club.liberty.asn.au

I've noticed lately at 100% throttle, the acceleration feels a little jerky.

I've done some logging... here you can see where I floor it (white line in top graph)... the boost comes on fairly quickly (green in top graph), and the rpm and engine load climb (top graph) - however the engine load seems to stutter a bit at the time of hitting full boost.

VVT right/left seem fairly well aligned (middle graph) and learned ignition timing I think is acceptable?

Injector pulse width goes a bit strange as full boost is coming on... and the MAF sensor has a few dips in it in roughly the same spots as the dips in the engine load.

I've not done much diagnostic logging before... does anyone want to take a stab at what might be the problem? Slight boost leak somewhere at high boost? Blocked injectors?

Thanks in advance for any help

Kim

Put a 12psi tune on (previous was a 15psi tune) and don't appear to get the same issue.

So for some reason, the jerkiness is only with higher boost conditions... :/

I'm suspecting there is a boost towards 17psi... it looks like the boost spike is cutting the injectors and causing the roughness with the engine load?

It's been suggested I open up the restrictor pill a little, which will slow the boost down a little and prevent the boost spike...?

I'm no expert but I would not faff with the restrictor until really required, if at all.

Maybe log the wastegate duty cycle too?

Can you be more specific describing the sensation?

Yeah, I'm not that keen to go modifying the restrictor valve just yet. I should have mentioned this is with a ChipTorque Ezy-Flash Pro, with the "15psi 98 octane" tune. The "12psi 98 octane" tune feels very smooth, just not as zippy as the 15psi tune. The 15psi tune did actually feel alright when I first loaded it, it just seems to have developed this stutter after me taking it for a nice spirited drive on the weekend (which is why I suspected maybe a vacuum leak or something, rather than the tune itself).

I'll try to describe the sensation... I'll be driving along at maybe 60kmh in 4th gear... then get to the open highway and put my foot to the floor. It reaches boost quickly and accelerates, but rather than being a nice smooth acceleration (with the RPM increasing smoothly), it feels a bit like a there is the occasional splutter - so the RPM increases in bursts rather than linear, and doesn't feel like it's pulling as hard as it should (I guess it's losing it's torque when the revs pause for a second. Not quite as bad as a full-on misfire, and there's no back-firing.
Here's a graph hehe



It did cross my mind that my clutch is towards the end of it's life, and perhaps running a bit more boost is causing it to slip. Although I would have thought that would cause the revs to jump up, rather than level out. I'll generate a new log with the wastegate duty cycle tonight.

I ask because I get a little of the same, barely enough to register on a log, not entirely sure if I can replicate the sensation, clutch at end of life etc.
I just notice it from time to time. Could it be related to the '08 STi fueling stumble?

I'm intrigued, yet its so slight I've never really let it worry me

Apparently if one of the restrictor pills gets blocked or dirty, it can cause this same sensation.

Noticed some oil around my turbo inlet: viewtopic.php?f=18&t=23637 (suggesting blow by and maybe a leak somewhere?).

I'm definitely considering the Pierburg 3 port EBCS for even better boost control over the stock 2 port...

Nasioc Boost Control Sticky
Rexnet: no pill for Pierburg

Inlet and TMIC overhaul pending too.

It couldnt be too hard to remove the section of hose with the restrictor for cleaning it?

Yeah, probably much better control with the 3 port boost control. Was looking at the Grimmspeed ones earlier...

If you log and plot rpm vs vehicle speed it should reveal if slipping clutch...

It was the GFB GT Hybrid BOV that convinced me that aftermarket boost control is a good thing. No more spikes.

Actually it was Bruce that suggested it

Hmmm, interesting clutch slip stuff on this page: http://performancetrends.com/drag_racing_datamite.htm

It's a possibility I guess...

I did a bit of logging with the ECUdataScan on the way home - up a hill which seemed to make the issue more noticeable.

I did a few runs, some felt worse than others. Comparing the logs, the only difference with the runs that felt the worst was that the dip after the boost spike was bigger... which makes me think it's a boosting issue... either a leak (and the pressure drops), or the tune not quite right. The tune is a 15psi tune and it's spiking 2psi over. Could a leak cause a boost spike? The log below felt particularly jumpy - I think because it has 2 boost spikes?

Ignore the jagged RPM/speed line - I think thats just the ECUdataScan being annoying, as all the logs looked like that regardless of how bad the issue was.

Also here a graph of the first log (which was done with the EzyFlash pro) - this time with the Wastegate Duty Cycle included. I'm not really sure if this tells anything more?

I've taken readings from the various points on the graph, and here is what is going on:

The duty cycle jumps up when I floor it, and peaks when the boost is at 6psi before the duty cycle starts dropping back (does that mean the stock wastegate actuator opens up at 6psi?)
The duty cycle backs off rapidly then levels out, with a boost spike of 17psi occurring then levelling at around 15psi (which is the target boost in the tune).

Is the duty cycle a setting that is tweaked when a tune is done? Perhaps it's set to high and causing the boost spike?

Returned the ECU to the stock tune, and I can just feel the same jerkiness on acceleration - just not quite as noticeable as the quicker tunes...

Just because there's no such thing as too many graphs - heres a graph of the stock tune... flat road, foot to the floor. Here is the RPM (red line) and vehicle speed (blue line)... just trying to determine clutch slip. Not really sure what this shows though... other than slightly jerky acceleration...