GR Yaris Ecutek, A-cams, & misfires

There was a post on this subject on fb that might provide some direction.

In the post it was proposed that the dual mass flywheel may have failed, causing changes in crankshaft speed and triggering the misfire count.

Your cams may also affect idle crankshaft speed. The 3-cylinder engine doesn't have the smoothest operation even in stock form at low speed.


1771477691371.webp


1771477721614.webp
 
I argue a failed DMF would introduce vibration into the cabin, triggering the sensor at all engine speeds. I would expect it to be noticeable when moving through gears. Likely enough to meet the threshold for a CEL.
 
Flashed new file. No change.

Updated dashboard for knock values. I have no clue what these mean but hopefully they make sense to someone.

FWIW they barely changed from -3.00 for Correction Angle (I saw min of -3.20 and max around -2.96) and 24.1 for Learn Value (only saw this change to 24.0), so I think they're good? 🤞

1000021556.webp
1000021557.webp
1000021559.webp
1000021560.webp
1000021562.webp
1000021566.webp
1000021568.webp
1000021572.webp
1000021574.webp


1000021560.webp
 
Those are stored values for learned timing. If there's knock detected at a certain load and rpm the timing pull is stored as a minus number. -3.00 is probably a default as it's the same throughout. Is there a knock counter?
 
-3 is the real time short term knock adjustment. -3 is the default meaning no advance or knock

24 is the long term knock adjustment. 24 means no knock / ignition retard and your calibration is happy with your fuel.
ACN91 here in California will be around 16-18 on stock tune (ignition retard). Mixing in E20-30ish ethanol will get it back to 24

Long term knock adjust has 2 bands. One for low RPM range, one for high RPM range. You might notice the switch if you do a full RPM pull.
 
Awesome. Thank you both! So it seems mine's barely knocking, if at all, and my misfires seem fairly normal. Which surprises me that nobody's noticed this before, if so!

How'd you get those graphs @tonic ?
 
I use Torque Pro to log signals to a .csv file, which I then convert to Motec i2 .ld format using MotecLogGenerator. But really, any software that can export data to .csv will work.

There’s a bit of manual prep needed before the file is accepted by MotecLogGenerator. Mainly deleting the empty lines and converting the log time to seconds.

After that, I just open the .ld file in Motec i2 and analyze it there.
 
tonic said:
I use Torque Pro to log signals to a .csv file, which I then convert to Motec i2 .ld format using MotecLogGenerator. But really, any software that can export data to .csv will work.

There’s a bit of manual prep needed before the file is accepted by MotecLogGenerator. Mainly deleting the empty lines and converting the log time to seconds.

After that, I just open the .ld file in Motec i2 and analyze it there.
Click to expand...
And you can get much the same with the Ecutek ODB dongle and app, it just exports CSV directly
 
On this topic Litchfield aren't responsive, so i'm health checking my own logs for piece of mind.
  1. Knock Correction Angle, which appears to be the instantaneous knock response from the sensor. The EcuTek dev car has always been based around -3deg and it counts up until it commits that increase into the Knock Correction Learned parameter, when it resets to -3deg and continues to measure knock sensor output and correct the ignition. tis value can be changed, if set to 0 it wont add ignition only ever remove it.

To be clear i've understood -3 is no correction needed. Anything lower e.g -3.1 means real time correction based on possible knock. When this gets down to -4 it starts committing to long term correction. Same applies in the upward direction -2.9 & up gives you timing advance. -1.9 starts adding timing to the long term stored number.

Is that accurate yes/no?
 
You must log in or register to reply here.
Back
Top