-3 Throttle position sensor. Signal faulty.
4-3-1 Temperature sensor. Signal missing.

That's 4 out of 8 possible TCU codes. LOL

- AC blocking relay? Where is it?
- Faulty pressure? How? Codes have been reset after replacing intake-to-MAP line.
- TPS. Put a known good one in. Does the same thing as it did with the old one.

All of this weirdness leads me to believe the problem is actually with grounds. However, I have cleaned all the grounding points behind headlights, on both A-posts and intake manifold.
I have no idea what is going on. Car is OK to drive till 3500rpm and that's the way I drive it for now, but it just drives me insane.

BTW, it also doesn't happen with TCU unplugged. Leads me to believe:
a) That way TCU isn't messing with FCU and ICU, so everything is fine
b) It's boost related, hence doesn't show with no TCU meaning no real boost

No idea. I could just unhook TCU forever and run an aftermarket EBC or hech even MBC, which I will eventually do if all else fails, but I'd really like to get this TCU thing working, since it is really a great and idiotproof system when working...

For a moment I thought of some weird spark blowout on less-than-1000km-old Bosch plugs, so a set of new NGK BPR6ES later, we can rule that out as well.

Same old, same old...

Aded another battery to chassis ground for the lolcats and ground intake manifold directly to the battery as well. It did seem better for a moment but at the end it seems like i just got a steady 4500rpm limiter, still builds full boost. Still clean on A6, EGT code on A2 and all four TCU codes mentioned before on A5...

EGT sensor

Sounds to me like your EGT sensor is shot. It uses a K type thermocouple that would be generally available but I didn't get one to work on the electronics. It would need calibration but the circuit is difficult to fathom let alone do the calibration. What I know is the output is an open collector pulse width modulated signal. To test the sensor, apply 12V and ground, use a pull up resistor of 1k from signal to 12V and use a scope to look at the signal. You'll see a square pulse if the electronics work properly. Heat the tip of the sensor and you should see the duty cycle of the signal change. Close to 100% duty cycle equals some 900 deg C.
I ended up getting a used EGT sensor, the second one worked.

This is what the EGT electronics looks like on the inside after cleaning out all the goo. Some chips there that seem custom made. Probably an early Bosch development?