Weird cut on boost

I have a problem on one of my friend's car and we try to figure out why.

It's a Nissan S13 dedicated for drifting with a CA18DET, a tad over 1.8L (1853cc), Holset HX35, Toyota 1ZZ COP, 1600cc injectors, Bosch 044 fuel pump and few others things. (I can elaborate the setting if necessary)

Like on the title, the engine have some weird cut on boost (the higher the boost the quicker it happens) with corresponding lean spikes on the logs.

My friend was thinking about the fuel pump that could be too small considering the size of the injectors (and it definitely is for me), but I guess it would lean out instead of cutting.

I was thinking about an ignition problem.

We already tried to reduce the gap on the spark plugs, but no improvement so far.

Next weekend we will try to increase the dwell (although I'm pretty sure the values are correct) but after that I'm out of ideas...

The log screen correspond to the first 20sec of this video:

https://youtu.be/-1oeKLpEqc4

(he didn't lift the throttle @0:14 -> that the cut)

So the question is did the ECU command the fuel cut, or did it run out of fuel? Do you log fuel pressure? What about Injector pulse width, or duty cycle? That data would help you determine what is happening.

If the ECU is commanding the change, then you need to find out what sensor or other part of the fueling equation led to that change.

If the fuel pressure drops, then you might have a pump, pickup or tank venting problem.

What is it that's getting activated at the start of the video?

If the top left gauge is boost it clearly boost right up once it's going so makes me wonder if there's some kind of speed related limiter kicking in.

Do you have the injector duty logged?

At the beginning of the video my friend was testing the launch control function, but the problem is not related to that.

It happen pretty much everywhere on the log, but it where you can ear it at best on the video.

Yes I do have injector duty cycle and fuel pressure on the log.

I don't think the problem is from the ECU. Could it be something else than the fuel delivery or the ignition ?

I forgot to say that he is running on "pump" E85.

It looks like your throttle position and injector duty are dropping together, are you using TPS as a load axis?

No it's MAP based, and the duty follow the MAP on the graph.

Here is another part of the log, maybe it's easier to read.

Looks like you got some serious timingscatter going on there. Stock cas for triggering? Im thinking it might be timing related

oh, and on the last pic it looks like it goes lean the moment map reaches 296`ish kpa. So that might be a limiter of some sort.

What map sensor is he running?

the map sensor is a Bosch 0281002456 (350kpa max)

The stock CAS has been replaced by a 36-1 trigger wheel and a hall effect sensor.

As for what might look like timing scatter, it's mostly due to the scale and the fact that it is logged at ~20hz I think

Hey Ludo, I know it's been mentioned already but I would be considering the trigger system too. Your rpm trace in those logs is very much what I'd expect to see if the trigger input was inconsistent. This can cause all sorts of weird issues like you're seeing. It's been many years now since I've dealt with a CA18 but if memory serves me correct they are the usual Nissan 360 deg optical which is problematic in many installations, particularly if you have non OE cams fitted.

I've had an almost identical situation on an S14 SR20DET. A 24/1 optical trigger disc totally fixed the issue and the engine picked up around 70 kW atw too.

Yes the original Nissan optical CAS is a nightmare and it have been replaced by the 36-1 trigger wheel + hall sensor for that very reason.

My friend had a very hard crash 2 month ago so maybe the bracket that hold the sensor as moved so we will check that just to be sure.

But I don't see why a trigger problem would happen only on boost ?

For example, when he run with 2bar (29psi) it append almost immediately and the engine can't rev over 5500rpm (so only 500rpm after we reach the boost target)

while with 1.5bar (22psi) the engine rev without any problem till ~7500rpm

I don't know what I was reading as tps, the joy of jumping on here between jobs.

Have you checked the calibration of your map sensor? Seeing if you can activate a limit of some sort by using a regulated or manual air supply and putting 29psi in and above?

It could be a soft limit kicking in for boost cut,are you using fuel cut for boost? What is the limit and control range?

im not clued up on how electro magnetic noise works but ive found with some cars ive tuned the triggers would give problems at higher boost not just rpm , I put it down to the different duty of injectors etc (could be wrong lol) I would 100% check the trigger system

Yes, I checked the map sensor calibration and the min/max I could read with it before the installation.

(with a fridge compressor for the vacuum, a standard compressor for the pressure and a set of Wika mechanical dial gauge)

The upper limit is around 270kpa above barometric pressure (39psig), so around 370kpa absolute (53.5psia)

If the ECU had activated any kind of limit, I would have had a "status" change (Alarm).

Status1: 136 = "RPM Synced" (8) and "Full-RPM sync" (128) (8+128=136) -> if it have a trigger problem it should be Status1: 0

Here is a non exhaustive list of what I could have had:

Spark cut: Status2: 32 / Over boost: Status2: 64 / Fuel cut: Status3: 1 / Soft limiter: Status3: 32 / MAP sample error: Status6: 128

As for the over-boost protection, I will have to recheck, but again I would have had an alarm if it reached that point.

Ludo, that triggerwheel is placed on the cam. That dont help at all when you want to know where the CRANK is at.

I know it's not a Link but on those the soft limit can be reached and it won't throw up an activation notification when on the soft limit, it's only when it reaches the hard limit. I haven't ever hit the boost limiter on an MS system so don't know what it gives as a notification.

If you want to post up a copy of your file I don't mind taking a look at it for you

kickerzx , yes it's on the cam, because we need a signal at cam speed for sequential.

Sure it would have been better with 2 sensors (one on the crank for the speed and one on the cam for the synch) but that is not an option.

Chris, on the megasquirt any soft limit will throw a notification (Status3: 32)

I'have just check, and there is no overboost protection configured at the moment (yes I know, not safe)

What file do you want, the config or the log ?

both would give the full picture if that's alright

No problem for me.

Just in case if you check the alarm (status), I might as well give you their signification. (at least the one you will see)

Status2: 4 = launch in

Status2: 8 = launch active (you will always see it with launch in, hence -> 12 on the log)

Status2: 32 = Spark cut

Status3: 128 = Launch on (spark retard)

Status6: 64 = Fan

4, 5, 7 and 8 are not used

the full list is available here (page 61/62)

http://www.msextra.com/doc/pdf/Megasquirt3_TunerStudio_MS_Lite_Reference-1.4.pdf

Hey Ludo, I might be off track here but I'll add a couple of thoughts anyway and please bare with me as Megasquirt is not a platform I've had much to do with so far.

The trigger situation on the cam can show up related to boost which I know might not seem logical but I've experienced it. What typically happens as a side effect of the ECU losing track of engine position is that the timing scatters all over the place. If you get to a point where the timing is massive,y retarded as an aspect of this scatter, it can feel very similar to a cut.

I've found the 24/1 trigger on the cam works pretty well in stock engines but with large cams this can still be problematic due to resonance in the valve train at certain rpm.

I'd start however by viewing or logging any active cuts and this should prove that you aren't seeing an ECU induced cut for any reason.

Small update.

We checked the ignition advance with a timing light and a fixed timing table -> the timing stay consistent = no scatters (that was expect since we didn't had any warning on the logs)

More dwell (a lot) -> no improvement.

The car is going on the dyno (for the first time) this Saturday.

I'll let you know if we find anything.

check the resistance of your injectors with an Ohmmeter and then look at how your injectors are set up in the ECU. I've had a similar problem not too long ago.

It turned out, that the internal resistance of the injectors was too high, so the ECU had a hard time opening them in time at high rpm. On a side note, it was also related to boost.

Any update Ludo? How did you get on on the dyno?

Honestly, nothing interesting... My friend did go to the dyno but with bad tires so they couldn't do anything.

He had a curve but the only thing you can see on it is where the turbo spool.

We live 500km apart so I can't do anything myself, and most of the time my friend can work on it only on Sunday.